-
Medical and health sciences
- General diagnostics
-
Engineering and technology
- Other (bio)medical engineering not elsewhere classified
Clinical reasoning is crucial in medical decision-making but is influenced by cognitive biases and knowledge deficits, leading to diagnostic errors. Bayesian clinical reasoning offers a probabilistic approach to support the diagnostic process. However, the predominant reliance on encoded data results in information loss, which negatively impacts the clinical reasoning process.
This study investigates whether textual context, in addition to encoded data, adds value to the diagnostic process. To address this question, a Clinical Decision Support System (CDSS) based on Bayesian networks and neurosymbolic AI (DeepProbLog) is developed. The system will be tested in a pilot study using an existing database from general practices. The effectiveness of the system will be evaluated by comparing clinical reasoning outcomes between practices with and without CDSS implementation. This research contributes to optimizing automated support in clinical decision-making and reducing diagnostic errors.