CR-ViT Glioma PFS explorer
Subtype-adaptive prognostic model · three-institution imaging cohort

Clinical–Radiologic
ViT Survival Model

Clinical endpoint Progression-free survival
Development sample 843 training · 1,880 total
Prediction horizon 6–36 months
01

Patient profile

Choose a molecular subtype, then enter its frozen model variables.

Molecular subtype

Preset profiles
02

Prediction profile

Continuous risk, training-derived stratum, and time-specific PFS estimates.

Risk stratum
—
Standardized risk score —
6-month PFS—
12-month PFS—
18-month PFS—
24-month PFS—
30-month PFS—
36-month PFS—
Predicted progression-free survival
Model familySubtype-specific Cox models
Primary integrationClinicoradiologic + 3D-ViT
Validation C-index—
Model specification Variables, coding, and calculation

Cox linear predictor

Calculation: S(t) = S0(t)exp(η). The displayed risk score standardizes η to the training cohort; the low/high label uses the frozen training-cohort median. Ordinal MRI descriptors are coded 0–3 from none/minimal to extensive.