Loading prediction engine

Connecting to prediction server...

0s, cold boot

About the adjacent segment disease model

Adjacent segment disease (ASD) is degeneration at the level next to a lumbar fusion. This tool predicts ASD-free survival after surgery, with risk at 1, 2, 3 and 5 years and a median time to ASD, and ranks the patient against the training cohort.

Inputs cover diagnoses, the levels fused, surgical technique (open, percutaneous, XLIF, LLIF, ALIF, osteotomies) and spinopelvic alignment (pelvic incidence, PI-LL mismatch, pelvic tilt, lumbar lordosis, SVA, sacral slope). SHAP and time-dependent survSHAP(t) show which factors matter early versus late.

Predicts: ASD-free survival over time; ASD probability at 12, 24, 36 and 60 months. 24 months is the most reliable horizon.
Inputs: Age, BMI, sex, prior surgery, diagnoses, levels fused (T12-L1 to L5-S1), technique, alignment parameters, post-operative complications and length of stay.
Model: Survival forest with principal-component composites for sagittal mismatch, operative burden and post-op alignment.
Cohort: 546 patients, 122 (22%) of whom developed ASD during follow-up.

Key terms

Research demonstration only. Not for clinical decision-making.