Advancing Interstitial Cystitis/Bladder Pain Syndrome (IC/BPS) Diagnosis: A Comparative Analysis of Machine Learning...

Area of Science:

• Urology
• Biomedical Informatics
• Machine Learning

Background:

• Interstitial cystitis/bladder pain syndrome (IC/BPS) is a challenging urinary bladder disorder.
• Accurate diagnosis of IC/BPS is crucial for effective patient management.
• Current diagnostic methods may lack precision, necessitating improved approaches.

Purpose of the Study:

• To enhance machine learning models for IC/BPS diagnosis.
• To compare classical machine learning techniques with advanced AutoML methods.
• To leverage urinary biomarker data and patient-reported outcomes for improved diagnostic accuracy.

Main Methods:

• Applied logistic regression, SVM, random forests, k-NN, and AutoGluon to predict IC/BPS.
• Utilized biomarker data from 2009 participants across two nationwide studies.
• Compared performance of classical ML and AutoML approaches.

Main Results:

• Expanded datasets improved model performance metrics.
• AutoML methods demonstrated superior accuracy over classical techniques.
• Top models achieved a receiver-operating characteristic area under the curve (ROC-AUC) up to 0.96.

Conclusions:

• This study achieved improved model performance for IC/BPS diagnosis compared to prior research.
• Objective urinary biomarker levels were key in the top-performing binary classification model.
• These advancements pave the way for a reliable IC/BPS classification model.