PM.20.038 – Artificial intelligence (AI)-based personalised diagnosis in women with urinary incontinence

Route: Personalised medicine: the individual at the centre

Cluster question: 095 How can we personalise health care, for example by using biomarkers?

The aim of the proposed project is to develop a digital diagnostic tool that provides a personalised diagnosis based on a multi-dimensional characterisation of women with urinary incontinence (UI) and that leads to better-informed clinical decision making.
UI is very common: Almost one third of middle-aged parous women has bothersome UI or had treatment for UI. Treatments are often not very effective, with e.g. pelvic floor physiotherapy and surgery reaching continence rates of 16% and 70% respectively. Socio-economic impact of UI is significant.
Current diagnoses of UI are overly simplified, have difficulty incorporating its many comorbidities, and thereby tend to result in ineffective treatments. Through advanced AI-based analyses of different types of data sets, the tool will arrive at a much more refined perspective on UI. By identifying and targeting new molecular pathways and quantitative biological parameters, the project will further radically change the currently used array of diagnostic assessments. Through a data-driven optimised sequence of assessments, the digital tool will significantly improve the accuracy of each individual woman’s diagnosis and reduce the misuse of diagnostics. This will lead to faster and more cost-effective diagnoses and lower the risk for harmful treatment.
From a technological point of view, at the end of the project we will have developed a novel artificial intelligence (AI)-based framework for the combination of heterogeneous data sets. The framework will be implemented in a multi-layered architecture that can be used by biomedical researchers, clinicians, and developers alike and can be extended with new AI tools developed by other software developers. It incorporates privacy and the principles of Findable, Accessible, Interoperable, Reusable (FAIR) data and FAIR services by design.

Keywords

artificial intelligence-based framework, comorbidities, data-driven optimised sequence of assessments, digital diagnostic tool, genomics, molecular pathways, Privacy by Design, transcriptomics, urinary incontinence, women

Other organisations

BB4all, Radboud Universiteit Nijmegen (RU), WFIPP

Submitter

Organisation Radboud University medical centre (RUMC), Nijmegen
Name Dr. K.B. (Kirsten) Kluivers
E-mail kirsten.kluivers@radboudumc.nl
Website https://www.radboudumc.nl/en/people/kirsten-kluivers