Validity Gates for Longitudinal Parkinson Disease Prediction

Actual PPMI records / Baseline-selected comparison

Motor scores of 10 and 11. Different evidence behind them.

P1-A has recent imaging and cognitive measurements; P1-B has no eligible DaT-SPECT value but does have NfL and GCase assay evidence. A missing measurement is not a normal result. Neither record supports a treatment recommendation.

The two complete measured histories are also available in the patient comparison figure.

Selection, score meaning and uncertainty

The pair was selected without future outcomes or forecast errors. The earliest eligible records were matched within sex and analysis fold, with baseline calipers for motor score (3 points), age (5 years), disease duration (2 years) and prior motor slope (2 points/year). The closest clinical pair among 30 candidates was retained. Treatment was not matched; this is an illustration, not a causal comparison.

Evidence coordinates are oriented, training-fold-relative ranks on a scale from -1 to +1, not probabilities or clinical severity grades. Positive values indicate relatively greater evidence on that axis. Reliability summarizes available source weights, with a 24-month age half-life; it is not the probability of a correct forecast. An older source can retain its relative score while losing reliability.

Measurements behind the evidence coordinates
Measurement and scale	What it contributes	Interpretation limit
DaT-SPECT SBR: regional imaging	Specific-binding ratio, used as dopaminergic evidence	Not a direct count of surviving neurons
MoCA: clinical cognition	Cognitive screening score out of 30	Not proof of a specific molecular co-pathology
NfL: fluid biomarker	Neurofilament light evidence within the injury/glial coordinate	NfL alone does not establish inflammation
GCase: enzyme assay	Glucocerebrosidase-related lysosomal evidence	Not a GBA genotype or a treatment-response test
LEDD: treatment record	Levodopa-equivalent daily dose and its availability	An absent record does not mean untreated

The displayed prediction interval belongs only to the separately fitted nested uncertainty model, not to the main point forecasts. Valid interval coverage did not establish reliable error ranking. The cohort-level selective-uncertainty verdict is ABSTAIN; this is not an individualized diagnosis or a claim that either patient failed a test.

The result at a glance

Seven questions produce seven verdicts, not one headline score

Each gate protects a different claim. Passing chronology cannot rescue failed transport, and correct interval coverage cannot substitute for useful error ranking.

ChronologyPass

ParticipantPass

SitePass

MagnitudeInformation only

CalendarFail

Selective UQAbstain

PDBP transportFail

The scientific output is the decision boundary. Dated multimodal history contains a reproducible internal increment, but the effect is clinically tiny, temporally fragile, and not externally transportable in the available export. The protocol therefore stops the claim before deployment language begins.

Study design

Every forecast begins at a dated cutoff

Each patient's record is paused at a dated cutoff. Only measurements already available at that moment may describe the state. Their age and absence remain visible; the future motor examination, 9 to 15 months later, stays hidden until evaluation.

A future-only patient timeline followed by separate participant, site, calendar, uncertainty, and transport gates — The model receives each source's most recent eligible value, age, availability, and observation context. The timeline is a protocol schematic, not an individual record. Every gate permits only its own claim.

Why ordinary validation is insufficient

A repeated clinical history is not a flat table

The same person appears at multiple visits. A scan may be months older than the examination it accompanies. Missingness may encode a study protocol rather than biology. Medication state changes motor scores, and a new enrollment era can change both who is measured and how sources are collected.

These are different scientific risks. The benchmark therefore assigns each one a distinct population, uncertainty interval, and promotion rule.

Analysis	Scale	Question it answers
Rolling-origin benchmark	1,058 people; 5,404 cutoffs	Do dated sources add internal information when every person stays in one fold?
Strict-calendar challenge	442 later-era participants	Does the frozen relation survive a later enrollment and observation regime?
Site-disjoint challenge	951 people across 52 sites	Is the increment dependent on acquisition center?
Gate stress test	8,000 controlled experiments	Does the protocol suppress false promotion and detect leakage or shift?
PDBP clinical transport	77 calibration; 156 untouched test participants	Does a frozen clinical model travel to an independent cohort?

Internal result

The signal is measurable; the improvement is not clinically persuasive

Held-out performance across participant, site, enrollment-era, and strict-calendar evaluations — Dated DaT-SPECT and MoCA history improve internal prediction over clinical trajectory by delta R-squared 0.015 and delta MAE -0.060 MDS-UPDRS Part III points. The interval supports a small information increment, but it misses the prespecified promotion magnitude and becomes uncertain under strict calendar testing.

0.015Internal delta R-squared

-0.060Delta MAE in motor points

Information onlyPrespecified magnitude verdict

A statistically stable improvement can still be practically negligible. The 0.060 point MAE reduction does not establish individual prognosis, patient-management value, or clinical-trial enrichment.

Uncertainty and stress testing

An interval may cover well and still fail to identify safer predictions

Coverage asks whether the future falls inside an interval. Selective prediction asks a harder question: when the model says it is more certain, is it actually less wrong?

Selective-risk curves, participant-level interval coverage, and controlled gate behavior — Source-aware 90% intervals covered every eligible forecast for 90.6% of participants, yet interval width did not rank future error better than the clinical reference and reversed under calendar shift. The locked selective-prediction verdict is ABSTAIN. Across 8,000 controlled experiments, the full gate limited null promotion to 0.35%, detected constructed leakage, and blocked 94.75% of calendar reversals.

External transport

A large archive is not the same as a compatible cohort

PDBP contains valuable Parkinson disease data, but a transport claim requires the same diagnosis, modalities, timing, endpoint, and eligible participants to intersect. File presence alone is not validation.

Model-capacity and external-transport stress tests from PPMI to PDBP — A six-variable clinical model achieved internal PPMI R-squared 0.134 but fell to -0.077 under raw PDBP transport. Recalibration reached R-squared 0.005 and did not beat the PDBP mean reference on MAE. The local imaging export contained no eligible diagnosis-modality-time intersection, so this is a failed transport and interoperability result, not external multimodal validation.

Claim contract

What the evidence permits us to say

Gate	Verdict	Permitted statement
Future-only chronology	Pass	Every predictor precedes its endpoint
Participant separation	Pass	The internal signal is not produced by repeated-person leakage
Site separation	Pass	A similarly small increment survives a center-held-out challenge
Effect magnitude	Information only	Dated sources add an internal increment below promotion magnitude
Calendar stability	Fail	Added value is not established across enrollment era
Selective uncertainty	Abstain	Coverage is adequate, but uncertainty does not reliably rank error
External PDBP transport	Fail	Recalibration and archive size do not establish transport

What this work contributes

A future-only rolling-origin patient representation
Separate people for model fitting, calibration, and testing
Independent gate verdicts instead of one favorable score
Executable stopping decisions under null effects, leakage, and shift

What this work does not claim

Individual clinical prognosis
Calibrated mechanism probability
Treatment selection or treatment benefit
Successful external multimodal validation

Connected research / Different questions, different claims

From patient phenotypes toward mechanism-aware evidence

01 / Pathway context CVC pathway-anchored clustering
Organizes structural, microstructural and dopaminergic imaging by predefined biological pathways.
02 / Patient phenotypes Posterior-aware phenotyping
Addresses clinical heterogeneity and uncertainty in patient representations.
bioRxiv preprint
Related MICCAI project
03 / Biological specificity Signed dopaminergic laterality
Retains direction, tests contralateral agreement and preserves unresolved conflict.
04 / Validity limits Longitudinal validity gates
Tests when a patient-state forecast is informative and when its clinical claim must stop.

Our earlier integrated genetic, molecular and wearable-biomarker study (medRxiv preprint) provides complementary multimodal stratification context. It is distinct from CVC's pathway-anchored imaging study.

This is a research progression, not a single jointly validated diagnostic pipeline. Patient-specific intervention simulation and treatment selection remain future work. The AI4PD program provides the broader systems context; these pages replay fixed research examples, not live clinical predictions.

Data: PPMI; access to underlying participant data follows the source study's application process. Code is available from the authors on request; the paper-specific repository is pd-validity-gates (access required). Putamen outlines adapt the NIH 3D Human Reference Atlas "Brain, Male" (3DPX-020960), based on the Allen Human Reference Atlas under CC BY 4.0, rendered here as neutral two-dimensional views. They are not patient scans.