The Use of Artificial Intelligence to Guide Decision-Making and Care in Neurosurgery: An Updated Review of Ethical Considerations

Introduction: The use of artificial intelligence (AI) in Neurosurgery, specifically through Large Language Models (LLMs) and Machine Learning (ML), is rapidly evolving, promising enhanced decision-making. However, these technologies raise significant ethical concerns. This rapid evolution of these technologies demands an ongoing and critical evaluation of the ethical implications associated with their use in neurosurgical care, ensuring that patient outcomes are optimized while ethical standards are upheld.

Methods: A comprehensive search was conducted across databases including Ovid MEDLINE, EMBASE, Cochrane, and Scopus, covering literature from 1970 to August 9, 2024. Studies were selected based on their relevance to the integration of AI into neurosurgical practice. Eighteen studies were included after excluding studies that did not provide substantial ethical discussions. Studies were grouped by focus on LLMs or ML. Three studies discussed both.

Results: For LLMs, the most commonly discussed ethical concerns were privacy and security (91.7%, 11/12), and bias, equity, fairness and discrimination (83%, 10/12). 58.3% (7/12) of studies discussed transparency and algorithmic explicability. 25% (3/12) of studies addressed concerns about accuracy, validity, and hallucinations, as well as liability and accountability. ML studies highlighted issues around generalizability (77.8%, 7/9), explicability or the "black box" problem (55%, 5/9), and data quality (33.3%, 3/9), with 22.2% (2/9) addressing privacy and security.

Conclusion : The integration of AI in neurosurgery can potentially enhance decision-making and patient outcomes but requires careful ethical scrutiny. LLMs raise critical concerns such as patient privacy, algorithmic fairness, and accountability for AI-driven decisions. ML models face issues such as generalizability, explicability, and data quality. Addressing these concerns will be crucial to safeguarding patient autonomy, promoting justice, and maintaining trust in AI-guided neurosurgical care. Additionally, Further research is needed to establish comprehensive ethical frameworks to ensure AI benefits all patient populations.