Back to all papers

Diffusion MRI Radiomics Predict Renal Function Decline in ADPKD.

July 28, 2026pubmed logopapers

Authors

Bargagli M,Altabella L,Fedon Vocaturo M,Negrelli R,Ambrosetti MC,Puppini G,Cavedon C,Scoglio M,Fuster DG,Ferraro PM

Affiliations (5)

  • Department of Nephrology and Hypertension, Inselspital, Bern University Hospital, University of Bern, Bern, Switzerland.
  • Medical Physics Unit, Azienda Ospedaliera Universitaria Integrata, Verona, Italy.
  • Radiology Unit, Azienda Ospedaliera Universitaria Integrata, Verona, Italy.
  • Graduate School for Health Sciences, University of Bern, Bern, Switzerland.
  • Section of Nephrology, Department of Medicine, Università degli Studi di Verona; Nephrology Unit, Azienda Ospedaliera Universitaria Integrata Verona, Italy.

Abstract

Autosomal dominant polycystic kidney disease (ADPKD) is the most common hereditary kidney disorder and a leading cause of end-stage kidney disease (ESKD). Current risk stratification relies on height-adjusted total kidney volume (ht-TKV), which captures cyst burden but not alterations in the noncystic parenchyma. Radiomics enables extraction of quantitative features reflecting tissue microstructure across both cystic and noncystic kidney compartments. We investigated whether radiomic features from apparent diffusion coefficient (ADC) maps improve prediction of renal function decline in ADPKD. We analyzed 112 patients with ADPKD from the Bern ADPKD Registry with baseline magnetic resonance imaging (MRI, T1-, T2-, and diffusion-weighted). Kidneys were manually segmented, and radiomic features were extracted using PyRadiomics. Features were selected using least absolute shrinkage and selection operator, and models were developed using support vector machines with 20-fold cross-validation. The outcome was rapid renal function decline (estimated glomerular filtration rate [eGFR] ≤ -3 ml/min per 1.73 m<sup>2</sup>/year) over a median follow-up of 5.5 years. ADC-based radiomics provided the highest area under the curve (AUC: 0.82), compared with clinical (0.77) and T1/T2w models (0.77). Ensemble models integrating ADC, T1/T2w, and clinical data achieved the best AUC (0.85) with improved calibration. Net reclassification improvement (NRI) analyses demonstrated improved risk reclassification when ADC-derived information was incorporated into clinical and conventional MRI-based models. ADC-derived radiomic features showed promising prognostic value for predicting renal function decline in ADPKD and provided incremental information beyond ht-TKV and current clinical standards when integrated with clinical and conventional MRI-based models, capturing both cystic and noncystic tissue alterations. If confirmed in independent cohorts, diffusion MRI-based radiomics may support earlier risk stratification and treatment decision-making in ADPKD.

Topics

Journal Article

Ready to Sharpen Your Edge?

Subscribe to join 11k+ peers who rely on RadAI Slice. Get the essential weekly briefing that empowers you to navigate the future of radiology.

We respect your privacy. Unsubscribe at any time.