Overall survival model explorer

1,408 patients · 387 events

Patient inputs

Model variables

Prediction model PathoClinPathological IDH status 2.5D-qMR-Clin2.5D-IDHwt score

Age at diagnosis years MGMT promoter Extent of resection Pathological IDH status

Calculation runs entirely in this browser. No patient information is transmitted.

Model output

PathoClin

—

1-year OS—

2-year OS—

3-year OS—

Cox risk score—

Estimated overall survival curve

The risk label uses the optimal training-cohort Cox risk-score cutoff fixed separately for each model. Continuous survival estimates remain the primary output.

Selected clinical factorsAge · MGMT · GTR

Primary comparison2.5D-qMR-Clin vs PathoClin

Validation design1 internal + 3 external cohorts

Model specification Variables, coding, and calculation

Both models estimate 1-, 2-, and 3-year overall survival from the date of initial surgery using a Cox proportional hazards framework. They share the same clinical inputs and differ only in the IDH information source.

PathoClin

Age, MGMT promoter methylation, gross total resection, and pathological IDH status.

2.5D-qMR-Clin

Age, MGMT promoter methylation, gross total resection, and the continuous 2.5D-IDHwt score.

Cox risk scores

PathoClin = 0.573 × (Age / 10) − 0.736 × MGMT methylation − 0.608 × GTR + 0.555 × IDHwt

2.5D-qMR-Clin = 0.413 × (Age / 10) − 0.662 × MGMT methylation − 0.447 × GTR + 0.349 × (2.5D-IDHwt score / 0.1)

Coding: MGMT methylation, GTR, and IDHwt are coded as 1 when present. Age is scaled per 10 years, and the 2.5D-IDHwt score is scaled per 0.1 increment. Individualized survival probabilities are calculated as S(t) = S₀(t)^exp(η), where η is the Cox risk score and S₀(t) is the model-specific baseline survival function.