Reliability and Accuracy of Generative Artificial Intelligence Tools in Providing General Information on Migraine Surgery.

Background: Numerous publicly available generative artificial intelligence tools have been introduced, aimed at research use or for the general public. The aim of this study was to investigate what information could be obtained by querying 9 of these publicly available software tools regarding the possible outcomes and complications of migraine surgery.

Methods: We consulted 9 of the most well-known and widely used generative artificial intelligence tools: ChatGPT, Gemini, Perplexity, Elicit, SciSpace, Consensus, PaperPal, Julius, and Mistral AI. Each tool was asked the same question: "Detail the outcomes and complications of migraine surgery."

Results: The results we obtained were, on the whole, very satisfactory. All the tools used, in addition to being extremely fast, provided sensible and scientifically reliable answers, albeit with varying degrees of accuracy. The accuracy of the answers provided was found to be directly proportional to the number of scientific references each tool used to generate its response. The tools that provided the most complete and detailed answers were, in order, Consensus, SciSpace, Perplexity, and ChatGPT.

Conclusions: In the answers provided by all the software, the universally accepted principles of migraine surgery were accurately listed, both in terms of outcomes and the rate and severity of complications. Regarding outcomes, the high success rate of the procedure was consistently emphasized.