Artificial Intelligence–Generated Patient Information Videos for Partial Nephrectomy by Medical Students With Expert Validation: A Pilot Study

Abstract

The role of artificial intelligence (AI) in the medical field has been expanding exponentially and has been met with excitement and uncertainty. Although there has been much research performed in AI’s role in objective parameters that affect surgical planning, there is a paucity of research that investigates its utility of enhancing intangible but equally important aspects of patient care, such as patient education. There is strong evidence supporting the use of patient information videos (PIV) preoperatively to improve anxiety in patients. Optimal preoperative counseling resulted in decreased anxiety, as well as decreased rehabilitation requirements. We sought to validate an AI-generated PIV (AIGPIV) for partial nephrectomy (PN) to augment the preoperative counseling process. MATERIAL AND METHODS Commercially available AI software (ChatGPT, MidJourney, ElevenLabs, D-ID) were used to create the PIV in a process first described by Mr Kenneth Leung (https://tinyurl.com/kennethleunggenai). Prompts were entered into ChatGPT to outline the scope of discussion of the PIV to encompass various aspects of preoperative counseling. The unedited script from ChatGPT was then converted to an audio file and subsequently to a video using other platforms mentioned above. The PIV was validated with the DISCERN Questionnaire by urologists to ensure the accuracy of the information presented in the Video. The DISCERN questionnaire is a validated questionnaire, consisting of 3 sections to objectively evaluate quality of information. RESULTS Fifteen expert urologists were involved in the validation of the AIGPIV. The mean DISCERN questionnaire score was 53.1 (±8.3) or 70.8% of the maximum score. This places the PIV squarely in the “Good” category of the DISCERN questionnaire. Overall, 80% (n = 12) and 73% (n = 11) of the urologists agreed that the AIGPIV would be informative for patients and useful for counseling a patient for PN, respectively. CONCLUSIONS This pilot study has validated the information generated by AI regarding PN. Although the information was the video assessed to be accurate, improvements could be made for to be more engaging and interactive. The AIGPIV creation may be performed at an exceptionally resourceful rate with minimal costs and manpower requirements. In addition, this can be repeated for various types of procedures for potential mass adoption. Further research is anticipated in assessing the effect of these AIGPIVs in the perioperative context as the utility of AI in healthcare continues to develop. STATEMENT OF ETHICS The study protocol was approved by the National Healthcare Group Domain Specific Review Board. All subjects were issued a participant information sheet, and written informed consent was taken before commencement of the study. CONFLICT OF INTEREST STATEMENT The authors declare that they have no conflict of interest. FUNDING SOURCES This manuscript did not receive any funding. {"href":"Single Video Player","role":"media-player-id","content-type":"play-in-place","position":"float","orientation":"portrait","label":"Video","caption":"AI-generated PIV.","object-id":[{"pub-id-type":"doi","id":""},{"pub-id-type":"other","content-type":"media-stream-id","id":"1_tpbvob09"},{"pub-id-type":"other","content-type":"media-source","id":"Kaltura"}]}

Ähnliche Arbeiten