A cost effective expert system to assist physicians: epileptologists' assistant

Area of Science:

• Medical Informatics
• Artificial Intelligence in Healthcare
• Clinical Decision Support Systems

Background:

• Medical expert systems have shown AI potential but few achieve practical success.
• Practical success requires cost-effective systems handling routine physician workload.
• Technology must be intuitive and user-invisible, anticipating needs.

Purpose of the Study:

• To demonstrate the practical success of expert systems in routine medical care.
• To introduce the Epileptologists' Assistant as a model for practical AI in specialty clinics.
• To improve efficiency and quality of care in a specialty clinic setting.

Main Methods:

• Developed the Epileptologists' Assistant, integrating a graphical user interface with an expert system and database.
• Focused on creating an intuitive system that anticipates user needs.
• Aimed to make technology appear invisible to the end-user.

Main Results:

• The Epileptologists' Assistant system currently reduces physician time by 66%.
• The system enables two nurses and a physician to manage the workload of three physicians.
• The approach aims to increase the overall quality of patient care.

Conclusions:

• Expert systems can achieve practical success by effectively managing routine medical tasks.
• The Epileptologists' Assistant demonstrates a viable model for AI implementation in specialty clinics.
• The ultimate goal is a unified family of AI systems for various medical specialties.