Researchers built algorithms, by using a novel machine learning (ML) approach, to predict patient outcomes after the overactive bladder treatments OnabotulinumtoxinA (OBTX-A) injection and sacral neuromodulation (SNM). They found novel ML algorithms were precise, better than expert urologists in predicting OBTX-A results, and noninferior to expert urologists in predicting SNM results. ML could complement, but not supplant, a physician's judgment.

• Using ROSETTA datasets for overactive bladder patients randomized to OBTX-A or SNM, experts developed novel ML algorithms using reproducing kernel techniques.

• Outstanding accuracy in predicting treatment response was offered by trained algorithms (OBTX-A: AUC 0.95; SNM: 0.88).

• Employment of algorithms enabled accurate prediction of mean decrease in urge urinary incontinence episodes (MSE < 0.15) in OBTX-A and SNM.

• Compared with human experts, algorithms were found to be better in response prediction for OBTX-A.

• Algorithms were noninferior to human experts in response prediction for SNM.