AI-Powered Support for Sudden Hearing Loss Diagnosis: Achieving Personalized Treatment Planning and Prognosis Prediction

Details

This achievement was realized through collaboration involving institutions such as NTUH-Taipei Branch. Sudden Sensorineural Hearing Loss (SSNHL) is an acute condition where traditional diagnosis struggled to provide objective, personalized treatment planning. The developed system analyzed over 864 patient records, building an AI prediction model using more than 22 variables, including demographics, symptoms, and comorbidities. This provides prognosis prediction tailored to individual patients and a 'Treatment Simulation' function that simulates optimal timing and frequency for middle ear corticosteroid injections. Furthermore, the use of SHAP technology visualizes the AI's decision-making process, making it a trustworthy clinical support tool rather than a 'black box.' The platform adheres to international standards such as FHIR (Fast Healthcare Interoperability Resources) and TW Core IG, enabling seamless integration with electronic health records. This capability suggests broad potential for future application.