An automatic diagnostic system for the urodynamic study applying in lower urinary tract dysfunction.

Area of Science:

• Urology
• Medical Informatics
• Machine Learning

Background:

• Lower urinary tract dysfunction (LUTD) diagnosis relies on urodynamic studies.
• Accurate and efficient analysis of urodynamic data is crucial for effective patient management.
• Existing diagnostic methods can be time-consuming and require specialized expertise.

Purpose of the Study:

• To develop an automated diagnostic system using machine learning for preliminary analysis of urodynamic studies.
• To evaluate the performance of different machine learning algorithms in diagnosing common LUTD conditions.
• To enhance the diagnostic process for lower urinary tract dysfunction.

Main Methods:

• Inclusion of 527 patients with complete urodynamic data from 2015-2020.
• Utilizing patient age, sex, and 13 urodynamic parameters as input features.
• Applying and comparing Decision Tree (DT), Logistic Regression (LR), and Support Vector Machine (SVM) algorithms.

Main Results:

• Machine learning models achieved high Area Under the Curve (AUC) values for common LUTDs (DT: 0.63-0.98, LR: 0.73-0.99, SVM: 0.64-1.00).
• A specialized classification model for underactive and acontractile detrusor showed strong performance (DT: 0.82-0.85 AUC, SVM: 0.86-0.90 AUC).
• Logistic Regression and Support Vector Machine models exhibited superior diagnostic accuracy, with the best model achieving an average AUC of 0.90 ± 0.08.

Conclusions:

• An automated diagnostic system for urodynamic studies was successfully developed using machine learning.
• The developed system shows promise in assisting and enhancing the diagnosis of lower urinary tract dysfunction.
• This automated approach provides a valuable reference for guiding LUTD treatment strategies.