Mining balance disorders' data for the development of diagnostic decision support systems

Area of Science:

• Medical Informatics
• Neurology
• Otolaryngology

Background:

• Balance disorders affect a significant patient population, often requiring complex diagnostic pathways.
• Accurate and timely diagnosis of balance disorders is crucial for effective patient management.
• Existing diagnostic processes can be time-consuming and may benefit from structured decision support.

Purpose of the Study:

• To present the methodology for developing the EMBalance diagnostic Decision Support System (DSS).
• To enhance the diagnostic accuracy of balance disorders for general practitioners (GPs) and specialists.
• To provide guidance on data collection and diagnostic steps for balance disorders.

Main Methods:

• Utilized data mining techniques to analyze medical data from patients with balance disorders.
• Integrated diverse data sources including demographics, clinical examinations, and auditory/vestibular tests.
• Developed a DSS to support the diagnostic process for 12 specific balance disorders.

Main Results:

• The EMBalance DSS demonstrated diagnostic accuracies ranging from 59.3% to 89.8% for GPs.
• For experts, the system achieved diagnostic accuracies between 74.3% and 92.1%.
• The system provides recommendations for data acquisition at each diagnostic stage.

Conclusions:

• The EMBalance DSS offers a valuable tool for improving the diagnosis of balance disorders.
• The system effectively supports both general practitioners and medical experts in clinical decision-making.
• Methodology provides a framework for developing similar diagnostic support systems in other medical fields.