Mithat Ekşi1, Abdullah Hizir Yavuzsan2, İsmail Evren3, Ali Ayten3, Ali Emre Fakir3, Fatih Akkaş3, Kerem Bursali3, Azad Akdağ3, Selcuk Sahin3, Ali İhsan Taşçi3. 1. Department of Urology, University of Health Sciences, Istanbul Bakirkoy Dr. Sadi Konuk Training and Research Hospital, Zuhuratbaba Mh. Tevfik Saglam Cd. No:11 Bakirkoy, Istanbul, Turkey. mithat_eksi@hotmail.com. 2. Department of Urology, University of Health Sciences, Istanbul Şişli Hamidiye Etfal Training and Research Hospital, Halaskargazi Cd, Şişli, 34371, Istanbul, Turkey. 3. Department of Urology, University of Health Sciences, Istanbul Bakirkoy Dr. Sadi Konuk Training and Research Hospital, Zuhuratbaba Mh. Tevfik Saglam Cd. No:11 Bakirkoy, Istanbul, Turkey.
Abstract
PURPOSE: To compare the models developed with a classical statistics method and a machine learning model to predict the possibility of orchiectomy using preoperative parameters in patients who were admitted with testicular torsion. MATERIALS AND METHODS: Patients who underwent scrotal exploration due to testicular torsion between the years 2000 and 2020 were retrospectively reviewed. Demographic data, features of admission time, and other preoperative clinical findings were recorded. Cox Regression Analysis as a classical statistics method and Random Forest as a Machine Learning algorithm was used to create a prediction model. RESULTS: Among patients, 215 (71.6%) were performed orchidopexy and 85 (28.3%) were performed orchiectomy. The multivariate analysis revealed that monocyte count, symptom duration, and the number of previous Doppler ultrasonography were predictive of orchiectomy. Classical Cox regression analysis had an area under the curve (AUC) 0.937 with a sensitivity and specificity of 88 and 87%. The AUC for the Random Forest model was 0.95 with a sensitivity and specificity of 92 and 89%. CONCLUSION: The ML model outperformed the conventional statistical regression model in the prediction of orchiectomy. The ML methods are cheap, and their powers increase with increasing data input; we believe that their clinical use will increase over time.
PURPOSE: To compare the models developed with a classical statistics method and a machine learning model to predict the possibility of orchiectomy using preoperative parameters in patients who were admitted with testicular torsion. MATERIALS AND METHODS: Patients who underwent scrotal exploration due to testicular torsion between the years 2000 and 2020 were retrospectively reviewed. Demographic data, features of admission time, and other preoperative clinical findings were recorded. Cox Regression Analysis as a classical statistics method and Random Forest as a Machine Learning algorithm was used to create a prediction model. RESULTS: Among patients, 215 (71.6%) were performed orchidopexy and 85 (28.3%) were performed orchiectomy. The multivariate analysis revealed that monocyte count, symptom duration, and the number of previous Doppler ultrasonography were predictive of orchiectomy. Classical Cox regression analysis had an area under the curve (AUC) 0.937 with a sensitivity and specificity of 88 and 87%. The AUC for the Random Forest model was 0.95 with a sensitivity and specificity of 92 and 89%. CONCLUSION: The ML model outperformed the conventional statistical regression model in the prediction of orchiectomy. The ML methods are cheap, and their powers increase with increasing data input; we believe that their clinical use will increase over time.
Authors: Gwen M Grimsby; Bruce J Schlomer; Vani S Menon; Lauren Ostrov; Melise Keays; Kunj R Sheth; Carlos Villanueva; Candace Granberg; Daniel Dajusta; Martinez Hill; Emma Sanchez; Clanton B Harrison; Micah A Jacobs; Berk Burgu; Halim Hennes; Linda A Baker Journal: Urology Date: 2018-03-20 Impact factor: 2.649
Authors: Frederic Geissmann; Markus G Manz; Steffen Jung; Michael H Sieweke; Miriam Merad; Klaus Ley Journal: Science Date: 2010-02-05 Impact factor: 47.728
Authors: Bertrand S Y Lian; Caroline C P Ong; Li Wei Chiang; Rambha Rai; Shireen Anne Nah Journal: Eur J Pediatr Surg Date: 2015-10-28 Impact factor: 2.191