A novel CT-based radiomics approach for kidney function evaluation in ADPKD: a pilot study.

Area of Science:

• Nephrology
• Radiology
• Medical Imaging

Background:

• Autosomal dominant polycystic kidney disease (ADPKD) management requires tools to predict progression to end-stage kidney disease (ESKD).
• Radiomic features from computed tomography (CT) scans offer a potential new avenue for risk stratification in ADPKD patients.

Purpose of the Study:

• To explore the predictive potential of radiomic features from CT scans for kidney function decline in ADPKD patients.
• To develop and evaluate a radiomic model for discriminating rapid versus non-rapid disease progression.

Main Methods:

• Retrospective analysis of 58 ADPKD patients with CT scans for total kidney volume (TKV) assessment.
• Extraction of 217 radiomic features from segmented cystic kidneys.
• Development of a radiomic model using significant features to predict rapid progression based on estimated glomerular filtration rate decline.

Main Results:

• A radiomic feature, F_cm.corr, demonstrated significant association with rapid progression (P=0.04).
• The F_cm.corr model achieved an Area Under the Curve (AUC) of 0.78 and a sensitivity of 0.92.
• This radiomic model outperformed a model based on height-adjusted TKV (ht-TKV), which had an AUC of 0.65 and sensitivity of 0.62.

Conclusions:

• A radiomic model incorporating the F_cm.corr feature effectively discriminates rapid progressors in ADPKD.
• Further validation in larger, external cohorts is necessary to confirm the utility of radiomics in ADPKD management.