MRI-Based Radiomics Analysis of Levator Ani Muscle for Predicting Urine Incontinence after Robot-Assisted Radical Prostatectomy
Background: The exact role of the levator ani (LA) muscle in male continence remains unclear, and so this study aims to shed light on the topic by characterizing MRI-derived radiomic features of LA muscle and their association with postoperative incontinence in men undergoing prostatectomy. Method: In this retrospective study, 140 patients who underwent robot-assisted radical prostatectomy (RARP) for prostate cancer using preoperative MRI were identified. A biomarker discovery approach based on the optimal biomarker (OBM) method was used to extract features from MRI images, including morphological, intensity-based, and texture-based features of the LA muscle, along with clinical variables. Mathematical models were created using subsets of features and were evaluated based on their ability to predict continence outcomes. Results: Univariate analysis showed that the best discriminators between continent and incontinent patients were patients age and features related to LA muscle texture. The proposed feature selection approach found that the best classifier used six features: age, LA muscle texture properties, and the ratio between LA size descriptors. This configuration produced a classification accuracy of 0.84 with a sensitivity of 0.90, specificity of 0.75, and an area under the ROC curve of 0.89. Conclusion: This study found that certain patient factors, such as increased age and specific texture properties of the LA muscle, can increase the odds of incontinence after RARP. The results showed that the proposed approach was highly effective and could distinguish and predict continents from incontinent patients with high accuracy.
Mohammed Shahait,1 Ruben Usamentiaga,2 Yubing Tong,3 Alex Sandberg,4 David I. Lee,5 Jayaram K. Udupa3 and Drew A. Torigian3
Department of Surgery, Clemenceau Medical Center, Dubai P.O. Box United Arab Emirates
Department of Computer Science and Engineering, University of Oviedo, Gijon, Spain
Medical Image Processing Group, Department of Radiology, University of Pennsylvania, Philadelphia, PA USA
Temple Medical School, Temple University, Philadelphia, PA USA
Department of Urology, University of California Irvine, Irvine, CA USA