Generative Artificial Intelligence (GenAI) Adoption in Healthcare: A Systematic Scoping Review

Abstract

Importance: Generative artificial intelligence (AI) has revolutionized healthcare domains by offering solutions to complex issues, better disease detection and treatment, and patient care experience and related activities. However, despite its immense promises, the technology exhibits some problems that compromise the exploitation of its full potential.

Objectives: To provide a comprehensive overview of how generative artificial intelligence technologies have been implemented in healthcare         

Evidence Review: PubMed, Web of Science, and Google Scholar were the electronic databases searched for studies published between 2022 and 2024 discussing GenAI in healthcare. An array of MeSH terms and keywords was used to query the databases and retrieve studies for inclusion based on eligibility criteria.

Findings: Of the 2178 studies retrieved from the search process, 67 were considered for final inclusion. Clinically, the GenAI application cuts across radiology, cardiology, cancer screening, and management of comorbidities. Some notable benefits of GenAI healthcare adoption include healthcare decision support, personalized care delivery, enhanced disease detection and management through improved diagnosis and imaging, advanced medical research support, and support for administrative (documentation) tasks. However, the main pitfalls included model hallucination, data privacy and security, and information accuracy and accountability (bias).

Conclusions and Relevance: Our scoping review identifies various applications of GenAI, with a comprehensive overview of promises or benefits and possible issues compromising the efficacy of the models. This article contributes to Gen AI literature in healthcare, emphasizing the need to address the mentioned pitfalls to ease the full utilization of generative AI potentials in health delivery.