Purpose:
The purpose of this article is to provide an analytical review of the applications of artificial intelligence in various branches of rehabilitation and to examine the challenges and ethical considerations associated with it in order to clarify the future development path of this technology in the service of patients with disabilities.
Methods:
In this review, relevant articles were searched in PubMed, Scopus, and Web of Science using keywords related to artificial intelligence, machine learning, rehabilitation sciences, and bioethics. The search covered publications from January 2000 to June 2025. After removal of duplicates and screening for relevance, 28 studies were selected for content analysis and narrative synthesis.
Results:
The results of this study show that this technology, relying on machine learning, natural language processing, and complex data analysis, can be effective in all areas of rehabilitation science, including motor rehabilitation, cognitive rehabilitation, treatment effect prediction, personalized treatments, remote services, and assisting patients with independence. However, challenges such as lack of native data, information heterogeneity and the need for training and clinical acceptance of therapists, high costs, and legal considerations in the use of health data are still the main obstacles to the development of this technology.
Conclusion:
This technology has great potential to improve the quality of rehabilitation services and improve treatment outcomes for patients with disabilities. However, for the effective and responsible use of this technology, it is necessary to pay attention to ethical considerations such as algorithm transparency, equity in access, privacy, and protection of the rights of people with disabilities. Developing ethical frameworks, strengthening data infrastructure, and training therapists can pave the way for the sustainable development of artificial intelligence in rehabilitation science.