We aimed to develop machine learning (ML) algorithms for the automated prediction of post operative ureteroscopy outcomes for paediatric kidney stones based on pre-operative characteristics.
Data from paediatric patients who underwent ureteroscopy for stone treatment by a single experienced surgeon, between 2010 and 2023 in Southampton General Hospital, were retrospectively collected. 15 ML classification algorithms were used to investigate correlations between preoperative characteristics and postoperative outcomes: primary stone free status (SFS, defined as stone fragments <2mm at the end of the procedure confirmed endoscopically and no evidence of stone fragments >2mm at XR KUB or US KUB at 3 months follow up) and complications. For the task of complication and stone status, ensemble model was made out of bagging classifier, extra trees classifier and LDA. Also, a multi-task neural network was constructed for the simultaneous prediction of all post-operative characteristics. Finally, Explainable AI techniques were used to explain the prediction made by the best models.
The ensemble model produced the highest accuracy (90%) in predicting SFS, finding correlation with overall stone size (-0.205), presence of multiple stones (-0.127) and preoperative stenting (-0.102). Complications were predicted by SMOTE oversampled dataset (93.3% accuracy) with relation to preoperative positive urine culture (-0.060) and SFS (0.003). Training the ML for the multi task model, accuracies of 83.3% and 80% were respectively reached.
ML has a great potential of assisting healthcare research, with possibilities to investigate dataset at a higher level. With the aid of this intelligent tool, urologists can implement their practice and develop new strategies for outcome prediction and patient counselling and informed shared decision making. Our model reached an excellent accuracy in predicting SFS and complications in the paediatric population, leading the way to the validation of patient-specific predictive tools.
Journal of endourology. 2024 Jul 23 [Epub ahead of print]
Carlotta Nedbal, Sairam Adithya, Shilpa Gite, Nithesh Naik, Stephen Griffin, Bhaskar K Somani
University Hospital Southampton NHS Foundation Trust, Urology, 117 High street, 23Castleplace, Southampton, United Kingdom of Great Britain and Northern Ireland, SO142EA; ., Symbiosis institute of technology, Pune, India, CSE Department , Pune, India; ., Symbiosis institute of technology, Pune, India, CSE Department , Pune, India; ., Manipal Academy of Higher Education, Mechanical and Manufacturing, Manipal Institute of Technology, Manipal, Karnataka, India, 576104; ., University Hospital Southampton NHS Foundation Trust, Paediatric Urology, Southampton, United Kingdom of Great Britain and Northern Ireland; ., University Hospital Southampton NHS Foundation Trust, Urology, Southampton, United Kingdom of Great Britain and Northern Ireland; .