Evaluation of split renal dysfunction using radiomics based on magnetic resonance diffusion-weighted imaging.

Area of Science:

• Radiology
• Medical Imaging
• Renal Function Assessment

Background:

• Accurate, noninvasive assessment of split renal dysfunction is clinically needed.
• Current methods for evaluating split renal function are limited.

Purpose of the Study:

• To explore the feasibility of diffusion-weighted imaging (DWI)-based radiomics models for evaluating split renal dysfunction.
• To develop and validate radiomics models for discriminating renal function levels.

Main Methods:

• Patients with varying renal function underwent renal DWI.
• Radiomics features were extracted from DWI images to create radiomics scores (Rad-score).
• Models were trained and validated to differentiate between normal, mild, moderate, and severe renal function groups based on GFR measurements.

Main Results:

• Three radiomics models (A, B, C) were developed to distinguish different levels of renal function.
• Models demonstrated efficient discrimination with high areas under the curve (AUC) in both training and testing cohorts (e.g., 0.822-0.897).

Conclusions:

• DWI-based radiomics models show significant potential for evaluating split renal dysfunction.
• These models can effectively discriminate between normal and impaired renal function groups and their subgroups.