Assessing Artificial Intelligence Models ChatGPT-4 and Gemini 2.0 for Interpreting Arterial Waveform Data in Adults Undergoing Elective Surgery

What this Trial Is About

Researchers are evaluating the ability of artificial intelligence AI models, specifically ChatGPT-4 and Gemini 2.0, to interpret arterial waveform analysis data from adult patients undergoing elective surgery. The study aims to determine how accurately these AI models can detect hemodynamic abnormalities, how well their diagnostic suggestions match expert anesthesiologist assessments, and whether their treatment recommendations are clinically appropriate. This observational study focuses on improving decision-support tools in perioperative and critical care settings by enhancing the interpretation of complex hemodynamic data. Participants will undergo standard hemodynamic monitoring using the MostCare arterial waveform analysis device during elective surgery. Data collection will occur at multiple time points before anesthesia induction, after intubation but before surgery, and during key surgical events requiring hemodynamic intervention. The anonymized waveform data will be analyzed by AI models to detect abnormalities, suggest diagnoses, and recommend treatments. Experienced anesthesiologists will independently review these AI-generated results to validate accuracy and relevance. Throughout the study, participants hemodynamic data and AI analyses will be compared against expert assessments. Researchers will measure the accuracy of AI in detecting abnormalities, the agreement between AI and anesthesiologist diagnoses, and the clinical suitability of AI treatment suggestions. No additional interventions beyond routine care are involved. The studys duration is approximately six months, aiming to explore AIs potential to support clinical decision-making and improve patient safety during surgery.

Evaluation of the Success of Artificial Intelligence Models in Interpreting Arterial Waveform Analysis Data