Artificial Intelligence in Health: Overview

Objective: To evaluate the use of ChatGPT and DeepSeek in clinical practice to provide healthcare professionals with accurate information on the prevention, diagnosis, and management of post-dural puncture headache (PDPH), in particular to evaluate ChatGPT-4o, ChatGPT-4o mini, DeepSeek-V3 and DeepSeek with Deep Think(R1)'s responses with consensus practice guidelines for headache after dural puncture.

Background: Post-dural puncture headache (PDPH) is a common complication of dural puncture. Currently, there is a lack of evidence-based guidance on the prevention, diagnosis and management of PDPH. The 2023 Consensus guidelines provide comprehensive information. With the development and popularization of AI, more and more people are using ai models, including patients and doctors. However, the quality of the answers provided by ai has not yet been tested.

Methods: Responses from ChatGPT-4o, ChatGPT-4o mini, DeepSeek-V3, and DeepSeek-R1 were evaluated against PDPH guidelines using four dimensions: Accuracy (guideline adherence), Overconclusiveness (unjustified recommendations), Supplementary information (additional relevant details), and Incompleteness (omission of critical guidelines). A 5-point Likert scale further assessed response accuracy and completeness.

Results: All four models show high accuracy and completeness.Of the 10 clinical guidelines evaluated,ChatGPT-4o, ChatGPT-4o mini, DeepSeek-V3 and DeepSeek-R1 all showed 100% accuracy in responses (10/10)(p = 1). None of the four models showed overly conclusive results(p = 1). In terms of supplementary information, ChatGPT-4o,ChatGPT-4o mini and DeepSeek-R1 are 100% (10/10), DeepSeek-V3 is 90% (9/10)(p = 1). In terms of incompleteness, ChatGPT-4o is 80%(8/10), DeepSeek-R1 is 70%(7/10), ChatGPT-4o mini and DeepSeek-V3 are 60% (6/10) (p = 0.729).

Conclusion: All four AI models demonstrate clinical validity, with ChatGPT-4o and DeepSeek-R1 showing stronger guideline alignment. Though largely accurate, their responses achieve only 60-80% completeness relative to medical guidelines. Healthcare professionals must exercise caution when using AI tools and should critically evaluate outputs before clinical application. While promising, their partial guideline coverage requires careful human oversight. Further validation research is essential before these models can reliably support clinical decision-making for complex conditions like PDPH.

Anesthesiology and critical care medicine contain a vast repository of patient data that can be analyzed and decoded by artificial intelligence applications. Although there has been a rapid growth in the development of predictive analytics to improve providers' decision making, the field is still in its infancy. Many studies evaluating the use of predictive analytics are subject to bias and lack external validation. Additionally, there are many technical, regulatory, and ethical concerns that need to be addressed before these novel tools are widely adopted within anesthesiology and critical care.

In the next 25 years, we will see unprecedented technological progress due to the artificial intelligence (AI) revolution. AI-driven advances will be made in five strongly interconnected domains: reenergised drug discovery, continuous real-time monitoring, personalised medicine, automation and decentralised care. Drug discovery will move from in vitro to in silico approaches. We will identify molecules that fit protein targets and protein targets that fit molecules, allowing us to better understand mechanisms of anaesthesia and pain. Improved monitoring of physiological and biological variables will lead to better assessment and prediction of haemodynamics, anaesthetic depth and nociception. Anaesthetic care will be personalised by AI-driven analysis of population, personal and biomarker data, as well as an understanding of the values and wishes of patients. Automation will streamline routine tasks in hospitals, especially in relation to documentation and service delivery. Remote monitoring and assistance systems will help patients receive more of their care at home. The introduction of AI may have unforeseen consequences, especially for equity, sustainability and the future of work. We must think deeply about the future role of anaesthesiologists and ensure that we are involved in redesigning our role. AI can simulate empathy, and will do so increasingly well, but most of us will not be deceived because it is not the human-to-human connection that we crave. Perhaps the greatest gift of AI in anaesthesia practice will be that we come to value what it means to be human.