Development of urination recognition technology based on Support Vector Machine using a smart band

Area of Science:

• Biomedical Engineering
• Urology
• Wearable Technology

Background:

• Urinary behavior monitoring is crucial for diagnosing and managing various urological conditions.
• Current methods for tracking urination patterns can be intrusive or inconvenient for patients.
• The increasing prevalence of wearable devices presents opportunities for non-invasive physiological monitoring.

Purpose of the Study:

• To develop and evaluate a smart band-based algorithm for recognizing urination intervals in women.
• To assess the feasibility of using wearable technology for non-invasive urinary behavior management.
• To establish the accuracy and robustness of a machine learning algorithm for detecting urination events.

Main Methods:

• Development of a smart band algorithm utilizing posture and posture change detection.
• Application of a Radial Basis Function kernel-based Support Vector Machine for data analysis.
• Validation of the algorithm through comparison with actual urination events in 10 female participants over 3 days.

Main Results:

• The smart band algorithm achieved a high average accuracy of 91.0% in recognizing urination.
• The algorithm demonstrated robustness in distinguishing urination events based on body movement patterns.
• The study confirmed the feasibility of using wearable sensors for monitoring urination behavior.

Conclusions:

• The developed smart band algorithm is a highly accurate and robust tool for recognizing urination intervals in women.
• This technology has significant potential for clinical application in characterizing urinary patterns and aiding in patient management.
• Wearable device algorithms detecting sequential body movements offer a promising new methodology for studying human physiological behaviors.