Enhancing Prognosis in FSGS and MCD through Computational Tubular Feature Analysis

Highlight

– Computational analysis of tubular features from whole slide images offers enhanced prognostic discrimination in FSGS and MCD compared to conventional assessments.
– Nine specific tubular features, including basement membrane thickness and epithelial flattening, correlate strongly with disease progression and proteinuria remission.
– Tubular morphology changes show a gradient with interstitial fibrosis severity, highlighting spatial relationships important for understanding pathophysiology.
– Integration of deep learning and pathomics enables quantitative, reproducible insights into tubular pathology beyond traditional visual scoring.

Study Background and Disease Burden