Surgical Innovation最新文献

{"title":"Virtual Reality Training Improves Procedural Skills in Mannequin-Based Simulation in Medical Students: A Pilot Randomized Controlled Trial.","authors":"Ryan M Knobovitch, Junko Tokuno, Fabio Botelho, Howard B Fried, Tamara E Carver, Gerald M Fried","doi":"10.1177/15533506251334693","DOIUrl":"10.1177/15533506251334693","url":null,"abstract":"<p><p>ObjectivesThe goal of this study was to evaluate whether immersive virtual reality (VR) training used in conjunction with interactive online learning improved procedural skills in medical students, using chest tube insertion as a model.MethodsMedical students (n = 30) with limited or no experience with chest tube insertion were randomized into control and VR groups. All participants received access to a previously developed online module to learn the equipment and steps involved in performing chest tube insertion. The VR group received additional training using commercially available software. All participants were then asked to perform chest tube insertion on a standardized mannequin. Technical skills were assessed by surgical experts, blinded to the group allocation, using a modified Objective Structured Assessment of Technical Skill (OSATS) rating scale (11-items, each scored 1-5). Multiple-choice tests and a 5-point Likert-scale were used to assess theoretical knowledge and to rate confidence level before and after training. Data are presented as median and interquartile range.ResultsAfter training, all participants showed significant improvement in knowledge from baseline; rate of correct answers was 50% pre-training [40.0-66.7]; 80% post-training [73.0-93.3]; <i>P</i> < 0.0001). There was no statistically significant difference between the two groups in knowledge before and after training. The VR group spent <60 min in VR training and had better procedural performance (OSATS scores: controls: 39 [33-45]; VR: 46 [42.0-50]; <i>P</i> = 0.03) and higher confidence (controls: 3 [3-4]; VR: 4 [4-5]; <i>P</i> = 0.002).ConclusionsAdding VR simulation to online learning improved technical skills and confidence in medical students learning chest tube insertion.</p>","PeriodicalId":22095,"journal":{"name":"Surgical Innovation","volume":" ","pages":"364-373"},"PeriodicalIF":1.2,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12202818/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144064692","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Enabling Robotic Capsule Endoscopy Through Effective in-Body Capsule Localization.","authors":"Muzaffer Kanaan, Memduh Suveren","doi":"10.1177/15533506251334690","DOIUrl":"10.1177/15533506251334690","url":null,"abstract":"<p><p>A critical problem for wireless capsule endoscopy is to be able to correlate a received image with a location, in order to more precisely locate a pathological condition (such as a lesion, or a tumor) inside the gastrointestinal tract and guide a potential subsequent surgical intervention. We treat the problem of determining the location and orientation parameters of a wireless capsule endoscope inside the human gastrointestinal tract through magnetic techniques. Considering the emerging trend in wireless capsule endoscopy towards remotely maneuverable robotic devices (also known as robotic capsule endoscopy), solving the problem of fast, accurate localization of the capsule becomes even more critical. It is also vital that the performance of different localization algorithms be characterized effectively in a repeatable manner in a controlled environment. Towards this goal, we present a design for a novel cyber-physical system to characterize the magnetic localization performance in robotic capsule endoscopy. We present some sample results for localization performance. The results indicate that highly accurate in-body localization of the capsule (with approximately 3.1 mm localization and 2.2° orientation error) is possible.</p>","PeriodicalId":22095,"journal":{"name":"Surgical Innovation","volume":" ","pages":"353-363"},"PeriodicalIF":1.2,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144035686","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Intrastromal Suturing Technique Compared With Interrupted Corneal Suturing Technique, Loose Suture and Knot Exposure: A Comparative Rabbit Study.","authors":"Kaan Ozkan, Bahri Aydın, Ahmet Yucel Ucgul, Kemal Bayrakceken, Mehmet Cuneyt Ozmen, Rustu Fikret Akata","doi":"10.1177/15533506251328456","DOIUrl":"10.1177/15533506251328456","url":null,"abstract":"<p><p>PurposeThis study aimed to assess the effectiveness of an innovative intrastromal suturing technique in an experimental rabbit model, comparing it to standard interrupted suturing, loose suture, and suturing with knot exposure.MethodsFourteen adult male New Zealand White rabbits were included in this study. Each rabbit underwent suturing in both eyes, divided into four groups based on suturing techniques. The novel intrastromal suturing technique involved burying the entire suture material within the corneal stroma. Corneal neovascularisation (CoNV) areas were evaluated by image analysis and immune cell densities by in vivo confocal microscopy (IVCM).ResultsThe intrastromal suturing group demonstrated significantly smaller CoNV areas at both 1 week and 1 month post-suturing compared to other interventional groups, indicating effective mitigation of CoNV development and progression. Moreover, this group exhibited lower immune cell densities in the superficial stromal layer and endothelial layer, suggesting a reduced inflammatory response. Both the loose suture and the knot exposure groups exhibited significant levels of CoNV and heightened immune cell densities.ConclusionThis experimental study demonstrated effectiveness of intrastromal suturing technique in limiting CoNV and immune cell infiltration, common contributors to graft rejection and complications. Furthermore, the study revealed that loose sutures and those with exposed knots are likely to cause more severe CoNV and inflammation, compared to the traditional interrupted suturing technique and intrastromal suturing.</p>","PeriodicalId":22095,"journal":{"name":"Surgical Innovation","volume":" ","pages":"325-333"},"PeriodicalIF":1.2,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143812448","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Extended Reality Technologies for Visualization in DIEP Flap Breast Reconstruction: A Systematic Review.","authors":"Berk B Ozmen, Ali Foroutan, Joseph D Quick, Robert Siska, William Albabish, Graham S Schwarz","doi":"10.1177/15533506251328473","DOIUrl":"10.1177/15533506251328473","url":null,"abstract":"<p><p>BackgroundDeep inferior epigastric artery perforator (DIEP) flap breast reconstruction is the gold standard for autologous breast reconstruction but is complex due to variable vascular anatomy. Traditional imaging modalities like computed tomography angiography (CTA) provide two-dimensional visualization, which can limit spatial understanding. Extended reality (XR) technologies, including virtual reality (VR), augmented reality (AR), and mixed reality (MR) offer immersive three-dimensional visualization that may enhance surgical planning and intraoperative guidance. This systematic review synthesizes current evidence regarding XR applications in DIEP flap breast reconstruction.MethodsA systematic review was conducted following PRISMA 2020 guidelines. PubMed, Embase, and Scopus were searched for original research articles reporting clinical use of XR technologies in DIEP flap breast reconstruction. Data on study characteristics, XR technologies used, outcomes, and key findings were extracted and analyzed.ResultsThirteen studies involving 265 patients met inclusion criteria. Seven studies utilized AR, six used VR, and none focused on MR. XR technologies were associated with improved accuracy in perforator identification, reduced operating times, and enhanced surgical planning. Identified limitations included soft tissue deformation affecting intraoperative application, ergonomic issues with headsets, and technological constraints such as processing power and network.ConclusionXR technologies show significant potential in improving preoperative planning and intraoperative guidance in DIEP flap breast reconstruction. While preliminary results are promising, further large-scale studies are needed to establish efficacy, address limitations, and facilitate integration into clinical practice.</p>","PeriodicalId":22095,"journal":{"name":"Surgical Innovation","volume":" ","pages":"385-391"},"PeriodicalIF":1.2,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143664634","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Validating a Novel 3D Printed Depth Gauge With Mandible Models.","authors":"Eric M Smith, Rhorie P Kerr, Ashley E Kita","doi":"10.1177/15533506251334694","DOIUrl":"10.1177/15533506251334694","url":null,"abstract":"<p><p>Background/needTraditional bone depth gauges are notoriously inaccurate tools, often used in head and neck surgery, that estimate the screw length needed for fracture fixation after bicortical drilling. Complications related to inaccurately sized screws may include soft tissue irritation or weakness of the repair and subsequent refracture. To improve size selection accuracy, a prototype depth gauge was 3D printed and tested in mandibles.MethodsThe prototype was constructed with a rotating deployable hook and intra-operative disassembly feature to extract the device if it became stuck. Ten 3.2 mm holes were drilled in a synthetic mandible, and 12 medical students, 12 residents, and 6 fellows/attendings measured them with industry standard and prototype depth gauges. User measurements from the prototype were compared to the holes' true depths and accuracy for each device was based on a user's closeness to the true depths. Differences between devices and training levels were analyzed with paired t tests and two-way ANOVAs. The device was also tested by 2 attendings in 2 cadavers with 8 holes drilled in each mandible.ResultsIn the synthetic model, differences between true depths and measured depths for the 2 gauges were not significantly different. Total accuracy was greater with the prototype, along with increased medical student accuracy compared to the industry standard. Prototype malfunctions were noted in the cadaveric model with no significant differences in device accuracy.ConclusionA novel 3D-printed depth gauge was tested and found to improve first time user accuracy and perform non-inferiorly to an industry standard depth gauge.</p>","PeriodicalId":22095,"journal":{"name":"Surgical Innovation","volume":" ","pages":"334-344"},"PeriodicalIF":1.2,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12202827/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144039284","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Validation of a Novel \"Windshield Wiper\" for Laparoscopes in Cadaver and Live Porcine Models.","authors":"Christopher R Idelson, Etse-Oghena Y Campbell, Lee M Fuentes, Simin E Golestani, Jawad T Ali, Austin Fagerberg, Aaron Laviana, John M Uecker","doi":"10.1177/15533506251329635","DOIUrl":"10.1177/15533506251329635","url":null,"abstract":"<p><p>BackgroundDuring laparoscopic surgeries, laparoscopes are inserted through a trocar port into the body cavity, which is then insufflated with carbon dioxide. Laparoscope lens clarity frequently becomes compromised via condensation or smearing of blood and adipose. This problem is well-known in the field, yet a viable in vivo solution has yet to address the issue and be successfully clinically adopted.Research DesignA structured cadaveric study evaluated the cleaning performance and clinician satisfaction with a novel laparoscope lens cleaning device against 2 gold-standard lens cleaning products. The novel device was also tested in a live animal porcine model to assess cleaning performance in a warm body environment qualitatively. The validation in the porcine model did not have the same evaluation process and comparison with other lens clearing methods as the cadaveric experiment.ResultsCleaning events were timed individually and analyzed post hoc. Average times to clean scopes for the novel device, Clearify™, and Fred™ Anti-Fog solution were 5 ± 5, 16 ± 7, and 14 ± 6 seconds, respectively. In 100 cleaning events with the novel device, the laparoscope was removed from the body zero times, with an average of 2 ± 1.29 cleaning actuations per event. Clearify™ and Fred™ Anti-Fog were removed from the body 102 and 116 times, with an average number of cleaning actuations of 1.07 ± 0.26 and 1.19 ± 0.53 per event, respectively. In the live porcine model, the novel device consistently cleared all debris deposited on the lens, including fog, tissue, blood, and bile fluid.ConclusionThis study demonstrates the novel device's reduction in cleaning duration and scope removals compared to gold-standard technologies, suggesting a potential for improved workflow and reduced intra-operative interruptions.</p>","PeriodicalId":22095,"journal":{"name":"Surgical Innovation","volume":" ","pages":"345-352"},"PeriodicalIF":1.2,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143710641","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Immersive Virtual Reality Simulation for Medical Student Procedural Training: Assessment of Cognitive Load and Usability.","authors":"Junko Tokuno, Ryan M Knobovitch, Fabio Botelho, Howard B Fried, Tamara E Carver, Gerald M Fried","doi":"10.1177/15533506251339920","DOIUrl":"10.1177/15533506251339920","url":null,"abstract":"<p><p>ObjectiveIn a previous randomized controlled trial, we found immersive virtual reality (VR) simulation to be effective for teaching procedural skills to medical students. We further investigated this interface's usability and cognitive load.MethodsThis was a secondary analysis of data from a previous randomized controlled trial. Twenty-two medical students with no or limited experience with VR and chest tube insertion received training for chest tube insertion using a commercially available immersive VR simulation. Participants completed post-training surveys on usability (System Usability Scale, SUS, from 0-100) and cognitive load (Leppink's scale, 11-point, 10 items). Three types of cognitive loads (intrinsic, extraneous, and germane) were evaluated. Modified Objective Structured Assessment of Technical Skills (OSATS, 5-point, 11 items) for technical skills in a mannequin simulation were assessed after VR training, and in knowledge scores before and after training were extracted to analyze their relationships with usability and cognitive load. Data are presented as median (interquartile range).ResultsMedian scores (%) for the knowledge test were 46.7 (40.0-53.3) at baseline and 86.7 (80.0-90.3) after training. The OSATS score was 40.5 (35.5-49.3), and SUS was 82.5 (73.8-88.8, with significant correlation between these variables (r = 0.51, <i>P</i> = 0.04). The intrinsic, extrinsic, and germane cognitive loads were 3.7 (1.8-6.1), 0.15 (0-1.4), and 9.2 (6.0-10), respectively.ConclusionCognitive load and usability of immersive VR simulation were reported to be excellent. Along with its effectiveness shown previously, VR simulation is a highly acceptable approach for teaching technical skills to medical students.</p>","PeriodicalId":22095,"journal":{"name":"Surgical Innovation","volume":" ","pages":"378-384"},"PeriodicalIF":1.2,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12202820/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143999924","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Efficacy of Uterine Manipulator in Total Abdominal Hysterectomy: A Triple-Blinded Randomized Controlled Trial.","authors":"Pınar Yıldız, Esra Keles, Gazi Yıldız, Pınar Birol İlter, Kasım Turan, Mehmet Mete Kirlangic, Murat Levent Dereli, Fatih Şanlıkan, Batuhan Çağlar, Emre Mat","doi":"10.1177/15533506251360840","DOIUrl":"https://doi.org/10.1177/15533506251360840","url":null,"abstract":"<p><p>BackgroundThe objective of this study was to investigate whether the advantages of uterine manipulators in laparoscopic hysterectomy could also be applied to abdominal hysterectomy. In the present study, we introduce a novel surgical technique that employs a uterine manipulator during open surgery and analyze the outcomes of this approach.MethodsA prospective, single-center, randomized controlled trial was conducted at a tertiary research hospital with patients who underwent hysterectomy for benign gynecological indications between October 2023 and March 2024. Patients were randomly assigned to either conventional abdominal hysterectomy or manipulator-assisted abdominal hysterectomy. Primary outcomes included operative time, with secondary outcomes including Visual Analog Scale (VAS) pain scores, intraoperative and postoperative complications, and length of hospital stay.ResultsAmong the 142 participants, the manipulator group exhibited a mean operative time of 92.2 ± 11.1 minutes compared to 107.6 ± 14.4 minutes in the control group (<i>P</i> < .001). Length of hospital stay was also significantly shorter in the manipulator group (42.3 ± 14.7 hours vs 47.4 ± 15.8 hours; <i>P</i> = .046). No significant differences were observed in postoperative complications or pain scores between groups. The mean postoperative hemoglobin decrease was significantly lower in the manipulator group (0.74 ± 0.28 g/dL) compared to the control group (1.52 ± 0.33 g/dL) (<i>P</i> < .001).ConclusionThe incorporation of a uterine manipulator in abdominal hysterectomy significantly reduces operative time, colpotomy time, and hospital stay while maintaining a comparable safety profile to conventional techniques. These findings suggest that uterine manipulators may enhance surgical efficiency and could be beneficial in clinical practice.</p>","PeriodicalId":22095,"journal":{"name":"Surgical Innovation","volume":" ","pages":"15533506251360840"},"PeriodicalIF":1.6,"publicationDate":"2025-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144754357","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Appraisal of the Current Guidelines for the Management of AIN Using the Appraisal of Guidelines Research and Evaluation II (AGREE II) Instrument.","authors":"Massimiliano Mistrangelo, Serena Mantova, Alberto Arezzo, Roberto Paolo Iachetta, Andrea Lauretta, Simone Arolfo, Marta Mozzon, Mario Morino, Paola De Nardi","doi":"10.1177/15533506251362905","DOIUrl":"https://doi.org/10.1177/15533506251362905","url":null,"abstract":"<p><p>AimAnal Intraepithelial Neoplasia has been a hot topic in colorectal surgery and many Scientific Societies have produced guidelines for their diagnosis and treatment. This study aims to appraise the quality of the existing guidelines in this field.MethodsA systematic review of the Literature was conducted in PubMed, EMBASE and Scholar Google databases. Seven authors independently valued the quality of guidelines using the AGREE II instrument.ResultsThis study identified and included 9 guidelines of varying quality. The highest scoring guidelines were different considering each domain. NYSDOH ones gained the higher scoring in 2 domains. However, there was considerable variability across the studies and the various domains. The highest scoring domains were domain VI: Editorial Independence (median = 89% across all studies); IV: Clarity of Presentation (median = 61% across all studies) and domain I: Scope & Purpose (median = 59% across all studies). The lowest scores were observed in domain V: Applicability (22%) and domain III: Rigour of Development (29%). Only 2 of the 9 gained unanimous support for their use, whilst 5 guidelines were unanimously declared unsuitable for clinical use. The last 2 guidelines were considered beneficial only for limited purposes.Conclusions4 out of the 9 guidelines examined obtained moderate/good scores in various domains of the AGREE II tool. The review of the guidelines highlighted poor attention to stakeholder involvement and scarce care, mainly in the Rigour of development and Applicability of the guidelines. The AGREE II instrument could improve the drafting of new guidelines or help update published ones.</p>","PeriodicalId":22095,"journal":{"name":"Surgical Innovation","volume":" ","pages":"15533506251362905"},"PeriodicalIF":1.6,"publicationDate":"2025-07-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144733319","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Comparative Performance Analysis of AI Engines in Answering American Board of Surgery In-Training Examination Questions: A Multi-Subspecialty Evaluation.","authors":"Nawaf AlShahwan, Ibrahim Majed Fetyani, Mohammed Basem Beyari, Saleh Husam Aldeligan, Maram Basem Beyari, Rayan Saleh Alshehri, Ahmed Alburakan, Hassan Mashbari, Abdulaziz AlKanhal, Thamer Nouh","doi":"10.1177/15533506251361664","DOIUrl":"https://doi.org/10.1177/15533506251361664","url":null,"abstract":"<p><p>BackgroundThe rapid advancement of artificial intelligence (AI) has led to its increasing application in the medical field, particularly in providing accurate and reliable information for complex medical queries. PurposeThis study evaluates the performance of four AI engines-Perplexity, Chat GPT, DeepSeek, and Gemini in answering 100 multiple-choice questions derived from the American Board of Surgery In-Training Examination (ABSITE). A set of questions focused on five surgical subspecialties including colorectal surgery, acute care and trauma surgery (ACS), upper GI Surgery, breast and endocrine surgery, and hepatopancreatobiliary surgery (HPB).Data collectionWe evaluated these AI engines' ability to provide accurate and focused medical knowledge as the main objective. The research study consisting of a two-month duration was conducted from January 1, 2025, to March 28, 2025. All AI engines received identical questions through then a comparison between correct and wrong responses was performed relative to the ABSITE key answers. Each question was entered manually into the chatbots, ensuring no memory retention bias.Statistical analysisThe researchers conducted their statistical analysis with JASP software for performance evaluation between different subspecialties and AI engines through univariate and multivariate investigations.ResultsAmong the available AI tools, DeepSeek produced the most accurate responses at 74% while Chat GPT delivered 70% accuracy Gemini achieved 69% and Perplexity attained 65%. Results showed that Chat GPT achieved 83.3% accuracy in colorectal surgery yet DeepSeek scored the best at 84.6% and 67.6% for HPB Surgery and ACS respectively. Perplexity achieved a 100% accuracy rate in breast and endocrine surgery which proved to be the highest score recorded throughout the study. The analysis showed that Chat GPT exhibited performance variability between different Surgical subspecialties since it registered significant variations (<i>P</i> < .05), especially in acute care and trauma Surgery. The results of logistic regression indicated that Gemini along with Perplexity scored the most consistent answers among AI systems with a significant odds ratio of 2.5 (<i>P</i> < .01). AI engines show different combinations of precision and reliability when solving medical questions about surgery yet DeepSeek stands out by remaining the most reliable overall.ConclusionsMedical application AI models need additional development because performance results show major differences between medical specialties.</p>","PeriodicalId":22095,"journal":{"name":"Surgical Innovation","volume":" ","pages":"15533506251361664"},"PeriodicalIF":1.2,"publicationDate":"2025-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144643608","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}