Artificial intelligence (AI) and machine learning may play a future role in syncope management, but technical challenges as well as medicolegal and ethical concerns remain, according to a recent JACC: Advances multidisciplinary collaborative statement published May 10 in JACC: Advances.

In their State-of-the-Art Review, Giselle M. Statz, MD, et al., report that opportunities exist for AI to assist clinicians in managing syncope, a complex medical condition where adverse cardiovascular conditions, although rare, may occur. They propose that through machine learning, AI could help “define the transient loss of consciousness event, diagnose the cause, assess short- and long-term risks, predict recurrence, and determine need for hospitalization and therapeutic intervention.”

The authors note that creating clinical decision support tools, developed through machine learning models, may lead to improved patient outcomes, streamlined diagnostics and reduced health care costs; however, they warn that many retrospective and administrative databases currently available are insufficient for machine learning.

Making prospective, multicenter and multinational datasets incorporating cases from emergency departments, ambulatory units and syncope units would be ideal, as accurate machine learning models require datasets with clear “features” and “classification labels” from which to learn.

Despite its potential, the use of AI also poses several medicolegal and ethical concerns, where AI’s lack of emotional basis and ethics are both limiting. While authors emphasize the impracticality of using AI-derived algorithms and platforms alone for clinical care decisions, they foresee collaboration between treating clinicians and AI experts as an “optimal approach to patient care.”

“The development of sizeable, high-quality datasets and clinically relevant [machine learning] models will require collaborative partnerships among clinicians, data scientists, medicolegal experts and leaders in the field,” the authors state. “Such collaboration should foster a reality where AI will complement rather than compete with the current state-of-the-art in syncope management.”

Keywords: Intelligence, Syncope, Hospitalization, Algorithms, Prospective Studies, Consciousness, Retrospective Studies, Artificial Intelligence

< Back to Listings