Toward responsible artificial intelligence in medicine: Reflections from the Australian epilepsy project

Area of Science:

• Medical Artificial Intelligence (AI)
• Clinical Decision Support Systems
• Neurology and Epilepsy Research

Background:

• Artificial intelligence (AI) is increasingly explored as a clinical support tool in medicine.
• Concerns exist regarding the potential harmful effects and impact of AI on expert medical decision-making.
• The integration of AI into medical research necessitates a thorough examination of associated ethical and practical issues.

Purpose of the Study:

• To outline the AI guidelines developed by the Australian Epilepsy Project.
• To address critical areas in medical AI, including Trust, Responsibility, and Safety.
• To prepare for an era of augmented-intelligence-based medicine through adaptable guideline development.

Main Methods:

• Development of AI guidelines for a large-scale platform focused on epilepsy.
• Utilizing prospective multimodal datasets (MRI, genetic, clinical, cognitive data) from individuals with epilepsy.
• Focusing on key literature-identified areas: Trust, Responsibility, and Safety in medical AI.

Main Results:

• The Australian Epilepsy Project has established specific AI guidelines.
• The guidelines address the critical aspects of Trust, Responsibility, and Safety in AI implementation.
• The project aims to democratize specialist epilepsy care through AI-driven clinical decision support.

Conclusions:

• AI is transforming medical practice, necessitating adaptive guidelines.
• Establishing clear guidelines for Trust, Responsibility, and Safety is crucial for responsible AI adoption in medicine.
• The Australian Epilepsy Project's guidelines provide a framework for integrating AI into epilepsy care and research.