Maintenance Notice

Due to necessary scheduled maintenance, the JMIR Publications website will be unavailable from Wednesday, July 01, 2020 at 8:00 PM to 10:00 PM EST. We apologize in advance for any inconvenience this may cause you.

Who will be affected?

Accepted for/Published in: JMIR Cancer

Date Submitted: Sep 1, 2025
Date Accepted: Jan 5, 2026

The final, peer-reviewed published version of this preprint can be found here:

Reinforcement Learning–Based Digital Therapeutic Intervention for Postprostatectomy Incontinence: Development and Pilot Feasibility Study

Fan F, Huang H, Yan JW, Xu CY, Wu XH, Zhou CM, Wen DD, Huang H, Li WHC, Qiu YH

Reinforcement Learning–Based Digital Therapeutic Intervention for Postprostatectomy Incontinence: Development and Pilot Feasibility Study

JMIR Cancer 2026;12:e83375

DOI: 10.2196/83375

PMID: 41649869

PMCID: 12924038

Reinforcement Learning–Based Digital Therapeutic Intervention for Post-Prostatectomy Incontinence: Development and Pilot Feasibility Study

  • Fan Fan; 
  • Hao Huang; 
  • Jing-Wen Yan; 
  • Chao-Yue Xu; 
  • Xiu-Hua Wu; 
  • Chun-Mei Zhou; 
  • Dan-Dan Wen; 
  • Hai Huang; 
  • William Ho Cheung Li; 
  • Yi-Hong Qiu

ABSTRACT

Background:

Post-prostatectomy urinary incontinence (PPI) affects 4.2% to 87% of prostate cancer patients following robot-assisted radical prostatectomy (RARP), significantly impairing their quality of life (QoL). Despite strong evidence supporting behavioral interventions such as pelvic floor muscle training (PFMT) and bladder diaries, poor adherence and lack of personalized tailoring hinder their effectiveness.

Objective:

This study aims to develop and characterize a novel Clinical Behavioral Intervention-Supporting System (CBISs) that leverages reinforcement learning (RL) to deliver adaptive, personalized rehabilitation for PPI patients.

Methods:

A prospective cohort of 150 PPI patients contributed over 10,000 voiding events recorded in 3-day bladder diaries. Twenty-nine dynamic features across six categories were engineered, including temporal patterns, behavioral ratios, contextual markers, and adherence metrics. An XGBoost-based reinforcement learning model optimized individualized training plans via Bayesian optimization. The CBIS mobile platform integrated four main functional modules: rehabilitation training, bladder diary, assessment tools, and incontinence dermatitis care.

Results:

The CBIS dynamically adjusted interventions through continuous, reinforcement learning-driven optimization, incorporating: (1) patient-specific regimen generation based on multidimensional data, including voiding patterns, fluid intake, and adherence scores; (2) iterative plan updates informed by patient feedback and progress data; (3) clinician oversight portals featuring automated incontinence dermatitis alerts; and (4) SM3/SM4 encryption with hierarchical data access.

Conclusions:

The CBIS represents the first reinforcement learning-powered digital therapeutic system for PPI, enabling adaptive, evidence-based behavioral optimization. By bridging gaps in traditional rehabilitation, such as static protocols and declining adherence, it offers a scalable solution for personalized PPI management. Future multicenter trials are warranted to validate its efficacy.


 Citation

Please cite as:

Fan F, Huang H, Yan JW, Xu CY, Wu XH, Zhou CM, Wen DD, Huang H, Li WHC, Qiu YH

Reinforcement Learning–Based Digital Therapeutic Intervention for Postprostatectomy Incontinence: Development and Pilot Feasibility Study

JMIR Cancer 2026;12:e83375

DOI: 10.2196/83375

PMID: 41649869

PMCID: 12924038

Download PDF


Request queued. Please wait while the file is being generated. It may take some time.

© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.