Machine learning pipeline for automated segmentation and classification of complicated cystic renal masses on MRI: comparison with radiologists' assessments.

purposeTo develop and validate a machine learning (ML)-based pipeline for automated segmentation and classification of complicated cystic renal masses (cCRMs) on MRI. MATERIALS AND

methodsThis multicenter retrospective study enrolled 275 patients (median age, 48 years; 85 females) with pathologically confirmed 275 cCRMs (203 malignant) who underwent renal MRI from January 2013 to December 2023. cCRMs from one institution were used as a training set (n = 215), while those from the other three institutions served as a test set (n = 60). 3D V-Net and random forest algorithms were employed for segmentation and classification, respectively. Segmentation and classification performance was evaluated using the Dice similarity coefficient (DSC) and the area under the curve (AUC), respectively. Two junior and two senior radiologists independently classified cCRMs in the test set into Bosniak categories II-IV based on the Bosniak classification, version 2019.

resultsIn the test set, the ML pipeline achieved DSC of 0.718 for cCRMs (n = 60) on excretory phase images. Additionally, classification performance of the ML pipeline (AUC = 0.835, 95% confidence interval [CI]: 0.717-0.919) significantly surpassed the junior radiologists (0.835 vs. 0.641, P = 0.042) and matched the senior radiologists (0.835 vs. 0.799, P = 0.684).

conclusionThe ML pipeline demonstrates expert-level diagnostic accuracy for automated segmentation and classification of cCRMs, potentially mitigating interobserver variability while maintaining robust performance across multicenter institutional data.