Mental Health Intent Recognition for Arabic-Speaking Patients Using the Mini International Neuropsychiatric Interview...

Area of Science:

• Digital Health
• Artificial Intelligence in Healthcare
• Mental Health Technology

Background:

• Mental health disorders affect over 10.7% globally, with a 95% surge in digital mental health tool usage during the COVID-19 pandemic.
• Existing digital mental health solutions have limitations and are not universally applicable.
• There's a need for culturally sensitive and advanced diagnostic tools, particularly for underrepresented linguistic groups.

Purpose of the Study:

• To conduct a literature review on current mental health technologies.
• To propose an intelligent intent recognition tool for mental health diagnosis tailored for Arab-speaking patients.
• To evaluate the system's performance and clinical utility.

Main Methods:

• A literature review of state-of-the-art mental health solutions.
• Development of an intent recognition system using BERT and the International Neuropsychiatric Interview (MINI).
• Data collection and experimentation at the Military Hospital of Tunis, Tunisia.

Main Results:

• The proposed system achieved over 92% accuracy and over 94% precision, recall, and F1 scores.
• The tool demonstrated high performance in diagnosing depression, suicidality, panic disorder, social phobia, and adjustment disorder.
• Medical staff found the tool valuable for clinical decision-making and patient prioritization.

Conclusions:

• The developed AI tool shows excellent diagnostic performance for key mental health conditions in Arab-speaking patients.
• This technology can significantly support clinical decision-making and streamline patient management in high-volume healthcare settings.
• The system represents a promising advancement in accessible and accurate AI-driven mental healthcare.