A Computer-Based Platform for Aiding Clinicians in Eating Disorder Analysis and Diagnosis

Area of Science:

• Psychiatry
• Medical Informatics

Background:

• Eating disorders affect 13% of the population, causing severe physical and psychosocial issues.
• Early diagnosis is crucial for effective treatment and preventing long-term health consequences.
• Non-specialist clinicians, particularly in primary care, face challenges in accurately diagnosing eating disorders.

Purpose of the Study:

• To develop a computer-assisted diagnostic tool for eating disorders.
• To provide an accessible diagnostic aid for primary care clinicians.
• To improve the accuracy and timeliness of eating disorder diagnoses.

Main Methods:

• A web-based system was developed using an algorithm trained on expert-diagnosed cases.
• The system automates the diagnostic process by estimating the probability and type of eating disorder.
• The tool provides a diagnostic report to assist clinicians and serve as an educational resource.

Main Results:

• The system provides an automated estimation of eating disorder probability and type.
• The generated report aids clinicians in the diagnostic process.
• The tool functions as an educational resource for clinicians, especially those new to the field.

Conclusions:

• Computer-assisted diagnostic procedures can offer expertise to non-specialist clinicians.
• The developed web-based system can improve early and accurate diagnosis of eating disorders.
• This tool supports clinicians in identifying eating disorders, facilitating timely and effective interventions.