Artificial Intelligence Challenges in the Healthcare Industry: A Systematic Review of Recent Evidence

While AI is essential to the development of electronic health, it has challenges that, if resolved, might improve the standard of healthcare services. The purpose of this study is to classify and identify these issues in the healthcare field. The study utilised a systematic review approach, drawing data from the Scopus, Web of Science, and PubMed databases. The search results were imported into EndNote software, and experienced experts reviewed the relevant articles. The selection criteria focused on original research articles in English, published between 2019 and July 2024, that provided full text and sufficient data on AI challenges. Forty-seven articles were included in the final analysis out of the 1453 that were identified. There were 17 categories for the obstacles, and the most common ones were technical challenges (29.8%), technological adoption (25.5%) and reliability and validity (23.4%). There are 24 categories into which the healthcare domains were divided. This article emphasises the critical importance of addressing technical challenges, enhancing reliability and validity, safeguarding patient data, and overcoming the lack of knowledge and understanding of artificial intelligence among patients and the general public to ensure the responsible and equitable implementation of AI in healthcare.