Shortest conclusion
Recent human proxy advances are real, but they do not yet compose automatically into same-subject state-complete readout. The reason is not only missing modalities. It is also that current routes measure different quantity types, on different spatial and temporal units, in different cohort / physiological regimes, with different model and hardware burdens.
The public pages had already become stricter about human proxy families, but this composition page was still lagging behind the site's newer taxonomy. It still let readers compress mixed arousal proxy / receptor-transporter atlas prior / occupancy PET / challenge-linked displacement-release PET into one generic neuromodulatory row, five-metabolite 1H-MRSI similarity and high-resolution 1H-MRSI metabolite-distribution mapping into one spectroscopy row, 31P metabolite / pH balance, 31P MT exchange-flux, 31P NAD-content mapping, and 31P functional NAD-dynamics into one row, deuterium metabolite-mapping / absolute quantification into deuterium kinetic-rate imaging and then back into the same energetic row, myelin-water / MT-family / bilayer / qT1 remyelination-sensitive routes into one myelin meter, BBB water-exchange MRI into tracer-specific BBB PET transport, blood-CSF barrier / choroid-plexus perfusion / blood-to-CSF transport / DCE water cycling / apparent BCSFB exchange / simultaneous BBB-versus-BCSFB ASL exchange into one generic barrier row, SMBT-1 first-in-human target validation / SMBT-1 AD-spectrum disease context / SMBT-1 brain quantification / SMBT-1 whole-body biodistribution into one generic MAO-B row and then that same MAO-B row into I2BS astrocyte PET, TSPO disease-context / validation-bounded PET / CSF1R route-setting PET / COX-2 enzyme-defined PET into one generic immune PET row, macroscopic CSF oscillation / parenchyma-CSF water exchange / respiration-conditioned net-flow / exercise-conditioned contrast influx / intrathecal clearance / CSF mobility / biomarker efflux into one generic clearance row, and still left too much room to read SV2A PET as one generic synaptic-density row. The current primary literature does not support those shortcuts. This page now fixes the composition rule at the same granularity as the rest of the site.
This central composition page was also still too coarse inside the neuromodulatory lane. The current primary literature does not support reading human neuromodulatory evidence as one reusable state row. Carro-Domínguez et al. (2025) constrain a mixed arousal proxy in human sleep, Hansen et al. (2022) constrain a regional receptor / transporter atlas prior across more than 1,200 healthy individuals, and Nakuci & Bansal (2025) show how those atlas maps act as a modeling scaffold for spontaneous BOLD activity rather than a same-subject current-state readout. Wong et al. (2013) and Schlosser et al. (2025) then constrain administered-drug target engagement or non-engagement under named tracers and doses, while Koepp et al. (1998), Erritzoe et al. (2020), and Miederer et al. (2025) constrain challenge-linked dopamine or serotonin release proxies within bounded regions and time windows. These papers do not share one direct observable, one evidence role, one temporal object, or one safe bundle role. On this site, neuromodulatory evidence therefore has to be typed by proxy class, target or challenge, spatial scope, and time window before proxy composition is judged at all.
The site had already fixed astrocyte PET on front-door pages, but this central composition page was still one step behind. Villemagne et al. (2022) establish first-in-human SMBT-1 MAO-B target validation, Villemagne et al. (2022) establish an AD-spectrum disease-context contrast, Hiraoka et al. (2025) establish a brain-quantification problem, and Mesfin et al. (2026) establish a separate whole-body biodistribution burden. Matsuoka et al. (2026) then establish a separate SL25.1188 simplified / arterial-free AD quantification route, while Best et al. (2026) establish a severity- and smoking-conditioned SL25.1188 AUD route. Tyacke et al. (2018) then establish an I2BS route rather than an MAO-B route, and Jaisa-Aad et al. (2024) further show that even within MAO-B-linked interpretation, tissue pathology still changes the safe ceiling. On this site, astrocyte PET therefore has to be typed by target, tracer family, and route role before proxy composition is judged at all.
This central composition page was still too coarse inside the barrier-side human lane. The primary literature does not support reading blood-CSF barrier / choroid-plexus evidence as one reusable human row. Zhao et al. (2020) and Sun et al. (2024) constrain a choroid-plexus perfusion route, Petitclerc et al. (2021) constrain a blood-to-CSF water-transport route, Anderson et al. (2022) constrain a DCE water-cycling route, Wu et al. (2026) constrain an apparent BCSFB-exchange route with scan-rescan repeatability, and Petitclerc et al. (2026) constrain a simultaneous BBB-versus-BCSFB ASL exchange route that explicitly estimates both Kbl→GM and Kbl→CSF. These papers do not share one direct observable, one crossed boundary, one transport model, or one safe bundle role. On this site, blood-CSF-barrier evidence therefore has to be typed by route family, carrier / transport object, and validation ceiling before proxy composition is judged at all.
The route-family split above was necessary, but it still left one composition shortcut available. The current primary literature does not support reading barrier-side human papers as if they already played one interchangeable bundle role once the route family had been named. Morgan et al. (2024) are primarily a method-family non-equivalence warning inside BBB water-exchange MRI, not a healthy-reference or disease-burden row. Padrela et al. (2025) are primarily a healthy-adult lifespan reference route, while Padrela et al. (2026) are primarily a disease-burden contrast route for early cognitive / cerebrovascular load. On the BCSFB side, Zhao et al. (2020) are an early route-setting perfusion study, Sun et al. (2024) are a healthy-aging extension, Wu et al. (2026) are a repeatability anchor, and Petitclerc et al. (2026) are a simultaneous boundary-separation / model-comparison route. A third barrier-side role now has to be kept separate as well: Farinas et al. (2025) are a large-cohort paired-fluid protein-balance route whose ratios can reflect transport across brain barriers but also synthesis or degradation in either compartment, so the route is not doing the same job as water-exchange MRI or transport PET. These are not the same bundle job. On this site, barrier-side human evidence therefore has to be typed not only by route family and transport object, but also by route role / evidence role before proxy composition is judged at all.
This central composition page was still too coarse inside the human immune lane. The primary literature does not support reading target-defined human neuroimmune PET as one reusable human row. Biechele et al. (2023) show why TSPO is not a universal human activation-state meter, Wijesinghe et al. (2025) constrain a TSPO disease-context / validation-bounded route in PSP, Horti et al. (2022) and Ogata et al. (2025) constrain first-in-human CSF1R PET routes, and Yan et al. (2025) constrains an enzyme-defined COX-2 route with celecoxib blockade. These papers do not share one direct observable, one target class, one validation ceiling, or one safe bundle role. On this site, human neuroimmune PET therefore has to be typed by target class and route role before proxy composition is judged at all.
A local causal paper and a living-human proxy paper do not automatically add up to measured human controller state. On this site, causal relevance and living-human observability are different axes, so the bridge has to be disclosed rather than implied.
This page stays on the technology and natural-science side only. It does not discuss philosophy, law, personhood, or policy. The question here is narrower: what do current human routes directly observe, what do they infer through a model, and what do they still leave latent?
One remaining shortcut was still too permissive for the current site rule. A proxy bundle could still cite SV2A PET as if that label already fixed one reusable human row. The primary literature does not support that compression. Naganawa et al. (2021) constrain the quantification route itself, Johansen et al. (2024) constrain a healthy-human atlas / baseline route, Snellman et al. (2024) constrain a disease / risk-contrast route, Shatalina et al. (2024) constrain a task / cognition association route, Smart et al. (2021) constrain an activation-null timescale boundary, and Holmes et al. (2022) constrain an intervention-response null at 24 h. Those papers do not share the same comparison family, time window, or safe bundle role. On this site, SV2A PET therefore has to be split internally before bundle composition is judged at all.
One remaining shortcut still survived inside the central bundle rule. A proxy bundle could still cite human clearance / immune evidence as if that label already fixed one reusable human support-state row. The primary literature does not support that compression. Fultz et al. (2019) constrain a macroscopic CSF-oscillation route, Kim, Huang, & Liu (2025) constrain a parenchyma-CSF water-exchange route, Lim et al. (2025) constrain a respiration-conditioned net-flow route, Yoo et al. (2025) constrain an exercise-conditioned contrast-influx / parasagittal meningeal-lymphatic route, Eide et al. (2023) constrain an intrathecal-tracer / CSF-to-blood clearance-capacity route, Hirschler et al. (2025) constrain a CSF-mobility route, and Dagum et al. (2026) constrain a model-based brain-to-plasma biomarker-efflux route. Those routes do not share one carrier class, crossed boundary, intervention regime, direct observable, or safe bundle role. On this site, clearance / immune support therefore has to be split internally before bundle composition is judged at all.
Another shortcut still survived in this page's human-side reading rule. A proxy bundle could still cite human excitability evidence as if that already fixed one reusable human row. The primary literature does not support that compression. Tallman et al. (2025) constrain a human clinical single-unit allocation-related route in epilepsy patients, but explicitly treat firing only as an indirect index of excitability. Huber et al. (2013), Kuhn et al. (2016), and Fehér et al. (2026) constrain a sleep-history / plasticity-recalibration route through TMS-EEG or PAS-conditioned human cortex. Zrenner et al. (2018) and Khatri et al. (2025) constrain state-gated perturbation routes whose direct observables remain bounded plasticity-assay or corticospinal responses. Those routes do not share one spatial unit, one direct observable, one intervention regime, or one safe controller claim. On this site, human excitability evidence therefore has to be split internally before bundle composition is judged at all.
The remaining weakness after the recent composition updates was operational. A bundle can now be typed correctly by proxy class, operational maturity, and calibrator role, yet still fail as evidence if one row is unstable, method-family-specific, centre-bound, or only available in a narrow complete-case subset. The primary literature now makes that narrower reading necessary. Finnema et al. (2018) showed that even a comparatively stable SV2A PET route still needs route-specific kinetic modeling and reached mean absolute test-retest reproducibility of 3-9% for regional VT. Li et al. (2025) reported highly similar repeated dynamic DMRSI measurements only when the same brain, same setup, and same acquisition parameters were kept fixed, while Bøgh et al. (2024) showed that 3 T DMI can be repeatable with within-subject whole-brain CoVs around 10% at the best time point, but still under a coarse and long-acquisition protocol. Morgan et al. (2024) then showed that even when the named quantity is the same, DP-ASL and ME-ASL give materially different BBB water-exchange estimates and inconsistent age dependence. Amiri et al. (2023) showed in acute DoC that only 63 of 87 patients had both EEG and fMRI and that direct same-sample bundle comparisons used a 48-patient complete-feature subset, while Manasova et al. (2026) validated multimodal models across centres with different acquisition parameters, reported gains with more modalities, and found higher inter-modality disagreement in minimally conscious or improving patients. Therefore, on this site, human-proxy composition now includes route-local repeatability, method-family non-equivalence, cross-centre transfer, and partial-availability slices.
Another shortcut still had to be blocked. A shared quantity label does not guarantee that two routes are already interchangeable. Morgan et al. (2024) measured BBB water exchange in the same cohort with DP-ASL and ME-ASL and obtained significantly different Kw values with even a negative cross-participant correlation. Bøgh et al. (2024) showed that a clinical 3 T DMI protocol can be repeatable at its own operating point, but that same result does not erase the protocol dependence of field strength, coil, timing, spatial resolution, or processing. Therefore, on this site, a Human Proxy Composition Card must disclose method-family non-equivalence whenever nominally similar rows depend on different acquisition, fitting, input-function, or reconstruction routes. Otherwise, the bundle can still collapse because "same named quantity" is being mistaken for "same validated row."
One more shortcut still had to be blocked. Saying only deuterium imaging still leaves too much hidden. Li et al. (2025) reported repeated dynamic DMRSI measurements only when the same brain, same setup, and same acquisition parameters were kept fixed. Karkouri et al. (2026) mixed 12 healthy volunteers with 5 glioblastoma patients and reported only two healthy post-glucose scans in the abstracted protocol. Ahmadian et al. (2025) showed that [6,6'-2H2]glucose dose materially changes downstream metabolite visibility, and Bøgh et al. (2024) showed that the best whole-brain repeatability in a clinical 3 T DMI workflow appeared at the named 120-min time point rather than as a route-free constant. Therefore, a Human Proxy Composition Card now has to disclose whether the deuterium row is absolute quantification or kinetic-rate imaging, plus the dose, timing window, field strength / coil route, and whether repeatability was shown only within the same operating point or across protocols.
One more shortcut still had to be blocked. A paper can look richly multimodal while different rows were actually measured in very different subsets. Amiri et al. (2023) already showed that a direct EEG+fMRI comparison in acute DoC shrank to 48 complete-feature patients out of 87 enrolled. Manasova et al. (2026) then made the geometry even clearer in a larger multicentre study: in the main French dataset, EEG-LG was available in 290 patients, dMRI in 151, aMRI in 101, FDG-PET in 53, and fMRI-RS in only 44, while pairwise disagreement rose in MCS and improved patients. Therefore, on this site, a bundle must disclose not only the complete-case count but also the row-overlap geometry and whether missingness itself tracks site, severity, tolerance, or protocol. Otherwise, the apparent bundle gain can still be driven by a changing patient subset rather than by cleaner same-subject state constraint.
One more shortcut still had to be blocked. A bundle can show a net gain while hiding that its rows disagree most in the biologically or clinically difficult cases. Rohaut et al. (2024) showed that multimodal assessment can reduce prognostic uncertainty overall, while also noting that multimodal approaches increase the odds of discrepancies across markers that can create choice paralysis or biased decisions. Manasova et al. (2026) then showed that pairwise disagreements across modalities were higher in MCS than in UWS patients and also higher in improved than in not improved patients. Therefore, on this site, a Human Proxy Composition Card must now disclose the agreement / disagreement topology across modalities and key subgroups, together with the resolution policy: whether discordant cases trigger abstention, human adjudication, subgroup-specific follow-up, or are simply absorbed into one summary score.
Three promotion gates before row diversity counts
The composition rule on this page can now be read as three gates. This is a compression of the longer route-card fields, not a weaker rule. A bundle does not rise because the rows look impressive in aggregate. It rises only if it survives robustness, effective-window / regime-aware common-driver and quantity-bridge separation, and increment over the strongest single row. A passing bundle must also disclose where disagreement concentrates and how discordant cases were handled.
| Promotion gate | What must be shown | Ceiling if the gate is missing |
|---|---|---|
| Gate 1: Robustness | Show row-level repeatability at the actual operating point, distinguish within-setup repeatability from cross-centre / cross-protocol transfer, disclose method-family non-equivalence when nominally similar rows use different acquisition or fitting routes, and disclose the actual complete-case slice plus row-overlap geometry. Li et al. (2025), Bøgh et al. (2024), Morgan et al. (2024), and Manasova et al. (2026) show why those layers must stay separate. | The bundle stays as setup-bound, centre-bound, or complete-case-bound evidence rather than portable human observability. |
| Gate 2: Common-driver and quantity-bridge separation | Show that the rows are not only synchronized, but also temporally and physiologically compatible enough to be read on an explicitly named biological axis after shared-driver audit. A bundle that mixes a static similarity scaffold, a high-resolution metabolite-distribution route, a scan-window average, a task-evoked local-dynamics route, a minutes-long kinetic map, and an overnight perturbation does not define one state object by default. Lucchetti et al. (2025), Guo et al. (2025), Guo et al. (2024), Kaiser et al. (2026), Li et al. (2025), and Dagum et al. (2026) already span different temporal objects. Vafaii et al. (2024) found common and divergent cross-modal organization, Chen et al. (2025) found tightly coupled global progression plus distinct network patterns in simultaneous EEG-PET-MRI, Bolt et al. (2025) linked a major global fMRI mode to autonomic physiology, and Epp et al. (2025) showed that task BOLD and oxygen-metabolism changes can even move in opposite directions. | Cross-row agreement remains shared-factor evidence or proxy-rich correlation, not one validated state variable. |
| Gate 3: Increment over the strongest single row | Show what the bundle adds beyond the best individual row under matched cohort, condition, and held-out evaluation when available. This can be narrower latent-state ceiling, better calibration, or better prediction, but it must be shown rather than assumed. In the same step, report where pairwise or higher-order disagreements remain, especially in clinically or biologically important subgroups, and state the resolution policy for those cases. Rohaut et al. (2024) is informative because overall multimodal gain can still coexist with marker discrepancies, and Manasova et al. (2026) is informative precisely because it reports both performance gains with more modalities and higher inter-modality disagreement in clinically important groups. | Row diversity is treated as richer description, not as automatic state closure or same-subject state identification. |
One remaining shortcut is to treat same-session as if it solved temporal compatibility. The primary literature does not support that compression. Lucchetti et al. (2025) is a static parcel-similarity scaffold, Guo et al. (2025) is a high-resolution metabolite-distribution route, Guo et al. (2024) is a whole-brain NAD-content map, Kaiser et al. (2026) is a task-evoked localized NAD-dynamics route, Li et al. (2025) is a minutes-long kinetic mapping route, and Dagum et al. (2026) is an overnight perturbation-and-efflux route. Even before any fusion model is discussed, those rows do not all answer the question what is true right now in one matched state sample? On this site, the card therefore now asks for effective time window / state-axis compatibility and physiological / perturbation regime compatibility in addition to same-session or same-subject wording.
One remaining shortcut is to treat any reproducibility statement as if it solved deployment maturity. Li et al. (2025) explicitly reported repeated dynamic DMRSI measurements on the same brain with the same setup and acquisition parameters, and the result is important because it shows that a highly specialized route can be stable at its own operating point. But that is still different from showing the same rate maps across centres, scanners, RF hardware, blood-input workflows, or reconstruction pipelines. On this site, repeatability therefore has to be split into route-local repeatability and cross-site transfer before a proxy bundle is allowed to rise.
The composition problem in one table
| Human route | Direct observable | Spatial / temporal unit | Main model or acquisition burden | Safe ceiling on this site |
|---|---|---|---|---|
| Destructive local ultrastructure Lu et al. (2023); Shapson-Coe et al. (2024) |
Local ex vivo structural scaffold with nanoscale cell, axon, glia, and synapse geometry. | Cubic-millimeter surgical fragment; destructive one-time snapshot. | Preservation route, live-to-fix delay, section loss, segmentation, registration, proofreading. | Destructive local structural scaffold, not living whole-brain state readout. |
| SV2A PET quantification route Naganawa et al. (2021) |
Tracer-defined regional SV2A binding under an explicitly named quantification route. | Dynamic PET scan-window average in healthy humans; route-setting study rather than an atlas, disease contrast, or task-state result. | Tracer choice, arterial-versus-reference quantification, compartment model, scan window, and partial-volume handling. | Named synaptic-density quantification route, not a generic synaptic-state meter. |
| SV2A PET healthy atlas route Johansen et al. (2024) |
Regional SV2A distribution aggregated into a healthy-human atlas calibrated to autoradiography. | 33 healthy adults; cohort-level regional baseline rather than same-subject state tracking. | Atlas aggregation, PET-to-postmortem calibration logic, regional quantification route, and anatomy handling. | Healthy-human regional baseline / atlas route, not moment-to-moment synaptic efficacy or a universal bundle anchor. |
| SV2A PET disease / risk-contrast route Snellman et al. (2024) |
Group contrast in hippocampal [11C]UCB-J SUVR across APOE-risk strata in cognitively unimpaired older adults. | 46 participants across APOE ε4/ε4, ε3/ε4, and ε3/ε3 groups; risk-stratified contrast rather than task or intervention tracking. | Reference-region SUVR route, group comparison design, hippocampal ROI definition, and risk-cohort interpretation. | Bounded disease / risk-contrast route, not a direct cognition meter or a same-subject current-state readout. |
| SV2A PET task / activation / intervention route family Shatalina et al. (2024); Smart et al. (2021); Holmes et al. (2022) |
Task-linked association between regional SV2A binding and cognition or neural activity, plus bounded null results under brief activation or 24 h intervention designs. | 25 healthy adults for PET-fMRI task / cognition coupling, brief visual activation for K1-versus-binding dissociation, and a 24 h ketamine-response design rather than chronic state tracking. | Task design, time window, regional hypothesis choice, tracer-binding versus influx separation, and intervention timing. | Design- and timescale-conditioned association / null route family, not a universal activity meter or rapid-plasticity readout. |
| Whole-brain 1H-MRSI metabolic similarity Lucchetti et al. (2025) |
Within-subject pairwise correlations among five metabolite profiles (tCr, tNAA, Glx, Ins, Cho) across gray-matter parcels. | Parcel-level static similarity matrix in 51 adolescents aged 13-15 years, with independent replication in 13 healthy controls aged 15-35 years. | Spectral QC, partial-volume correction, parceling choice, z-scoring, similarity definition. | Macro biochemical similarity scaffold, not tractography, flux imaging, or controller-level state readout. |
| High-resolution 1H-MRSI metabolite-distribution route Guo et al. (2025) |
High-resolution ultrahigh-field metabolite-distribution maps reconstructed with extended spatiospectral encoding and subspace modeling. | Specialized ultrahigh-field route with explicit ghosting / aliasing / low-SNR handling rather than a parcel-similarity design. | Encoding / reconstruction route, artifact-control policy, and metabolite-map formation burden. | High-resolution metabolite-distribution proxy, not tractography, flux imaging, or controller-level state readout. |
| Human 31P-MRS metabolite / pH balance route Ren et al. (2015) |
ATP synthesis, phosphorus metabolite concentrations, and intra-/extracellular pH balance from 31P spectra. | Resting-brain spectroscopy in 12 healthy participants; scan-window biochemical balance rather than dynamic local kinetics. | 31P hardware, spectral quantification, and exchange-model assumptions such as EBIT versus saturation-transfer framing. | Macro metabolite / pH balance proxy, not branch-local mitochondrial positioning or synapse-specific ATP reserve. |
| Deuterium metabolite-mapping / absolute-quantification route Karkouri et al. (2026) |
Deuterated HDO / Glc / Glx / Lac metabolite distributions under an explicit absolute-quantification pipeline. | 12 healthy volunteers and 5 glioblastoma patients in a specialized 7 T workflow; only two healthy volunteers were scanned post-glucose in the abstracted protocol. | 7 T hardware, dedicated deuterium / proton coils, oral tracer timing, dose choice, B1-calibrated absolute quantification, and spectral fitting. | Specialized mixed-cohort macro deuterium metabolite-mapping / absolute-quantification route, not kinetic-rate imaging, branch-local mitochondrial positioning, or ATP nano-organization ground truth. |
| Deuterium kinetic-rate imaging Li et al. (2025) |
Glucose-transport and metabolic-rate maps under an explicit kinetic model. | 0.7 cc nominal voxels and 2.5 min/image whole-brain 7 T acquisition in five healthy participants, with repeat scans reported only at the same operating point. | 7 T hardware, dedicated deuterium / proton coils, oral tracer timing, blood-input function, kinetic model choice, and same-setup repeatability burden. | Specialized model-conditioned deuterium kinetic-rate route, not branch-local mitochondrial or ATP nano-organization ground truth and not a protocol-free portability claim. |
| Human myelin-water / calibrated T1w/T2w comparison route Arshad et al. (2017) |
ROI-level myelin-water fraction versus calibrated T1w/T2w ratio under repeat-scan reliability and concurrent-validity testing. | Healthy-adult white-matter ROI comparisons with back-to-back and repositioned scans; 20 adults for multi-echo T2 and 17 adults for calibrated T1w/T2w after exclusions. | Multi-echo T2 acquisition, image calibration, white-matter ROI sampling, and repeat-scan design. | Route-local comparison of MWF versus calibrated T1w/T2w, not a generic myelin truth and not an interchangeable bundle row for all human myelin claims. |
| Human relaxometry / MTsat myelin-comparison route Hagiwara et al. (2018) |
SyMRI-derived relaxometry myelin maps, MTsat macromolecular contrast, and T1w/T2w ratio compared within the same healthy-adult brains. | 20 healthy adults with whole-brain white- and gray-matter comparison rather than remyelination or bilayer-specific imaging. | Simultaneous relaxometry, MTsat acquisition, map calibration, and cross-method ROI analysis. | Cross-method white-matter myelin-comparison route, not the same object as MWF reference-style comparison, bilayer mapping, or remyelination-sensitive pathology readout. |
| Human myelin bilayer mapping route Baadsvik et al. (2024) |
Bilayer-sensitive macro myelin mapping in living human brain. | 1.4 mm mapping in two healthy volunteers under a proof-of-principle protocol. | Specialized high-performance hardware, non-routine acquisition, and route-specific bilayer sensitivity. | Specialized bilayer-sensitive macro myelin route, not a routine whole-brain myelin meter and not per-axon timing-state ground truth. |
| qT1 remyelination-sensitive pathology route Galbusera et al. (2025) |
qT1 sensitivity to cortical demyelination versus remyelination in postmortem multiple-sclerosis cortex. | Six whole postmortem human brains with cortical lesion classes rather than living-human same-subject monitoring. | Postmortem MRI-histology alignment, pathology-specific lesion typing, and route-specific sensitivity to remyelination rather than general myelin abundance. | Pathology-conditioned remyelination-sensitive route, not a living-human routine row and not interchangeable with MWF, MT-family, or bilayer-sensitive routes. |
| T1w/FLAIR tissue-health-sensitive ratio route Colaes et al. (2026) |
T1w/FLAIR ratio tested against MWF, diffusion metrics, and cognition with only weak MWF coupling. | 36 participants in a retrospective clinical-imaging cohort rather than a dedicated myelin-validation or longitudinal remyelination design. | Ratio construction from routine T1w and 2D FLAIR, lesion-region analysis, and route-specific dependence on broader tissue properties. | Tissue-health-sensitive ratio route, not a myelin-specific meter and not interchangeable with MWF, MT-family, bilayer-sensitive, or qT1 remyelination-sensitive routes. |
| Human clinical single-unit allocation route Tallman et al. (2025) |
Relative firing increase at encoding in hippocampal single units associated with sparse episodic-memory coding at retrieval in epilepsy patients. | Local clinical-unit route in implanted human hippocampus rather than a noninvasive whole-brain observability row. | Patient-specific invasive single-unit recording, memory-task design, and the explicit ceiling that firing is only an indirect index of excitability. | Local clinical-unit allocation-related route, not a whole-brain excitability meter and not direct readout of AIS / channel-state or recovery-controller identity. |
| Human sleep-homeostasis / plasticity proxy Huber et al. (2013); Kuhn et al. (2016); Fehér et al. (2026) |
TMS-EEG excitability and PAS-related plasticity outcomes that depend on wake history, sleep deprivation, or nap-mediated recalibration. | Intervention-backed noninvasive human route, but still assay-conditioned and controller-indirect. | TMS / PAS assay choice, sleep-history manipulation, and route-specific physiological-regime burden. | Perturbation-conditioned maintenance proxy, not direct identification of the responsible excitability or recovery controller. |
| Human state-gated perturbation proxy Zrenner et al. (2018); Khatri et al. (2025) |
EEG-defined or personalized whole-brain state-conditioned TMS outcomes, including plasticity efficacy and corticospinal response differences. | Operationally real closed-loop human route, but still mechanism-indirect and perturbation-conditioned. | State estimator quality, stimulation timing policy, corticospinal / plasticity assay choice, and route-specific closed-loop burden. | State-gated perturbation proxy, not direct measurement of AIS geometry, channel distribution, allocation bias, or long-horizon recovery control. |
| Human BBB water-exchange MRI Padrela et al. (2025); Morgan et al. (2024) |
Tex or Kw estimates for blood-to-tissue water exchange across the BBB from ASL-based MRI. | 194 healthy adults after QC in a lifespan cohort and 30 participants in a same-cohort DP-ASL versus ME-ASL comparison. | ASL sequence choice, fitting route, ATT / CBF coupling, and method-dependent quantification. | Macro BBB water-exchange proxy, not one generic permeability meter and not endothelial / pericyte controller ground truth. |
| Tracer-specific BBB PET transport Chung et al. (2025) |
Tracer-specific BBB permeability-surface-area product under a kinetic transport model. | Total-body dynamic PET scan-window inference across three radiotracers and selected human cohorts. | High-temporal-resolution PET, tracer-specific transport assumptions, arterial or image-derived input handling, and kinetic modeling. | Tracer-specific BBB transport proxy, not a generic BBB leakiness scalar and not direct molecular-controller readout. |
| Human choroid-plexus perfusion route Zhao et al. (2020); Sun et al. (2024) |
Apparent choroid-plexus blood flow / perfusion under ASL-based human MRI. | Early healthy-volunteer route extended by a large HCP-Aging cohort analysis, but still perfusion-side rather than transport-side human evidence. | pCASL acquisition, choroid-plexus segmentation, long-T1 interpretation, and aging-cohort transfer. | Choroid-plexus perfusion proxy, not blood-to-CSF transport truth, not epithelial-transporter identity, and not generic BBB evidence. |
| Human blood-to-CSF water-transport route Petitclerc et al. (2021) |
Ultra-long-TE ASL blood-to-CSF water transport / exchange time. | Specialized healthy-subject transport route rather than a routine perfusion or solute-clearance assay. | Ultra-long-TE ASL settings, compartment assumptions, and blood-to-CSF signal interpretation. | Blood-to-CSF transport proxy, not choroid-plexus perfusion, not DCE water cycling, and not route-free clearance truth. |
| Human choroid-plexus DCE water-cycling route Anderson et al. (2022) |
DCE-derived choroid-plexus water-efflux rate constant with separate contrast-agent leakage terms. | Older-adult cohort route with cognition-linked interpretation, but still a model-based DCE transport proxy. | DCE-MRI acquisition, modified two-site exchange modeling, and separation of kco from Ktrans. | Choroid-plexus water-cycling proxy, not perfusion truth, not blood-to-CSF transport equivalence, and not whole-brain clearance capacity. |
| Human apparent BCSFB-exchange route Wu et al. (2026) |
REXI-derived apparent kBCSFB in the human choroid plexus. | Small healthy-volunteer proof-of-principle route with scan-rescan repeatability, still below broad deployment or controller identity. | REXI sequence design, apparent-exchange interpretation, and route-specific repeatability limits. | Apparent BCSFB-exchange proxy, not route-free BCSFB function, not BBB water-exchange equivalence, and not epithelial-controller truth. |
| Human simultaneous BBB-versus-BCSFB ASL exchange route Petitclerc et al. (2026) |
Simultaneous ASL estimates of Kbl→GM and Kbl→CSF in one acquisition. | Model-comparison route that sharpens boundary separation rather than collapsing BBB and BCSFB into one transport scalar. | Multi-TE 3D-GRASE plus T2-preparation, two- versus three-compartment modeling, and boundary-specific parameter interpretation. | Boundary-separated BBB-versus-BCSFB exchange proxy, not one generic barrier-permeability row and not direct neurovascular / choroid-plexus controller measurement. |
| Human paired CSF-plasma protein-balance proteomics route Farinas et al. (2025) |
Individualized CSF/plasma ratios across 2,304 proteins from paired-fluid proteomics. | 2,171 healthy or cognitively impaired older individuals across multiple cohorts; a large-cohort paired-fluid route rather than transport imaging. | SomaScan paired-fluid assay, ratio construction from both compartments, cohort harmonization, and the explicit possibility that ratio shifts reflect synthesis or degradation as well as barrier transport. | Paired-fluid protein-balance barrier-system proxy, not a route-free BBB or BCSFB permeability scalar, not absolute concentration truth, and not cell-specific transporter identity. |
| SMBT-1 first-in-human MAO-B target-validation route Villemagne et al. (2022) |
Brain 18F-SMBT-1 binding under selegiline-blockade-supported MAO-B target validation in healthy humans. | 14 healthy volunteers; first-in-human dynamic brain PET route rather than disease-context contrast, brain-quantification generalization, or whole-body deployment burden. | Tracer synthesis, blockade logic, scan-window choice, quantification route, and healthy-volunteer regime. | MAO-B target-validation route, not a route-free astrocyte-state meter, not disease-context proof, and not I2BS equivalence. |
| SMBT-1 AD-spectrum MAO-B disease-context route Villemagne et al. (2022) |
Groupwise MAO-B-linked reactive-astrogliosis contrast across the Alzheimer disease continuum under one tracer family. | 77 volunteers spanning cognitively unimpaired Aβ-negative / Aβ-positive, MCI, and AD cohorts; pathology-context contrast rather than first-in-human validation or dosimetry. | AD-spectrum cohort composition, amyloid / tau context, partial-volume handling, quantification route, and disease-regime transfer limits. | AD-context MAO-B contrast route, not a generic human astrocyte baseline and not a route-free whole-brain astrocyte scalar. |
| SMBT-1 brain-quantification route Hiraoka et al. (2025) |
Kinetic and simplified quantitative behavior of 18F-SMBT-1 in human brain under explicitly compared modeling routes. | Human brain PET quantification study; route-setting / scan-window problem rather than target discovery, disease-general proof, or whole-body burden. | Reference-versus-kinetic model choice, scan-duration tradeoff, arterial-input burden, and transfer of the named quantification regime. | Brain-side MAO-B quantification route, not universal transfer across cohorts, not whole-body tracer-burden evidence, and not route-free astrocyte-state language. |
| SMBT-1 whole-body biodistribution route Mesfin et al. (2026) |
Organ distribution and excretion profile of 18F-SMBT-1 in healthy humans. | Six healthy volunteers under approximately 5.5 h whole-body dynamic PET; tracer-burden route rather than brain target validation or AD-context contrast. | Whole-body acquisition burden, hepatobiliary / intestinal distribution, and limited healthy-volunteer sample. | Whole-body biodistribution / tracer-burden route, not brain-side astrocyte quantification, not disease-context proof, and not a generic astrocyte meter. |
| I2BS astrocyte-related PET route Tyacke et al. (2018); Livingston et al. (2022) |
Human I2BS-related tracer binding with idazoxan-sensitive pharmacology and cognitively impaired cohort imaging. | 14 healthy male volunteers for first-in-human 11C-BU99008 pharmacology and 20-subject multimodal cognitive-impairment imaging; different target class from MAO-B routes. | Idazoxan competition, quantification-model choice, disease / amyloid context, and cross-target non-equivalence. | I2BS-related astrocyte proxy, not MAO-B equivalence, not a route-free astrocyte scalar, and not cell-specific controller identity. |
| Sleep-state CSF-oscillation route Fultz et al. (2019) |
Macroscopic CSF oscillations coupled to human NREM electrophysiology and hemodynamics. | Specialized sleep-state simultaneous physiology route rather than a solute-specific clearance assay. | Sleep staging, fast-fMRI / EEG coupling, macro fluid-motion interpretation, and sleep-state dependence. | Macroscopic sleep-state CSF-motion proxy, not molecular efflux or local immune-controller readout. |
| Human parenchyma-CSF water-exchange route Kim, Huang, & Liu (2025) |
Parenchyma-CSF water exchange measured in vivo with magnetization-transfer spin labeling. | Small-sample age-comparison MRI route rather than a protein-specific efflux assay. | MT spin-labeling design, age-comparison framing, water-exchange model, and region-level interpretation. | Parenchyma-CSF water-exchange proxy, not direct protein-clearance or local immune-controller truth. |
| Human respiration-conditioned CSF net-flow route Lim et al. (2025) |
Awake-state 2D phase-contrast MRI estimates of plane-specific CSF displacement and net flow under breathing manipulations. | Awake breathing-manipulation route with an explicitly plane-specific ceiling rather than whole-brain circulation truth. | Breathing task design, 2D PC-MRI plane choice, displacement-versus-net-flow interpretation, and explicit whole-brain-circulation caveat. | Respiration-conditioned CSF net-flow proxy, not route-free clearance flux or local controller identification. |
| Human exercise-conditioned contrast-influx route Yoo et al. (2025) |
Exercise-conditioned intravenous-contrast influx and parasagittal meningeal-lymphatic flow changes. | Healthy-young small-sample route with manual ROI burden and intervention-conditioned interpretation rather than baseline clearance truth. | Contrast administration, IR-ALADDIN / lymphatic-flow processing, manual ROI definition, and exercise-conditioned design. | Exercise-conditioned contrast-influx / parasagittal meningeal-lymphatic proxy, not route-free baseline glymphatic flux or local immune-controller ground truth. |
| Human intrathecal tracer / CSF-to-blood clearance route Eide et al. (2023) |
Intrathecal gadobutrol retention plus pharmacokinetic CSF-to-blood clearance variables linked to plasma biomarkers. | Neurological-disorder clinical route using an exogenous tracer, serial MRI, and population pharmacokinetic modeling rather than natural-sleep baseline physiology. | Intrathecal administration, PK model, clinical cohort composition, and plasma-biomarker linkage. | Intrathecal-tracer / CSF-to-blood-clearance-capacity proxy, not endogenous whole-brain clearance truth or local synaptic-maintenance readout. |
| CSF mobility MRI Hirschler et al. (2025) |
CSF mobility, explicitly distinguished from net flow or diffusion. | 0.45-mm isotropic 7 T MRI with whole-brain rest maps in 20 healthy younger individuals, plus driver analyses reported in 11 of 24 total healthy participants. | CSF-specific sequence design, mobility-encoding tensor model, 7 T acquisition, and region interpretation. | Macro support-state / mobility proxy, not direct clearance flux or local immune-controller identification. |
| Sleep-linked glymphatic efflux to plasma Dagum et al. (2026) |
Overnight plasma biomarker changes interpreted through a randomized crossover and a multicompartment brain-to-plasma model. | Randomized crossover trial with 39 healthy older participants aged 49-66 years; overnight sleep versus sleep deprivation. | Investigational device, plasma biomarker assays, multicompartment model, and overnight physiology assumptions. | Model-based human biomarker-efflux route, not local synaptic-maintenance or segment-specific lymphatic ground truth. |
Collapse errors to block
| Shortcut that fails | Why it fails in primary literature | What must be disclosed instead |
|---|---|---|
Quantity / target / transport collapsedensity + similarity + metabolite / pH balance + exchange flux + NAD map / dynamics + absolute metabolite map + kinetic rate + water exchange + transport + target-defined astrocyte binding + target-defined neuroimmune PET + mobility + efflux = same state variable |
Johansen measures regional SV2A density proxy, Lucchetti measures parcel-level similarity, Ren (2015) measures metabolite / pH balance, Ren (2017) measures exchange flux, Guo and Kaiser measure NAD content or localized dynamics, Li and Karkouri measure deuterated metabolite-mapping versus kinetic-rate routes, Morgan and Padrela measure BBB water exchange, Chung measures tracer-specific BBB transport, Zhao and Sun measure choroid-plexus perfusion, Petitclerc (2021) measures blood-to-CSF transport, Anderson measures DCE water cycling, Wu measures apparent BCSFB exchange, Petitclerc (2026) measures simultaneous BBB-versus-BCSFB exchange, Villemagne, Matsuoka, Best, and Tyacke measure tracer-family-separated target-defined astrocyte PET, Biechele and Wijesinghe measure TSPO disease-context / validation-bounded PET, Horti and Ogata measure CSF1R route-setting PET, Yan measures COX-2 enzyme-defined PET, Hirschler measures CSF mobility, and Dagum models overnight biomarker efflux. Those are different inferential objects. | Name the exact object: density, similarity, metabolite / pH balance, exchange flux, NAD content or dynamics, absolute metabolite map, kinetic rate, BBB water exchange, tracer-specific BBB transport, choroid-plexus perfusion, blood-to-CSF transport, water cycling, apparent BCSFB exchange, simultaneous BBB-versus-BCSFB exchange, target-defined astrocyte binding, TSPO disease-context / validation-bounded PET, CSF1R route-setting PET, COX-2 enzyme-defined PET, mobility, or model-based efflux. |
Family-internal route collapseone family name = one reusable bundle row |
Naganawa et al. (2021) fix an SV2A quantification route, Johansen et al. (2024) fix a healthy atlas / baseline route, Snellman et al. (2024) fix a disease / risk-contrast route, Shatalina et al. (2024) fix a task / cognition association route, Smart et al. (2021) fix an activation-null timescale boundary, and Holmes et al. (2022) fix an intervention-response null at 24 h. The same problem appears inside astrocyte PET: Villemagne et al. (2022) fix a first-in-human SMBT-1 MAO-B target-validation route, Villemagne et al. (2022) fix an SMBT-1 AD-spectrum disease-context route, Hiraoka et al. (2025) fix an SMBT-1 brain-quantification route, Mesfin et al. (2026) fix an SMBT-1 whole-body biodistribution route, Matsuoka et al. (2026) fix a separate SL25.1188 simplified-quantification route, Best et al. (2026) fix a severity- and smoking-conditioned SL25.1188 disease route, and Tyacke et al. (2018) plus Livingston et al. (2022) fix an I2BS route. The same problem appears inside human neuroimmune PET: Biechele et al. (2023) and Wijesinghe et al. (2025) fix a TSPO disease-context / validation-bounded route, Horti et al. (2022) and Ogata et al. (2025) fix a CSF1R route-setting route, and Yan et al. (2025) fix a COX-2 enzyme-defined route. The same problem appears inside human myelin MRI: Arshad et al. (2017) fix an MWF / calibrated T1w:T2w comparison route, Hagiwara et al. (2018) fix a relaxometry / MTsat comparison route, Baadsvik et al. (2024) fix a bilayer-sensitive mapping route, Galbusera et al. (2025) fix a qT1 remyelination-sensitive pathology route, and Colaes et al. (2026) fix a T1w/FLAIR tissue-health-sensitive ratio route. The same problem appears inside human blood-CSF barrier / choroid-plexus routes: Zhao et al. (2020) and Sun et al. (2024) fix a perfusion route, Petitclerc et al. (2021) fix a blood-to-CSF transport route, Anderson et al. (2022) fix a DCE water-cycling route, Wu et al. (2026) fix an apparent BCSFB-exchange route, and Petitclerc et al. (2026) fix a simultaneous BBB-versus-BCSFB exchange route. The same problem also appears inside human clearance-transport routes: Fultz et al. (2019) fix a macroscopic CSF-oscillation route, Kim, Huang, & Liu (2025) fix a parenchyma-CSF water-exchange route, Lim et al. (2025) fix a respiration-conditioned net-flow route, Yoo et al. (2025) fix an exercise-conditioned contrast-influx route, Eide et al. (2023) fix an intrathecal-tracer / CSF-to-blood clearance-capacity route, Hirschler et al. (2025) fix a CSF-mobility route, and Dagum et al. (2026) fix a model-based biomarker-efflux route. Those are not interchangeable bundle rows either. | Name the family-internal comparison type explicitly: for example, quantification route, healthy atlas / baseline, disease / risk contrast, task / cognition association, activation-null, or intervention-response for SV2A; SMBT-1 target validation, SMBT-1 AD-spectrum contrast, SMBT-1 brain quantification, SMBT-1 whole-body biodistribution, SL25.1188 simplified-quantification / disease / severity, or I2BS target-defined imaging for astrocyte PET; TSPO disease-context / validation-bounded, CSF1R route-setting, or COX-2 enzyme-defined for human neuroimmune PET; MWF / calibrated T1w:T2w comparison, relaxometry / MTsat comparison, bilayer-sensitive mapping, or qT1 remyelination-sensitive pathology for human myelin MRI; choroid-plexus perfusion, blood-to-CSF transport, DCE water cycling, apparent BCSFB exchange, or simultaneous BBB-versus-BCSFB exchange for blood-CSF-barrier evidence; and macroscopic CSF oscillation, parenchyma-CSF water exchange, respiration-conditioned net-flow, exercise-conditioned contrast influx, intrathecal tracer / CSF-to-blood clearance, CSF mobility, or model-based biomarker efflux for human clearance routes, then state the time window and design burden that go with that row. |
Human excitability-route collapseclinical single-unit allocation + sleep-history plasticity recalibration + state-gated perturbation = one human excitability row |
Tallman et al. (2025) constrain local hippocampal single-unit allocation-related firing in epilepsy patients and explicitly stop short of treating firing as direct excitability truth. Huber et al. (2013), Kuhn et al. (2016), and Fehér et al. (2026) constrain sleep-history / plasticity-recalibration routes through TMS-EEG or PAS outcomes. Zrenner et al. (2018) and Khatri et al. (2025) constrain EEG-defined or personalized whole-brain state-gated perturbation routes whose direct observables remain bounded plasticity-assay or corticospinal responses. Those papers do not share one spatial unit, one direct observable, one perturbation regime, or one safe controller claim. | Name the human evidence class explicitly: clinical single-unit allocation route, sleep-homeostasis / plasticity proxy, or state-gated perturbation proxy, then state the direct observable, intervention burden, and what controller still remains latent. |
Spatial-unit collapselocal fragment + regional atlas + parcel graph + macro voxel = same resolution ladder |
Shapson-Coe is a local surgical fragment, Johansen is regional atlas-level PET, Lucchetti is gray-matter parcel similarity, Ren is large-volume biochemical spectroscopy, Li is 0.7 cc voxelwise kinetic imaging, Baadsvik is 1.4 mm macro myelin mapping, and Dagum is overnight plasma-level efflux inference. These do not resolve the same biological unit. | Name the biological unit actually constrained: fragment, region, parcel, large spectroscopy voxel, macro voxel, or whole-body efflux trend. |
Timescale collapsestatic scaffold + scan average + resting balance + minutes-long kinetics + overnight physiology = one current state |
Naganawa and Johansen operate over PET scan windows, Lucchetti is a static similarity scaffold, Ren (2015) is a resting metabolite / pH balance route, Ren (2017) is a model-conditioned exchange-flux route, Li resolves minutes-long deuterium dynamics, and Dagum is an overnight sleep manipulation. They do not all answer what is true right now at the same timescale. | Name the time window explicitly: structural snapshot, scan-window average, resting biochemical balance, minutes-long kinetics, or overnight state transition. |
Model-burden collapsemultimodal means directly seen |
PET requires tracer and kinetic interpretation, 31P-MRS requires spectral and exchange-model interpretation, deuterium imaging requires quantification and kinetic modeling, ASL-derived BBB metrics require fitting-route choices, CSF mobility MRI requires mobility-tensor interpretation, and Dagum requires a multicompartment model. Model burden changes what is justified. | Disclose tracer, model family, correction route, input function, synchronization route, and abstention boundary. |
Method-family collapsesame quantity label = same validated row |
Arshad et al. (2017) showed that calibrated T1w/T2w and MWF can both look usable while still failing to become one criterion-valid myelin row, and Hagiwara et al. (2018) showed stronger white-matter agreement between SyMRI and MTsat than with T1w/T2w. Morgan et al. (2024) then showed that DP-ASL and ME-ASL can return materially different BBB water-exchange estimates with inconsistent age dependence inside one cohort, while Bøgh et al. (2024) showed that a 3 T DMI repeatability result is an operating-point property of a stated protocol rather than a generic guarantee for all deuterium imaging routes. A quantity label alone therefore does not fix row equivalence. | Name the exact acquisition family, fitting route, correction path, and whether cross-method agreement was actually shown rather than assumed. |
Deployment-maturity collapseif a route exists, it is already field-ready |
Li used custom 7 T hardware in five participants; Karkouri used 7 T with a dedicated 2H/1H array; Baadsvik used two healthy volunteers with high-performance hardware; Hirschler used ultra-high-field MRI with a specialized sequence; Dagum used an investigational device. These are advances, but not routine whole-brain deployment. | Name cohort size, hardware class, scan burden, device status, and whether the route is routine, specialized, or proof-of-principle. |
Cohort-regime collapseadolescent cohort + healthy younger adults + healthy older adults + pathology-specific remyelination + AD continuum = same human baseline |
Lucchetti et al. (2025) derive their main metabolic-similarity route from adolescents aged 13-15 years, Padrela et al. (2025) establish BBB water-exchange reference ranges across adulthood, Hirschler et al. (2025) map CSF mobility in healthy younger adults and separately examine CAA, Dagum et al. (2026) test healthy older adults aged 49-66 years, Villemagne et al. (2022) study healthy volunteers for first-in-human SMBT-1 target validation, Villemagne et al. (2022) image reactive astrogliosis across the Alzheimer disease continuum, and Galbusera et al. (2025) study postmortem MS cortex. These rows are informative, but they do not share one developmental, aging, or pathology regime by default. | Disclose age band, health/pathology regime, enrichment or exclusion criteria, and why transfer across cohorts is justified for the claimed latent variable. |
Causal-bridge collapselocal rodent controller causality + human proxy observability = measured human controller |
Williamson et al. (2025), Dewa et al. (2025), and Bukalo et al. (2026) show local rodent astrocyte-causal control of recall, multiday stabilization, and fear-memory representations, whereas Villemagne et al. (2022), Villemagne et al. (2022), and Tyacke et al. (2018) constrain target-defined human astrocyte-related PET, Hirschler et al. (2025) constrains CSF mobility, and Dagum et al. (2026) constrains model-based overnight biomarker efflux. These do not share the same species, spatial unit, direct observable, or controller identity. | Name the state family made causally relevant, the human direct observable, the species / unit gap, the bridge assumption, and the local controller that remains latent. |
Common-driver collapseif rows move together, they must be the same latent variable |
Vafaii et al. (2024) found both common and divergent cross-modal structure, Chen et al. (2025) found coupled global progression plus two distinct network patterns in simultaneous EEG-PET-MRI, and Bolt et al. (2025) showed that a major global fMRI mode is substantially coupled to autonomic physiology. Agreement can therefore reflect shared nuisance or shared global state rather than one solved target variable. | Disclose shared-vs-specific decomposition, nuisance calibration, matched perturbation, and what still remains vulnerable to autonomic / vascular / motion-linked common drivers. |
Discordance-collapseif the overall bundle score improved, disagreement no longer matters |
Rohaut et al. (2024) showed that multimodal assessment can reduce uncertainty while also warning that discrepancies across markers can create choice paralysis or biased decisions. Manasova et al. (2026) then showed that pairwise disagreements were higher in MCS and improved patients, meaning discordance can concentrate exactly where interpretation is hardest. | Disclose the pairwise / subgroup-conditioned disagreement topology and the resolution policy: abstain, adjudicate, collect extra rows, or explicitly accept that the final score silently absorbed the conflict. |
Increment-collapsemore rows listed = more state closure |
Current primary papers advance different rows, but they do not by themselves show that a bundle of similarity, energetic balance, rate imaging, myelin-family, BBB, astrocyte, and clearance routes outperforms the strongest row under matched subject, session, and condition. Diversity of rows is not yet evidence of incremental state closure. | Report what the bundle adds beyond the strongest single row under the same reading rule, ideally with matched-condition or held-out evaluation, and say explicitly when the gain is only rhetorical. |
The safe reading of SV2A PET is narrower than “synapses were measured, therefore this bundle now contains one human synaptic-state row.” Naganawa et al. (2021) showed that quantification depends on tracer kinetics, model choice, and scan window. Johansen et al. (2024) built a healthy-human regional atlas. Snellman et al. (2024) then showed lower hippocampal [11C]UCB-J signal in cognitively unimpaired APOE ε4/ε4 carriers than in APOE ε3/ε3, while also reporting no significant relation to cognitive scores in that risk cohort. Shatalina et al. (2024) further found a positive association with task-switching activation and switch cost in healthy adults, yet not with the N-back task in the same study. Smart et al. (2021) showed that visual stimulation increased tracer influx K1 but left binding measures unchanged, and Holmes et al. (2022) found no detectable overall SV2A change 24 h after ketamine despite antidepressant response. On this site, that means SV2A PET has to be split by comparison family before bundle promotion is judged at all: quantification, healthy atlas, risk contrast, task / cognition association, activation-null, and intervention-response do not share one automatic bundle role.
The phrase human spectroscopy is especially easy to overread. Lucchetti et al. (2025) define a within-subject parcel-similarity graph from five metabolites and show that it aligns only weakly with tractography-based structural connectivity. Guo et al. (2025) define a high-resolution 1H-MRSI metabolite-distribution route whose output depends on explicit reconstruction and artifact-control choices. Ren et al. (2015) constrain ATP synthesis, phosphorus metabolites, and pH balance in resting human brain. Karkouri et al. (2026) constrain deuterated metabolite distributions under an absolute-quantification pipeline, while Li et al. (2025) constrain glucose-transport and kinetic-rate maps at 7 T under a blood-input model. On this site, similarity, high-resolution metabolite distribution, energetic balance, deuterium metabolite mapping, and deuterium kinetic-rate imaging therefore remain separate rows and do not inherit one another's claim ceiling.
Within-family compression still breaks composition here as well. Arshad et al. (2017) showed that calibrated T1w/T2w can remain reasonably reliable while still giving only limited concurrent validity against MWF. Hagiwara et al. (2018) then showed that SyMRI and MTsat align more strongly in white matter than either does with T1w/T2w. Baadsvik et al. (2024) added a bilayer-sensitive route in only two healthy volunteers, Galbusera et al. (2025) showed that qT1, but not MWF or MTR, separated demyelinated from remyelinated cortical lesions in postmortem multiple-sclerosis cortex, and Colaes et al. (2026) showed that T1w/FLAIR keeps only weak associations with MWF and is safer to read as a broader tissue-health marker. On this site, that means myelin MRI has to be split by comparison family before bundle promotion is judged at all: MWF versus calibrated T1w/T2w comparison, relaxometry / MTsat comparison, bilayer-sensitive mapping, qT1 remyelination-sensitive pathology route, and T1w/FLAIR tissue-health-sensitive ratio route do not share one automatic bundle role.
Morgan et al. (2024) and Padrela et al. (2025) constrain BBB water exchange, whereas Chung et al. (2025) constrains tracer-specific transport-model permeability. Astrocyte PET now needs the same discipline. Villemagne et al. (2022) constrain a first-in-human SMBT-1 MAO-B target-validation route, Villemagne et al. (2022) constrain an AD-spectrum disease-context route, Hiraoka et al. (2025) constrain a brain-quantification route, Mesfin et al. (2026) constrain a whole-body biodistribution / tracer-burden route, Matsuoka et al. (2026) constrain a simplified SL25.1188 AD quantification route, Tyacke et al. (2018) constrain an I2BS target class, and Best et al. (2026) show that SL25.1188 MAO-B binding can still shift with cohort severity and smoking regime. On this site, those rows therefore stay transport-defined, target-defined, tracer-family-defined, and route-role-defined; they are not interchangeable family labels.
The safe reading of human clearance evidence is narrower than “a glymphatic route exists.” Fultz et al. (2019) constrain macroscopic CSF oscillation, Kim, Huang, & Liu (2025) constrain parenchyma-CSF water exchange, Lim et al. (2025) constrain respiration-conditioned net flow, Yoo et al. (2025) constrain exercise-conditioned contrast influx and parasagittal meningeal-lymphatic flow, Eide et al. (2023) constrain intrathecal-tracer / CSF-to-blood clearance capacity, Hirschler et al. (2025) constrain CSF mobility, and Dagum et al. (2026) constrain model-based overnight biomarker efflux. Those routes differ in carrier class, crossed boundary, intervention regime, time window, and model burden. On this site, clearance / immune support therefore has to be typed as macroscopic CSF oscillation, parenchyma-CSF water exchange, respiration-conditioned net-flow, exercise-conditioned contrast influx, intrathecal tracer / CSF-to-blood clearance, CSF mobility, or model-based biomarker efflux before bundle promotion is judged at all.
The safe reading of human neuroimmune PET is narrower than “immune PET now exists.” Biechele et al. (2023) and Wijesinghe et al. (2025) constrain a TSPO disease-context / validation-bounded route, Horti et al. (2022) and Ogata et al. (2025) constrain CSF1R route-setting PET, and Yan et al. (2025) constrains an enzyme-defined COX-2 route. Those routes differ in target class, disease context, pharmacological validation, and model burden. On this site, immune PET therefore has to be typed as TSPO disease-context / validation-bounded, CSF1R route-setting, or COX-2 enzyme-defined before bundle promotion is judged at all.
One remaining shortcut is to list human proxy rows together as if they came from one matched biological baseline. The primary papers themselves argue against that. Johansen et al. (2024) provide a healthy-adult SV2A atlas, Snellman et al. (2024) probe cognitively unimpaired but genetically high-risk older adults, Lucchetti et al. (2025) derive metabolic similarity mainly from adolescents aged 13-15 years with a separate 15-35-year replication cohort, Hirschler et al. (2025) map CSF mobility in healthy younger adults and then test a separate CAA cohort, Dagum et al. (2026) study healthy older adults aged 49-66 years under sleep-versus-deprivation crossover, Villemagne et al. (2022) studies healthy volunteers for first-in-human SMBT-1 target validation, and Villemagne et al. (2022) images reactive astrogliosis across the Alzheimer disease continuum in volunteers averaging 76 +/- 5.5 years. Therefore, on this site, proxy composition must also disclose developmental / aging regime, health versus pathology regime, and transfer claim before cross-row agreement is treated as more than cross-cohort analogy.
Proxy class, evidence role, operational maturity, and calibrator role are different questions
The remaining weakness after separating quantity type, spatial unit, and model burden is subtler. Readers can still think that once a route has a recognizable proxy class and looks technically real, the route must already calibrate a broad hidden-state family and already be usable for any bundle job. The current primary literature does not support that shortcut. On this site, each human route is therefore read along four separate axes: what kind of proxy it is, what evidence role it safely serves, how operationally mature it is, and what hidden-state family it safely calibrates.
Those four axes are defined per route. When several routes are combined, this site adds a separate cohort / regime compatibility audit so that developmental stage, aging band, and pathology enrichment are not silently treated as interchangeable background.
| Evidence role on this site | What it can support | What it cannot silently become |
|---|---|---|
| Normative atlas / cohort prior Johansen et al. (2024) |
A healthy or reference distribution that anchors where a quantity typically sits across a cohort. | A same-subject change tracker, perturbation-response witness, or generic current-state meter. |
| Cross-sectional contrast Snellman et al. (2024) |
A bounded risk-, disease-, or group-difference statement under a named cohort design. | A same-subject baseline anchor or within-subject intervention tracker. |
| Same-subject baseline readout Finnema et al. (2018); Bøgh et al. (2024) |
A route-local baseline anchor whose repeatability is characterized at the stated operating point. | A validated change witness outside that route and window, or a portable cross-centre bundle anchor without separate transfer evidence. |
| Within-subject change witness Smart et al. (2021); Kaiser et al. (2026) |
A bounded same-subject change statement under a named task, activation, or localized dynamic design. | A generic current-state meter, chronic tracker, or broad maintenance-state readout. |
| Perturbation-response witness Holmes et al. (2022); Dagum et al. (2026) |
A bounded response or response-null statement under a named intervention, challenge, or sleep-manipulation regime. | A stable baseline atlas, a general state-complete meter, or a chronic same-subject tracker across unmatched regimes. |
| Human route | Proxy class on this site | Operational maturity on this site | Safe calibrator role | What still remains outside calibration |
|---|---|---|---|---|
| Human pupil-size arousal proxy in sleep Carro-Domínguez et al. (2025) |
Mixed arousal proxy | Non-invasive human route with synchronized PSG / ECG, but still a pooled arousal readout rather than a transmitter-specific measurement. | Constrains coarse arousal-level fluctuations and stimulation-response gating under controlled lighting and named sleep-stage context. | Specific transmitter identity, receptor occupancy, endogenous release, and current whole-brain neuromodulatory field. |
| Human receptor / transporter atlas prior Hansen et al. (2022); Nakuci & Bansal (2025) |
Normative chemoarchitectural prior | Large collated healthy-cohort PET atlas and follow-on modeling scaffold rather than a same-subject current-state readout. | Constrains regional receptor / transporter density priors and model-based hypotheses about broad system-level modulation. | Same-subject change, task-evoked transmitter release, individual current occupancy, and moment-to-moment neuromodulatory state. |
| Human occupancy PET Wong et al. (2013); Schlosser et al. (2025) |
Ligand- and dose-limited target-engagement proxy | Established PET design for administered-drug target engagement, but still receptor-family-, tracer-, dose-, and window-specific. | Constrains whether a named administered compound occupies a selected target under the stated tracer, dose, and scan design, including informative null occupancy. | Endogenous transmitter release, unsampled receptor families, current whole-brain neuromodulatory field, and route-free antidepressant mechanism. |
| Human challenge-linked displacement / release-sensitive PET Koepp et al. (1998); Erritzoe et al. (2020); Miederer et al. (2025) |
Challenge-linked endogenous release proxy | Specialized task- or drug-challenge PET route with explicit kinetic-model and time-window burden. | Constrains bounded dopamine or serotonin release proxy under the named task block or pharmacological challenge in the measured regions. | Baseline target engagement, unsampled transmitter systems, task-general neuromodulatory state, and chronic whole-brain transmitter identity. |
| SV2A PET family Johansen et al. (2024); Snellman et al. (2024); Shatalina et al. (2024); Smart et al. (2021); Holmes et al. (2022) |
Family-split regional synaptic-density proxy | Real human PET route, but different papers answer healthy-baseline, disease / risk-contrast, task-association, activation-null, or intervention-response questions rather than one generic state-meter question. | After family-internal typing, constrains bounded regional density contrasts or design-conditioned associations. | Current release probability, postsynaptic efficacy, rapid plasticity, branch-local controller state, and any claim that ignores the family-internal split. |
| Whole-brain 1H-MRSI metabolic similarity Lucchetti et al. (2025) |
Macro biochemical similarity scaffold | Cohort-level whole-brain mapping with replication, but still a static similarity route. | Constrains parcel-level biochemical organization and cross-region metabolic similarity structure. | Kinetic metabolic rate, energetic reserve, axonal wiring, and local maintenance-controller identity. |
| High-resolution 1H-MRSI metabolite-distribution route Guo et al. (2025) |
High-resolution metabolite-distribution proxy | Specialized ultrahigh-field mapping route with explicit reconstruction and artifact-control burden rather than parcel-similarity analysis. | Constrains metabolite-distribution structure under extended spatiospectral encoding and subspace modeling. | Parcel-level similarity, kinetic-rate maps, deuterium absolute quantification, and local maintenance-controller identity. |
| Human 31P-MRS metabolite-balance / pH route Ren et al. (2015) |
Macro 31P metabolite-balance / pH proxy | Established resting-brain spectroscopy route, but still a coarse balance readout rather than dynamic local kinetics. | Constrains ATP synthesis, phosphorus metabolite balance, and pH under a specific spectral / exchange-model route. | Exchange-flux, NAD dynamics, glucose transport rates, branch-local mitochondrial positioning, ATP nano-organization, and cell-specific energetic reserve. |
| Human 31P MT exchange-flux route Ren et al. (2017) |
Model-conditioned macro 31P exchange-flux proxy | Specialized 7 T route with band-inversion / magnetization-transfer modeling rather than a routine resting-balance readout. | Constrains PCr→γ-ATP creatine-kinase exchange, Pi→γ-ATP synthesis, and ATP intramolecular exchange under a 5-pool model. | Cell-specific energetic reserve, branch-local mitochondrial positioning, task-evoked redox dynamics, and any claim that treats one flux estimate as a generic energetic scalar. |
| Human 31P NAD-content mapping route Guo et al. (2024) |
Macro 31P NAD-content map proxy | Specialized 7 T whole-brain route with low-concentration fitting, subspace denoising, and long acquisition burden. | Constrains whole-brain intracellular NAD content under explicit spectral fitting. | Task-evoked local NAD dynamics, whole-brain moment-to-moment redox control, branch-local mitochondrial residence, and task-general controller identity. |
| Human 31P functional NAD-dynamics route Kaiser et al. (2026) |
Localized functional 31P NAD-dynamics proxy | Specialized 7 T task fMRS route with prior fMRI localization and one occipital voxel rather than a whole-brain map. | Constrains bounded task-evoked NAD+ dynamics under explicit spectral fitting and named task blocks. | Whole-brain NAD-content mapping, task-general energetic-controller identity, branch-local mitochondrial residence, and whole-brain moment-to-moment redox control. |
| Deuterium metabolite-mapping / absolute-quantification route Karkouri et al. (2026) |
Specialized macro deuterium metabolite-mapping / absolute-quantification route | High-burden 7 T route with dedicated hardware, calibrated quantification, and limited healthy / patient cohorts. | Constrains deuterated metabolite distributions under explicit absolute quantification. | Glucose-transport terms unless explicitly modeled, branch-local mitochondrial positioning, ATP nano-organization, and synapse-specific energetic reserve. |
| Deuterium kinetic-rate imaging Li et al. (2025) |
Specialized model-conditioned deuterium kinetic-rate route | High-burden 7 T route with dedicated hardware, blood-input acquisition, and very small healthy cohorts. | Constrains macro glucose-transport / metabolic-rate maps under explicit kinetic models. | Branch-local mitochondrial positioning, ATP nano-organization, and synapse-specific energetic reserve. |
| Human myelin-water / calibrated T1w/T2w comparison route Arshad et al. (2017) |
Quantity-defined macro myelin comparison route | Healthy-adult repeat-scan comparison route with ROI-level validity and reliability limits. | Constrains the relation between MWF and calibrated T1w/T2w in named white-matter ROIs under the stated acquisition route. | Relaxometry / MTsat comparison, bilayer-sensitive mapping, remyelination-sensitive pathology readout, and per-axon timing support. |
| Human relaxometry / MTsat myelin-comparison route Hagiwara et al. (2018) |
Quantity-defined white-matter myelin-comparison route | Healthy-adult cross-method comparison route rather than living-human remyelination tracking. | Constrains the relation among SyMRI, MTsat, and T1w/T2w under one acquisition and ROI-analysis policy. | MWF reference-style comparison, bilayer-sensitive mapping, remyelination-sensitive pathology readout, and per-axon timing support. |
| Human myelin bilayer mapping route Baadsvik et al. (2024) |
Bilayer-sensitive macro myelin proxy | Specialized proof-of-principle route in two healthy volunteers. | Constrains bilayer-sensitive macro myelin mapping under the named high-burden acquisition route. | Routine deployment, remyelination-sensitive pathology inference, per-axon timing support, and generic myelin-family equivalence. |
| qT1 remyelination-sensitive pathology route Galbusera et al. (2025) |
Pathology-conditioned remyelination-sensitive proxy | Postmortem multiple-sclerosis-cortex validation route rather than living-human same-subject imaging. | Constrains cortical demyelination-versus-remyelination sensitivity for qT1 under pathology-specific MRI-histology alignment. | Living-human routine myelin mapping, bilayer-sensitive mapping, MWF / T1w:T2w equivalence, and per-axon timing support. |
| T1w/FLAIR tissue-health-sensitive ratio route Colaes et al. (2026) |
Tissue-health-sensitive ratio proxy | Retrospective clinical-imaging route in 36 participants rather than dedicated myelin-validation or longitudinal remyelination tracking. | Constrains only weak T1w/FLAIR associations with MWF, diffusion metrics, and lesion-linked cognition under the stated ratio-construction route. | Myelin-specific readout, MWF / T1w:T2w equivalence, bilayer-sensitive mapping, remyelination-sensitive pathology inference, and per-axon timing support. |
| Human BBB water-exchange method-comparison route Morgan et al. (2024) |
Method-family non-equivalence anchor within BBB water-exchange MRI | Same-cohort method-comparison route rather than a healthy-reference or disease-burden row. | Constrains that DP-ASL and ME-ASL can return materially different BBB water-exchange values and inconsistent age dependence under named acquisition / fitting routes. | Method interchangeability, healthy-reference norms, disease-burden interpretation, generic BBB leakiness language, and pericyte / endothelial controller identity. |
| Human BBB water-exchange healthy-lifespan reference route Padrela et al. (2025) |
Healthy-adult macro BBB water-exchange reference proxy | Lifespan healthy-adult route with explicit age / sex interpretation and a ceiling conditioned by CBF / ATT correction. | Constrains Tex-style age-related variation under one multi-echo ASL route while showing that the apparent gray-matter age effect does not survive the named perfusion / transit-time correction. | Method equivalence across ASL families, disease-burden contrast, tracer-specific transport, generic BBB leakiness language, and controller identity. |
| Human BBB water-exchange disease-burden contrast route Padrela et al. (2026) |
Disease-burden BBB water-exchange contrast proxy | Cross-sectional SCD / MCI and WMH-burden route rather than a healthy-reference or method-comparison row. | Constrains lower Tex in early cognitive / cerebrovascular-burden groups under one ASL route, while amyloid-group differences do not survive age / sex adjustment. | Amyloid-specific barrier truth, healthy-reference baseline, method equivalence, tracer-specific transport, and pericyte / endothelial controller identity. |
| Tracer-specific BBB PET transport Chung et al. (2025) |
Tracer-specific BBB transport proxy | Advanced total-body PET route with high modeling burden and tracer-specific interpretability. | Constrains permeability-surface-area product for explicitly named molecular radiotracers under a transport model. | Generic BBB-permeability language, direct controller readout, and transport equivalence across tracers. |
| Human choroid-plexus perfusion route-setting study Zhao et al. (2020) |
Route-setting choroid-plexus perfusion proxy | Early healthy-volunteer dynamic-ASL feasibility route rather than a healthy-aging reference or transport-side study. | Constrains apparent choroid-plexus blood flow and ATT under a named dynamic ASL perfusion model. | Healthy-aging baseline, blood-to-CSF transport truth, epithelial-transporter identity, BBB equivalence, and route-free clearance language. |
| Human choroid-plexus perfusion healthy-aging extension Sun et al. (2024) |
Healthy-aging choroid-plexus perfusion proxy | Large HCP-Aging cohort route linking choroid-plexus volume, perfusion, and diffusion across adulthood rather than an early route-setting or transport-model-separation study. | Constrains age-related declines in choroid-plexus CBF and their relation to volumetric / diffusion changes under one healthy-aging MRI framework. | Blood-to-CSF transport truth, epithelial-transporter identity, route-free BCSFB function, same-subject disease-burden tracking, and BBB equivalence. |
| Human blood-to-CSF water-transport route Petitclerc et al. (2021) |
Blood-to-CSF transport proxy | Specialized healthy-subject ultra-long-TE ASL route with narrow route-setting scope. | Constrains blood-to-CSF water transport under the named ultra-long-TE ASL route. | Choroid-plexus perfusion truth, DCE water cycling, generic BBB exchange, and whole-brain clearance capacity. |
| Human choroid-plexus DCE water-cycling route Anderson et al. (2022) |
Choroid-plexus water-cycling proxy | Older-adult DCE route with cognitive-dysfunction context and explicit exchange-model burden. | Constrains choroid-plexus water-efflux rate under DCE-MRI with separate leakage terms. | Perfusion equivalence, blood-to-CSF transport equivalence, generic clearance truth, and epithelial-controller identity. |
| Human apparent BCSFB-exchange repeatability route Wu et al. (2026) |
Repeatability-bounded apparent BCSFB-exchange proxy | Proof-of-principle healthy-volunteer REXI route whose key added role is scan-rescan repeatability rather than disease contrast or boundary separation. | Constrains apparent kBCSFB and its route-local repeatability under the named REXI protocol. | Route-free BCSFB function, BBB water-exchange equivalence, generic barrier truth, broad deployment, and epithelial-controller identity. |
| Human simultaneous BBB-versus-BCSFB exchange route Petitclerc et al. (2026) |
Boundary-separated BBB-versus-BCSFB exchange proxy | Boundary-separation / model-comparison route rather than a repeatability anchor, healthy-reference baseline, or route-free transport scalar. | Constrains simultaneous Kbl→GM and Kbl→CSF under a multi-TE ASL route. | One generic barrier row, route-free transport equivalence, and direct neurovascular / choroid-plexus controller measurement. |
| Human SMBT-1 first-in-human MAO-B target-validation route Villemagne et al. (2022) |
MAO-B-related target-validation proxy | Healthy-volunteer first-in-human dynamic PET with blockade evidence; still validation-stage rather than routine disease calibration. | Constrains that SMBT-1 can read MAO-B-linked binding in human brain under the named validation route. | AD-context transfer, stable brain-quantification regime, whole-body biodistribution, I2BS burden, astrocyte-ensemble identity, and a route-free whole-brain astrocyte-state scalar. |
| Human SMBT-1 AD-spectrum MAO-B disease-context route Villemagne et al. (2022); Jaisa-Aad et al. (2024) |
MAO-B-related disease-context proxy | AD-continuum and MAO-B-targeted cohort imaging show that interpretation remains disease- and tissue-pathology-specific rather than one general human baseline. | Constrains SMBT-1 MAO-B-linked reactive-astrogliosis contrast only under the named AD-spectrum and tissue-pathology regime. | Route-free astrocyte-state language, I2BS burden, astrocyte-ensemble identity, and transfer across unrelated tracer families or disease / smoking / severity regimes. |
| Human SL25.1188 MAO-B disease / severity route Matsuoka et al. (2026); Best et al. (2026) |
SL25.1188 MAO-B disease- / covariate-conditioned proxy | AD and AUD cohort imaging show that interpretation remains tracer-family-specific and depends on simplified quantification choices plus cohort covariates such as severity and smoking. | Constrains SL25.1188 MAO-B-linked contrast only under the named disease, quantification, and covariate regime. | SMBT-1 transfer, I2BS burden, astrocyte-ensemble identity, and route-free whole-brain astrocyte-state language. |
| Human SMBT-1 brain-quantification route Hiraoka et al. (2025) |
MAO-B-related brain-quantification proxy | Dedicated quantification study with explicit kinetic-versus-simplified modeling burden; route-setting rather than disease-general deployment. | Constrains which named quantification regime can summarize SMBT-1 brain uptake. | Cross-centre transfer, disease-context calibration, whole-body biodistribution, I2BS burden, astrocyte-ensemble identity, and a route-free astrocyte-state scalar. |
| Human SMBT-1 whole-body biodistribution route Mesfin et al. (2026) |
MAO-B-related whole-body biodistribution / tracer-burden proxy | Healthy-volunteer whole-body dynamic PET with limited cohort and long acquisition burden. | Constrains organ distribution, excretion profile, and radiation burden of SMBT-1. | Brain target validity, disease-context contrast, I2BS burden, astrocyte-ensemble identity, and a route-free whole-brain astrocyte-state scalar. |
| Human I2BS astrocyte PET route Tyacke et al. (2018); Livingston et al. (2022) |
I2BS-related astrocyte proxy | Tracer-specific PET with idazoxan-competition interpretation and disease-regime dependence. | Constrains I2BS-related binding under the named tracer / cohort regime. | MAO-B target-validation or disease-context transfer, SMBT-1 dosimetry, astrocyte-ensemble identity, and a route-free whole-brain astrocyte-state scalar. |
| Human TSPO disease-context / validation-bounded PET Biechele et al. (2023); Wijesinghe et al. (2025) |
TSPO disease-context / validation-bounded neuroimmune proxy | Disease-context PET route whose safe reading depends on species ceiling, cohort context, tracer family, and post-mortem validation rather than a route-free human microglia scalar. | Constrains TSPO-linked neuroimmune contrast only under the named disease context and validation regime. | Universal human activation-state truth, route-free transfer across diseases and tracers, local immune-controller identity, and synapse-specific maintenance. |
| Human CSF1R route-setting PET Horti et al. (2022); Ogata et al. (2025) |
CSF1R route-setting neuroimmune proxy | First-in-human PET route with arterial-input and model-burden dependence rather than a settled whole-brain immune-state meter. | Constrains CSF1R-linked target-defined neuroimmune binding under the named tracer and quantification route. | Universal microglia-state truth, route-free transfer across tracers, local immune-controller identity, and synapse-specific maintenance. |
| Human COX-2 enzyme-defined PET Yan et al. (2025) |
COX-2 enzyme-defined neuroimmune proxy | Healthy-human PET quantification route with celecoxib blockade and explicit arterial / reference-tissue modeling choices. | Constrains COX-2-linked enzyme-defined neuroimmune binding under the named blockade and modeling regime. | Generic microglia-state truth, route-free immune-controller identity, and synapse-specific maintenance. |
| Sleep-state CSF-oscillation route Fultz et al. (2019) |
Macro support-state / CSF-motion proxy | Specialized simultaneous sleep-physiology route rather than a solute-specific assay. | Constrains macroscopic sleep-linked CSF-motion coupling. | Molecular efflux, local immune control, and synapse-level maintenance state. |
| Human parenchyma-CSF water-exchange route Kim, Huang, & Liu (2025) |
Macro clearance-transport / water-exchange proxy | Small-sample MT spin-labeling route rather than routine protein-clearance imaging. | Constrains parenchyma-CSF water exchange under the named MT spin-labeling route. | Protein-specific efflux, local immune control, and synapse-level maintenance state. |
| Human respiration-conditioned CSF net-flow route Lim et al. (2025) |
Macro clearance-transport / net-flow proxy | Awake breathing-manipulation 2D PC-MRI route with an explicitly plane-specific ceiling. | Constrains breathing-conditioned CSF displacement / net flow in the measured plane. | Whole-brain bulk circulation, route-free directional transport, and local immune-controller identity. |
| Human exercise-conditioned contrast-influx route Yoo et al. (2025) |
Intervention-conditioned clearance-transport proxy | Small healthy-young contrast-based route with manual ROI steps. | Constrains exercise-conditioned contrast influx and parasagittal meningeal-lymphatic-flow changes under the named route. | Route-free baseline glymphatic flux, natural-sleep clearance, and local immune-controller identity. |
| Human intrathecal tracer / CSF-to-blood clearance route Eide et al. (2023) |
Model-conditioned intrathecal clearance-capacity proxy | Clinical exogenous-tracer route with serial MRI and population pharmacokinetic modeling. | Constrains intrathecal-tracer retention / CSF-to-blood-clearance relations under the named clinical route. | Natural-sleep whole-brain clearance, local immune control, and synapse-level maintenance state. |
| CSF mobility MRI Hirschler et al. (2025) |
Macro support-state / mobility proxy | Specialized 7 T route with a CSF-specific sequence and region-specific driver analyses. | Constrains region-specific CSF mobility and its candidate physiological drivers. | Net clearance flux, cell-specific immune control, and synapse-level maintenance state. |
| Sleep-linked glymphatic efflux to plasma Dagum et al. (2026) |
Model-based human biomarker-efflux route | Intervention-backed crossover design, but still device-dependent and model-heavy. | Constrains the sleep-linked directionality of brain-to-plasma biomarker transport under a multicompartment model. | Spatial localization, local clearance topology, and cell-specific synaptic-maintenance controller identity. |
The safe ceiling of a proxy bundle is not set by the most impressive hardware description or by the number of rows listed. It is set by the narrowest well-supported calibrator role that survives matched-condition comparison, common-driver audit, and external calibration, after any family-internal quantity / target / transport split has also been made explicit. Without that separation, proxy-rich is too easy to misread as broad hidden-state calibration.
Causal relevance is not yet human controller measurement
One shortcut still remained even after quantity type, model burden, and calibrator role were separated. Readers could still place a strong local rodent causal paper beside a living-human proxy paper and let them sound additive, as if the responsible controller had now been measured in humans. The current primary literature does not support that jump.
On the causal side, Williamson et al. (2025) showed that learning-associated hippocampal astrocyte ensembles regulate recall, Dewa et al. (2025) showed that emotional-memory-associated astrocytic ensembles contribute to multiday stabilization, and Bukalo et al. (2026) showed that basolateral-amygdala astrocyte Ca2+ signaling supports fear-memory retrieval / extinction representations. On the human side, Villemagne et al. (2022) established a first-in-human MAO-B SMBT-1 route with pharmacological blockade, Villemagne et al. (2022) measured MAO-B-linked reactive-astrogliosis burden across the Alzheimer disease spectrum, Hiraoka et al. (2025) fixed the brain-side quantification problem, Mesfin et al. (2026) added the whole-body biodistribution burden, Matsuoka et al. (2026) added a separate SL25.1188 simplified arterial-free AD quantification route, Tyacke et al. (2018) established a first-in-human I2BS PET route with dose-dependent idazoxan blockade and no isocarboxazid block, Livingston et al. (2022) showed region- and impairment-stage-dependent BU99008 behavior, Best et al. (2026) showed that SL25.1188 binding in AUD shifts with severity and daily cigarette use, Fultz et al. (2019) measured macroscopic CSF oscillation, Kim, Huang, & Liu (2025) measured parenchyma-CSF water exchange, Lim et al. (2025) measured respiration-conditioned net flow, Yoo et al. (2025) measured exercise-conditioned contrast influx and parasagittal meningeal-lymphatic flow, Eide et al. (2023) modeled intrathecal-tracer / CSF-to-blood clearance variables, Hirschler et al. (2025) measured region-specific CSF mobility, and Dagum et al. (2026) inferred overnight brain-to-plasma biomarker transport under a multicompartment model. Those human routes improve observability, but they do not directly reveal which astrocyte ensemble, clearance controller, or synapse-support mechanism generated the causal effect in the rodent studies.
| Evidence layer | What it directly supports | What it still does not support |
|---|---|---|
| Local rodent causal controller papers | Show that a named support-state family can matter causally for recall, stabilization, or fear-memory dynamics in a specific circuit and preparation. | Do not by themselves provide living-human whole-brain observability or a target-defined human readout. |
| Human target-defined astrocyte / neuroimmune PET / clearance proxies | Show that a bounded human route can observe target-defined astrocyte binding, target-defined neuroimmune PET binding, CSF mobility, or model-based biomarker efflux under a stated tracer / sequence / model. | Do not by themselves identify the local controller, cell ensemble, or causal mechanism that mattered in the animal studies. |
| Combined narrative without an explicit bridge | Supports at most that the state family is causally important somewhere and partly observable somewhere else. | Does not justify the sentence "the responsible controller is now measured in humans". |
If a proxy bundle mixes local causal evidence with living-human proxy evidence, this site now asks for a causal-bridge statement: name (1) the state family made causally relevant, (2) the human direct observable, (3) the species / spatial-unit / timescale gap, (4) the bridge assumption that would connect them, and (5) the controller that still remains latent in humans. Without that bridge statement, the bundle stays at causally relevant somewhere + human proxy somewhere, not measured human controller state.
Same-session multimodal does not erase the composition problem
It is tempting to think that if several modalities are acquired simultaneously, the composition problem disappears. The current primary literature does not support that shortcut. Simultaneous acquisition narrows one class of timing mismatch, but it still leaves open what is shared, what remains modality-specific, and what common factors may come from physiology rather than the target neural variable.
Vafaii et al. (2024) used simultaneous wide-field Ca2+ imaging and fMRI-BOLD in mice and found that the modalities reveal both common and divergent features of network organization. Chen et al. (2025) used simultaneous EEG-PET-MRI in humans and found tightly coupled global hemodynamic and metabolic progressions during descent into NREM sleep, while also identifying two distinct network patterns rather than one uniform multimodal state. Bolt et al. (2025) further showed that a major low-frequency global fMRI mode is substantially coupled to autonomic arousal physiology. Epp et al. (2025) then showed in same-session quantitative fMRI that approximately 40% of gray-matter voxels with significant ΔBOLD exhibited opposing oxygen-metabolism changes. On this site, that means same-session multimodal acquisition still needs a Fusion Card and does not auto-upgrade a result to same-subject state closure. It also means a proxy bundle still has to disclose whether the apparent agreement survives a common-driver audit, whether the quantity types remain commensurable, and whether the bundle adds anything beyond the strongest single row.
Same-subject is not yet same-state when the bridge is sequential
The next shortcut to block is subtler than ordinary multimodal overreading. Even if two measurements come from the same subject or the same brain, they still do not automatically form one state sample when the bridge itself is sequential. The problem is not only that the modalities differ. The problem is that the pipeline can cross time, physiological regime, and tissue transformation before the second stage is acquired.
Lu et al. (2023) showed that preservation route and fixation time course materially change extracellular-space retention and native geometry. Bosch et al. (2022) then showed that linking in vivo two-photon physiology to synchrotron microtomography and serial block-face EM requires a multistage correlative workflow with explicit landmarks and targeted subvolumes. MICrONS Consortium et al. (2025) showed that same-brain function plus EM remains a sequential local pipeline rather than simultaneous whole-state capture. Finally, Attardo et al. (2015) showed that adult CA1 dendritic spines are not static over relaxed windows, with mean lifetimes on the order of 1-2 weeks. On this site, that means same-subject wording solves specimen identity only; it does not solve state continuity by itself.
| Bridge failure mode | Why it matters | What must be disclosed instead |
|---|---|---|
Time-gap collapsesame subject = same time |
A live measurement and a later fixation or cross-day reacquisition can cross windows in which synaptic, excitability, sleep-related, or support-state variables drift. | Name the exact elapsed time and which hidden-state families could change inside that window. |
Regime-collapsesame specimen = same physiological state |
Arousal, anesthesia, sleep deprivation, task condition, pharmacology, or recovery status can differ across the bridge even when the specimen is the same. | Name whether the regime was matched, intentionally shifted, or left unmatched. |
Geometry-collapsesame brain = same coordinates |
Fixation, shrinkage, sectioning, and correlative registration change how points in live space are mapped to ex vivo space. | Name landmarks, deformation model, manual versus automated correspondence, and unresolved mismatch. |
Validation-collapsecorrelative workflow = validated bridge |
A bridge can be carefully engineered and still remain unvalidated for the exact state claim being made. | Name the bridge validation rung: repeated live measurement, vessel / cell recovery, stimulation-site correspondence, histology check, or none. |
If the claim depends on treating a sequential bridge as one latent-state sample, this site now asks for the Verification: State-Continuity Bridge Card. Without that card, the result stays at the strongest directly supported live or destructive stage plus, at most, an unvalidated bridge hypothesis; it is not promoted to same-state cross-regime evidence. If the bridge problem itself is your main question, the longer background is in Wiki: State-Continuity Bridge.
What must be fixed before a proxy bundle is promoted
| Required disclosure | Why it is required here |
|---|---|
| Claimed latent variable | Without naming the target variable, routes that constrain different objects are silently mixed. |
| Proxy class of each row | Prevents structural scaffold, mixed-arousal proxy, receptor / transporter atlas prior, target-engagement occupancy route, challenge-linked release proxy, density proxy, similarity scaffold, metabolite / pH balance route, exchange-flux route, NAD-content / dynamics routes, absolute-metabolite mapping, kinetic-rate imaging, quantity-defined myelin family, BBB water-exchange route, tracer-specific BBB transport route, choroid-plexus perfusion / transport / water-cycling / exchange routes, target-defined astrocyte PET, mobility proxy, and model-based efflux from being treated as the same kind of evidence. |
| Human evidence class of each perturbation-conditioned row | Prevents a local clinical-unit allocation route, a sleep-history / plasticity-recalibration route, and a state-gated perturbation route from being silently treated as one human excitability meter. |
| Direct observable of each stack | Prevents density, similarity, energetic balance, absolute metabolite map, kinetic rate, water exchange, tracer-specific transport, target-defined binding, mobility, efflux, local-unit firing allocation, plasticity-assay outcome, and corticospinal response from being treated as synonyms. |
| Same-subject / same-session / same-perturbation relation | Stops cross-cohort or cross-paper evidence from being rephrased as if it were one joint measurement in one person. |
| Effective time window / state-axis compatibility | Stops static scaffolds, scan-window averages, task-evoked local dynamics, minutes-long kinetics, and overnight perturbation routes from being silently compressed into one current state sample. |
| Physiological / perturbation regime compatibility | Stops rest, task, sleep stage, sleep deprivation, pharmacological challenge, and severity-specific rows from being treated as the same biological condition without an explicit bridge. |
| Model burden and acquisition burden | Kinetic models, blood inputs, high-field hardware, investigational devices, and specialized sequences all change the practical meaning of the result. |
| Per-row repeatability / reproducibility window | Stops a bundle from leaning on a row whose stability was never shown under its actual tracer, reconstruction, field strength, or processing route. |
| Cross-center / cross-scanner transfer | Stops a same-lab bundle from being overread as portable human observability when the relation was never reproduced across centres, scanners, or protocol families. |
| Partial-availability / missing-row slice | Stops a full bundle from being treated as generally available when the decisive comparison depended on a narrow complete-case subset or collapses once one row is unavailable. |
| Operational maturity | Stops proof-of-principle, specialized high-field, and routine clinical-style routes from being silently treated as equally deployable or equally reproducible. |
| Calibrator role | Stops a real human route from being overread as if it calibrated every hidden-state family, rather than one bounded family such as regional density, parcel similarity, macro energetic balance, deuterium rate, quantity-defined myelin, local clinical-unit allocation, perturbation-conditioned plasticity support, state-gated corticospinal responsiveness, tracer-specific transport, target-defined astrocyte burden, or support-state mobility. |
| Causal-bridge statement when local animal causality and human proxy evidence are mixed | Stops “causally important somewhere + human proxy somewhere” from being rewritten as measured human controller state without naming the bridge assumption and residual latent controller. |
| External calibration or perturbation route | Needed to separate a shared statistical factor from a validated biological variable. |
| Common-driver / independence audit | Prevents cross-row agreement from being read as one solved latent variable when the bundle may still be dominated by autonomic, vascular, motion-linked, or shared-preparation factors. |
| Agreement / disagreement topology and resolution policy | Stops an overall bundle gain from hiding that modalities disagree most in the hard strata, and forces the paper to say whether discordant cases were abstained, adjudicated, followed up, or silently absorbed into one final score. |
| Increment over the strongest single row | Stops row diversity from being mistaken for state closure when the bundle has not been shown to add anything beyond the strongest single route under matched conditions. |
| Residual latent-state ceiling | Keeps transcription / chromatin state, RNA-state, phospho-signaling, proteostasis, chloride homeostasis, mitochondrial positioning, and local immune control explicit when they remain unresolved. |
If these disclosures are missing, a proxy bundle remains at proxy-rich but ceiling-limited human evidence. It is not promoted to same-subject cross-stack state identification, unique internal-state recovery, or state-complete readout. In particular, row diversity without explicit proxy class, human evidence class when perturbation-conditioned routes are cited, effective time window / state-axis compatibility, physiological / perturbation regime compatibility, operational maturity, calibrator role, causal-bridge statement when mixed-species causal evidence is used, row-level repeatability, cross-centre transfer, partial-availability disclosure, common-driver audit, agreement / disagreement topology plus resolution policy, and a clear increment over the strongest single row is not treated as real state closure. The public-facing implementation of that rule is the Verification: Human Proxy Composition Card.
References (main)
- Lu, X., Han, X., Meirovitch, Y., et al. (2023). Preserving extracellular space for high-quality optical and ultrastructural studies of whole mammalian brains. doi:10.1016/j.crmeth.2023.100520
- Shapson-Coe, A., Januszewski, M., Berger, D. R., et al. (2024). A petavoxel fragment of human cerebral cortex reconstructed at nanoscale resolution. doi:10.1126/science.adk4858
- Bosch, C., Pacureanu, A., Patiño, J., et al. (2022). Functional and multiscale 3D structural investigation of brain tissue through correlative in vivo physiology, synchrotron microtomography and volume electron microscopy. doi:10.1038/s41467-022-30199-6
- Naganawa, M., Li, S., Nabulsi, N., et al. (2021). First-in-human evaluation of 18F-SynVesT-1, a radioligand for PET imaging of synaptic vesicle glycoprotein 2A. doi:10.2967/jnumed.120.249144
- Johansen, A., Beliveau, V., Colliander, E., et al. (2024). An in vivo high-resolution human brain atlas of synaptic density. doi:10.1523/JNEUROSCI.1750-23.2024
- Smart, K., Liu, H., Matuskey, D., et al. (2021). Binding of the synaptic vesicle radiotracer [11C]UCB-J is unchanged during functional brain activation using a visual stimulation task. doi:10.1177/0271678X20946198
- Holmes, S. E., Scheinost, D., Finnema, S. J., et al. (2022). Imaging the effect of ketamine on synaptic density (SV2A) in the living brain. doi:10.1038/s41380-022-01465-2
- Shatalina, E., Onwordi, E. C., Whitehurst, T., et al. (2024). The relationship between SV2A levels, neural activity, and cognitive function in healthy humans: A [11C]UCB-J PET and fMRI study. doi:10.1162/imag_a_00190
- Snellman, A., Tuisku, J., Koivumäki, M., et al. (2024). SV2A PET shows hippocampal synaptic loss in cognitively unimpaired APOE ε4/ε4 homozygotes. doi:10.1002/alz.14327
- Carro-Domínguez, M., Huwiler, S., Oberlin, S., et al. (2025). Pupil size reveals arousal level fluctuations in human sleep. doi:10.1038/s41467-025-57289-5
- Hansen, J. Y., Shafiei, G., Markello, R. D., et al. (2022). Mapping neurotransmitter systems to the structural and functional organization of the human neocortex. doi:10.1038/s41593-022-01186-3
- Nakuci, J., & Bansal, K. (2025). The neuroreceptors and transporters underlying spontaneous brain activity. doi:10.1038/s42003-025-08492-z
- Wong, D. F., Kuwabara, H., Brašić, J. R., et al. (2013). Determination of dopamine D2 receptor occupancy by lurasidone using positron emission tomography in healthy male subjects. doi:10.1007/s00213-013-3103-z
- Koepp, M. J., Gunn, R. N., Lawrence, A. D., et al. (1998). Evidence for striatal dopamine release during a video game. doi:10.1038/30498
- Erritzoe, D., Ashok, A. H., Searle, G. E., et al. (2020). Serotonin release measured in the human brain: a PET study with [11C]CIMBI-36 and d-amphetamine challenge. doi:10.1038/s41386-019-0567-5
- Miederer, I., Buchholz, H.-G., Rademacher, L., et al. (2025). Dopaminergic Mechanisms of Cognitive Flexibility: An [18F]Fallypride PET Study. doi:10.2967/jnumed.124.268317
- Schlosser, G., Murgaš, M., Godbersen, G. M., et al. (2025). Human in vivo assessment of ketamine binding of the serotonin transporter-follow up at a higher dose. doi:10.3389/fnins.2025.1651016
- Lucchetti, F., Céléreau, E., Steullet, P., et al. (2025). Constructing the human brain metabolic connectome with MR spectroscopic imaging reveals cerebral biochemical organization. doi:10.1038/s41467-025-66124-w
- Guo, R., Li, Y., Zhao, Y., et al. (2025). High-Resolution Brain Metabolic Imaging at Ultrahigh Field Using Extended Spatiospectral Encoding and Subspace Modeling. doi:10.1109/TBME.2025.3572448
- Ren, J., Sherry, A. D., & Malloy, C. R. (2015). 31P-MRS of healthy human brain: ATP synthesis, metabolite concentrations, pH, and T1 relaxation times. doi:10.1002/nbm.3384
- Ren, J., Sherry, A. D., & Malloy, C. R. (2017). Efficient 31P band inversion transfer approach for measuring creatine kinase activity, ATP synthesis, and molecular dynamics in the human brain at 7 T. doi:10.1002/mrm.26560
- Guo, R., Yang, S., Wiesner, H. M., et al. (2024). Mapping intracellular NAD content in entire human brain using phosphorus-31 MR spectroscopic imaging at 7 Tesla. doi:10.3389/fnins.2024.1389111
- Kaiser, A., Vind, F. A., Duarte, J. M. N., et al. (2026). Ultra-high field 31P functional magnetic resonance spectroscopy reveals NAD+ dynamics in brain energy metabolism during visual stimulation. doi:10.1177/0271678X261415784
- Li, X., Zhu, X.-H., Li, Y., et al. (2025). Quantitative mapping of key glucose metabolic rates in the human brain using dynamic deuterium magnetic resonance spectroscopic imaging. doi:10.1093/pnasnexus/pgaf072
- Karkouri, J., Novoselova, M., Rodgers, C. T., et al. (2026). Absolute Quantification of Brain Deuterium Metabolic Imaging in Healthy Volunteers and Glioblastoma Patients at 7T. doi:10.1002/mrm.70308
- Ahmadian, N., Karkouri, J., Deelchand, D. K., et al. (2025). Human Brain Deuterium Metabolic Imaging at 7 T: Impact of Different [6,6'-2H2]Glucose Doses. doi:10.1002/jmri.29532
- Arshad, M., Stanley, J. A., & Raz, N. (2017). Test-retest reliability and concurrent validity of in vivo myelin content indices: Myelin water fraction and calibrated T1w/T2w image ratio. PMCID:PMC5342928
- Hagiwara, A., Hori, M., Kamagata, K., et al. (2018). Myelin Measurement: Comparison Between Simultaneous Tissue Relaxometry, Magnetization Transfer Saturation Index, and T1w/T2w Ratio Methods. doi:10.1038/s41598-018-28852-6
- Baadsvik, E. L., Weiger, M., Froidevaux, R., et al. (2024). Myelin bilayer mapping in the human brain in vivo. doi:10.1002/mrm.29998
- Galbusera, R., Weigel, M., Bahn, E., et al. (2025). Quantitative T1 is sensitive to cortical remyelination in multiple sclerosis: A postmortem MRI study. doi:10.1111/bpa.70010
- Colaes, R., Radwan, A., Billiet, T., et al. (2026). Evaluating the T1w/FLAIR ratio as a proxy for myelin: Associations with myelin water imaging, diffusion metrics, and cognition. doi:10.1007/s00234-025-03875-9
- Morgan, C. A., Thomas, D. L., Shao, X., et al. (2024). Measurement of blood-brain barrier water exchange rate using diffusion-prepared and multi-echo arterial spin labelling: Comparison of quantitative values and age dependence. doi:10.1002/nbm.5256
- Padrela, B. E., Slivka, M., Sneve, M. H., et al. (2025). Blood-brain barrier water permeability across the adult lifespan: A multi-echo ASL study. doi:10.1016/j.neurobiolaging.2024.12.012
- Chung, K. J., Abdelhafez, Y. G., Spencer, B. A., et al. (2025). Quantitative PET imaging and modeling of molecular blood-brain barrier permeability. doi:10.1038/s41467-025-58356-7
- Farinas, A., Rutledge, J., Bot, V. A., et al. (2025). Disruption of the cerebrospinal fluid-plasma protein balance in cognitive impairment and aging. doi:10.1038/s41591-025-03831-3
- Villemagne, V. L., Harada, R., Dore, V., et al. (2022). First-in-Humans Evaluation of 18F-SMBT-1, a Novel 18F-Labeled Monoamine Oxidase-B PET Tracer for Imaging Reactive Astrogliosis. doi:10.2967/jnumed.121.263254
- Villemagne, V. L., Harada, R., Dore, V., et al. (2022). Assessing reactive astrogliosis with 18F-SMBT-1 across the Alzheimer disease spectrum. doi:10.2967/jnumed.121.263255
- Hiraoka, K., Mesfin, B., Wu, Y., et al. (2025). Kinetic and quantitative analysis of [18F]SMBT-1 PET imaging for monoamine oxidase B. doi:10.1007/s12149-025-02083-y
- Mesfin, B., Ishioka, Y., Ichinose, Y., et al. (2026). Whole-body biodistribution of [18F]SMBT-1: a novel PET tracer for monoamine oxidase B imaging in healthy humans. doi:10.1007/s12149-025-02144-2
- Matsuoka, K., Takado, Y., Kimura, Y., et al. (2026). Quantification of monoamine oxidase B expression with 11C-SL25.1188 for imaging reactive astrocytes in patients with Alzheimer's disease. doi:10.1007/s00259-025-07542-2
- Tyacke, R. J., Myers, J. F. M., Venkataraman, A., et al. (2018). Evaluation of 11C-BU99008, a PET ligand for the imidazoline2 binding site in human brain. doi:10.2967/jnumed.118.208009
- Livingston, N. R., Calsolaro, V., Hinz, R., et al. (2022). Relationship between astrocyte reactivity, using novel 11C-BU99008 PET, and glucose metabolism, grey matter volume and amyloid load in cognitively impaired individuals. doi:10.1038/s41380-021-01429-y
- Best, L. M., Truong, J., McCluskey, T., et al. (2026). MAO-B status in alcohol use disorder: a [11C]SL25.1188 PET imaging study of putative astrogliosis. doi:10.1038/s41380-025-03355-9
- Jaisa-Aad, M., Muñoz-Castro, C., Healey, M. A., Hyman, B. T., & Serrano-Pozo, A. (2024). Characterization of monoamine oxidase-B (MAO-B) as a biomarker of reactive astrogliosis in Alzheimer’s disease and related dementias. doi:10.1007/s00401-024-02712-2
- Fultz, N. E., Bonmassar, G., Setsompop, K., Stickgold, R. A., Rosen, B. R., Polimeni, J. R., & Lewis, L. D. (2019). Coupled electrophysiological, hemodynamic, and cerebrospinal fluid oscillations in human sleep. doi:10.1126/science.aax5440
- Kim, D., Huang, Y., & Liu, J. (2025). Non-invasive MRI measurements of age-dependent in vivo human glymphatic exchange using magnetization transfer spin labeling. doi:10.1016/j.neuroimage.2025.121142
- Eide, P. K., Lashkarivand, A., Pripp, A. H., et al. (2023). Plasma neurodegeneration biomarker concentrations associate with glymphatic and meningeal lymphatic measures in neurological disorders. doi:10.1038/s41467-023-37685-5
- Lim, C., Chen, C., Zhang, C., et al. (2025). Human cerebrospinal fluid net flow enhanced by respiration during the awake state. doi:10.1038/s41467-025-66548-4
- Yoo, R.-E., Kim, J.-H., Moon, H. Y., et al. (2025). Long-term physical exercise facilitates putative glymphatic and meningeal lymphatic vessel flow in humans. doi:10.1038/s41467-025-58726-1
- Hirschler, L., Runderkamp, B. A., Decker, A., et al. (2025). Region-specific drivers of CSF mobility measured with MRI in humans. doi:10.1038/s41593-025-02073-3
- Dagum, P., Elbert, D. L., Giovangrandi, L., et al. (2026). The glymphatic system clears amyloid beta and tau from brain to plasma in humans. doi:10.1038/s41467-026-68374-8
- Tallman, C. W., Siler, S. N., Lee, H., et al. (2025). Neuronal allocation and sparse coding of episodic memories in the human hippocampus. doi:10.1038/s41598-025-21967-7
- Huber, R., Mäki, H., Rosanova, M., Casarotto, S., Canali, P., Casali, A. G., Tononi, G., & Massimini, M. (2013). Human cortical excitability increases with time awake. doi:10.1093/cercor/bhs014
- Kuhn, M., Wolf, E., Maier, J. G., Mainberger, F., Feige, B., Schmid, H., et al. (2016). Sleep recalibrates homeostatic and associative synaptic plasticity in the human cortex. doi:10.1038/ncomms12455
- Zrenner, C., Desideri, D., Belardinelli, P., & Ziemann, U. (2018). Real-time EEG-defined excitability states determine efficacy of TMS-induced plasticity in human motor cortex. doi:10.1016/j.brs.2017.11.016
- Khatri, U. U., Pulliam, K., Manesiya, M., Vieyra Cortez, M., Millán, J. del R., & Hussain, S. J. (2025). Personalized whole-brain activity patterns predict human corticospinal tract activation in real-time. doi:10.1016/j.brs.2024.12.1193
- Fehér, K. D., Henckaerts, P., Hirsch, V., Bucsenez, U., Kuhn, M., Maier, J. G., et al. (2026). A nap can recalibrate homeostatic and associative synaptic plasticity in the human cortex. doi:10.1016/j.neuroimage.2026.121723
- Williamson, M. R., Kwon, W., Woo, J., et al. (2025). Learning-associated astrocyte ensembles regulate memory recall. doi:10.1038/s41586-024-08170-w
- Dewa, K., Kaseda, K., Kuwahara, A., et al. (2025). The astrocytic ensemble acts as a multiday trace to stabilize memory. doi:10.1038/s41586-025-09619-2
- Bukalo, O., et al. (2026). Astrocytes enable amygdala neural representations supporting memory. doi:10.1038/s41586-025-10068-0
- Vafaii, H., Mandino, F., Desrosiers-Grégoire, G., et al. (2024). Multimodal measures of spontaneous brain activity reveal both common and divergent patterns of cortical functional organization. doi:10.1038/s41467-023-44363-z
- Chen, J. E., Lewis, L. D., Coursey, S. E., et al. (2025). Simultaneous EEG-PET-MRI identifies temporally coupled and spatially structured brain dynamics across wakefulness and NREM sleep. doi:10.1038/s41467-025-64414-x
- Bolt, T. S., et al. (2025). Autonomic physiological coupling of the global fMRI signal. doi:10.1038/s41593-025-01945-y
- Epp, S. M., Castrillón, G., Yuan, B., et al. (2025). BOLD signal changes can oppose oxygen metabolism across the human cortex. doi:10.1038/s41593-025-02132-9
- Finnema, S. J., Nabulsi, N. B., Mercier, J., et al. (2018). Kinetic evaluation and test-retest reproducibility of [11C]UCB-J, a novel radioligand for positron emission tomography imaging of synaptic vesicle glycoprotein 2A in humans. doi:10.1177/0271678X17724947
- Bøgh, N., Vaeggemose, M., Schulte, R. F., et al. (2024). Repeatability of deuterium metabolic imaging of healthy volunteers at 3 T. doi:10.1186/s41747-024-00426-4
- Holiga, Š., Sambataro, F., Luzy, C., et al. (2018). Test-retest reliability of task-based and resting-state blood oxygen level dependence and cerebral blood flow measures. doi:10.1371/journal.pone.0206583
- Wirsich, J., Jorge, J., Iannotti, G. R., et al. (2021). The relationship between EEG and fMRI connectomes is reproducible across simultaneous EEG-fMRI studies from 1.5T to 7T. doi:10.1016/j.neuroimage.2021.117864
- Amiri, M., Fisher, P. M., Raimondo, F., et al. (2023). Multimodal prediction of residual consciousness in the intensive care unit: the CONNECT-ME study. doi:10.1093/brain/awac335
- Rohaut, B., et al. (2024). Multimodal assessment improves neuroprognosis performance in clinically unresponsive critical-care patients with brain injury. doi:10.1038/s41591-024-03019-1
- Manasova, D., Belloli, L. M. L., Rosenfelder, M. J., et al. (2026). Multimodal multicentre investigation of diagnostic and prognostic markers in disorders of consciousness. doi:10.1093/brain/awaf412
- MICrONS Consortium, et al. (2025). Functional connectomics spanning multiple areas of mouse visual cortex. doi:10.1038/s41586-025-08790-w
- Attardo, A., Fitzgerald, J. E., & Schnitzer, M. J. (2015). Impermanence of dendritic spines in live adult CA1 hippocampus. doi:10.1038/nature14467