Surgery最新文献

{"title":"Tabular foundation models as a new portable standard in local surgical risk prediction","authors":"Chris Varghese MBChB, BMedSc(Hons) ,&nbsp;Elizabeth Habermann PhD ,&nbsp;Kristine Hanson MPH ,&nbsp;Ashok Choudhary PhD ,&nbsp;Hojjat Salehinejad PhD ,&nbsp;Cornelius Thiels DO, MBA","doi":"10.1016/j.surg.2025.110078","DOIUrl":"10.1016/j.surg.2025.110078","url":null,"abstract":"<div><h3>Introduction</h3><div>Surgical risk is locality-specific, and database infrastructure to support accurate preoperative risk stratification is limited globally. Recently, foundation models for risk prediction trained on large corpus of synthetic data that are ready for domain-specific applications have emerged. We aimed to evaluate the role of a tabular foundation model in widening access to high-accuracy local risk stratification in emergency surgery.</div></div><div><h3>Methods</h3><div>We applied a transformer-based tabular pretrained foundation model with comparison to logistic regression and gradient boosting methods in our institutional data from the American College of Surgeons National Surgical Quality Improvement Program database of patients undergoing emergency surgery. We first compared performance overall, then at individual sites (n = 5), followed by comparison of the tabular prior-data fitted network model trained on site-level data against logistic regression and XGBoost models trained on all available multisite data not used for testing. Outcomes of interest were 30-day mortality and morbidity.</div></div><div><h3>Results</h3><div>Among 7,281 emergency surgery patients (4.8% mortality, 30.2% morbidity), tabular prior-data fitted network achieved the highest area under the receiver operating characteristic curve (0.82, 95% confidence interval 0.81–0.83, for morbidity; 0.89, 95% confidence interval 0.89–0.90, for mortality) and under the precision-recall curve (0.68, 95% confidence interval 0.66–0.7, for morbidity; 0.35, 95% confidence interval 0.29–0.39, for mortality), and best calibration (Brier score 0.15 for morbidity and 0.04 for mortality) compared with logistic regression and XGBoost models. The tabular prior-data fitted network's excellent performance persisted in smaller site-specific cohorts. A tabular prior-data fitted network model trained only on a single site's data performed comparable to logistic regression and XGBoost models trained on all available multisite data (<em>P</em> &lt; .4).</div></div><div><h3>Conclusions</h3><div>Access to high-performance surgical risk stratification can be improved through a tabular foundation model. This portable approach offers flexibility to missing data, and strong comparative performance in smaller data sets.</div></div>","PeriodicalId":22152,"journal":{"name":"Surgery","volume":"192 ","pages":"Article 110078"},"PeriodicalIF":2.7,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146158595","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The impact of frailty on postoperative outcomes of veterans with stage I non–small cell lung cancer","authors":"Nahom Seyoum MD ,&nbsp;Steven Tohmasi MD, MPHS ,&nbsp;Daniel B. Eaton Jr. MPH ,&nbsp;Tullis Liu BS ,&nbsp;Nikki E. Rossetti MD, MPHS, MSc ,&nbsp;Martin W. Schoen MD, MPH ,&nbsp;Theodore S. Thomas MD, MPHS ,&nbsp;Su-Hsin Chang PhD, SM ,&nbsp;Yan Yan MD, PhD ,&nbsp;Mayank R. Patel MD ,&nbsp;Ana A. Baumann PhD ,&nbsp;Whitney S. Brandt MD ,&nbsp;Daniel Kreisel MD ,&nbsp;Bryan F. Meyers MD, MPH ,&nbsp;Benjamin D. Kozower MD, MPH ,&nbsp;Brendan T. Heiden MD, MPHS ,&nbsp;Varun Puri MD, MSCI","doi":"10.1016/j.surg.2026.110089","DOIUrl":"10.1016/j.surg.2026.110089","url":null,"abstract":"<div><h3>Background</h3><div>Although surgical resection is the standard of care for early-stage non–small cell lung cancer, frailty influences treatment decisions. We evaluated the prognostic utility of the Veterans Affairs Frailty Index, a claims-based assessment tool, among veterans undergoing resection for stage I non–small cell lung cancer.</div></div><div><h3>Methods</h3><div>We conducted a retrospective cohort study of veterans who underwent curative-intent surgery for stage I non–small cell lung cancer in the Veterans Health Administration from 2006 to 2020. Using the Veterans Affairs Frailty Index, frailty was categorized as follows: nonfrail (≤0.1), prefrail (0.1–0.2), mildly frail (0.2–0.3), moderately frail (0.3–0.4), and severely frail (&gt;0.4). The primary outcome was overall survival, assessed using multivariable Cox regression. Secondary outcomes included major complications, prolonged hospitalization, 30-day readmission, 90-day mortality, and recurrence.</div></div><div><h3>Results</h3><div>Among 12,271 veterans, 7.7% were severely frail. Compared with nonfrail patients, severely frail patients were older (mean age 70.3 years vs 64.7 years; <em>P</em> &lt; .001) and were more likely to undergo minimally invasive surgery (57.7% vs 37.8%) and sublobar resection (41.2% vs 22.3%) (all <em>P</em> &lt; .001). Severe frailty was independently associated with higher risk of major complications (adjusted odds ratio 2.85, 95% confidence interval 2.18–3.71), prolonged hospitalization (adjusted odds ratio 2.67), 30-day readmission (adjusted odds ratio 1.76), 90-day mortality (adjusted odds ratio 2.87), and worse overall survival (adjusted hazard ratio 2.20, 95% confidence interval 1.97–2.46; all <em>P</em> &lt; .001). Recurrence was not significantly associated with frailty (adjusted hazard ratio 0.87; <em>P</em> = .410).</div></div><div><h3>Conclusion</h3><div>Frailty, as measured by the Veterans Affairs Frailty Index, independently predicts adverse postoperative and survival outcomes following resection for stage I non–small cell lung cancer. Preoperative frailty assessment may improve risk stratification and guide surgical decision making.</div></div>","PeriodicalId":22152,"journal":{"name":"Surgery","volume":"192 ","pages":"Article 110089"},"PeriodicalIF":2.7,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146182504","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Clinical outcomes of completing total pancreatectomy for isolated recurrence of pancreatic ductal adenocarcinoma in the remnant pancreas after initial pancreatectomy","authors":"Yejong Park MD, PhD ,&nbsp;Dae Wook Hwang MD, PhD ,&nbsp;Jae Hoon Lee MD, PhD ,&nbsp;Ki Byung Song MD, PhD ,&nbsp;Eunsung Jun MD, PhD ,&nbsp;Woohyung Lee MD, PhD ,&nbsp;Song Cheol Kim MD, PhD","doi":"10.1016/j.surg.2026.110088","DOIUrl":"10.1016/j.surg.2026.110088","url":null,"abstract":"<div><h3>Background</h3><div>This study evaluated the clinical features of patients with isolated recurrence of pancreatic ductal adenocarcinoma in the remnant pancreas after initial pancreatectomy. It specifically analyzed the postoperative and oncologic outcomes of patients who underwent completion total pancreatectomy to identify prognostic factors associated with 5-year survival.</div></div><div><h3>Methods</h3><div>A retrospective review was conducted on 93 patients with isolated recurrence of pancreatic ductal adenocarcinoma in the remnant pancreas after initial pancreatectomy between 2009 and 2020. Of these, 59 patients underwent completion total pancreatectomy, and 34 did not undergo completion total pancreatectomy. Clinicopathologic factors and overall survival were analyzed based on completion total pancreatectomy status.</div></div><div><h3>Results</h3><div>The 5-year overall survival after initial pancreatectomy for pancreatic ductal adenocarcinoma was 22.3% in the patients who did not undergo completion total pancreatectomy, with a median survival of 32.8 months. In contrast, patients who underwent completion total pancreatectomy exhibited a 5-year overall survival of 61.7% and a median survival of 91.2 months (<em>P</em> &lt; .001). Among patients with early recurrence, those who underwent completion total pancreatectomy exhibited a median survival of 33.5 months. In contrast, patients with late recurrence showed a median survival of 104.2 months. Early recurrence after pancreatectomy was identified as a significant poor prognostic factor for survival (hazard ratio 6.684, 95% confidence interval 2.474–18.062, <em>P</em> &lt; .001), along with tumor size &gt;4 cm and elevated carcinoembryonic antigen levels.</div></div><div><h3>Conclusion</h3><div>In the remnant pancreas, completion total pancreatectomy for isolated recurrence of pancreatic ductal adenocarcinoma may offer favorable oncologic outcomes in carefully selected patients—particularly patients with recurrence after 1 year, tumor size &lt;4 cm, and stable tumor marker levels.</div></div>","PeriodicalId":22152,"journal":{"name":"Surgery","volume":"192 ","pages":"Article 110088"},"PeriodicalIF":2.7,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146182463","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A missed opportunity to discuss treatment costs? A qualitative analysis of preoperative conversations about thyroid cancer treatment","authors":"Emily V. Crowley MS ,&nbsp;Catherine B. Jensen MD, MSc ,&nbsp;Elizabeth M. Bacon MPH ,&nbsp;Benjamin C. James MD, MS ,&nbsp;Susan C. Pitt MD, MPHS","doi":"10.1016/j.surg.2025.110075","DOIUrl":"10.1016/j.surg.2025.110075","url":null,"abstract":"","PeriodicalId":22152,"journal":{"name":"Surgery","volume":"192 ","pages":"Article 110075"},"PeriodicalIF":2.7,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146078865","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Integration of spatiotemporal features into machine learning assessment of open surgical skills","authors":"Armin Alipour MS ,&nbsp;Jeffrey Balian BS ,&nbsp;Kevin Tabibian BS ,&nbsp;Oh Jin Kwon MD ,&nbsp;Nguyen Le MS ,&nbsp;Areti Tillou MD ,&nbsp;Peyman Benharash MD","doi":"10.1016/j.surg.2025.110079","DOIUrl":"10.1016/j.surg.2025.110079","url":null,"abstract":"<div><h3>Introduction</h3><div>Accurate and objective assessment of operative skills is essential for improving training paradigms, patient safety, and quality of surgery. Advances in machine learning have facilitated automated assessment of minimally invasive and robotic operations. This study aims to develop a novel machine learning model for evaluation of open surgical proficiency.</div></div><div><h3>Methods</h3><div>This study used the AIxSuture data set. A global rating score was assigned for each video, categorizing individuals into novice (<em>n</em> = 119), intermediate (<em>n</em> = 79), and proficient (<em>n</em> = 116) classes. Hybrid convolutional neural network and long-short-term-memory networks were employed to train the video classifier model. ResNet50, an image classification model, served as a spatial feature extractor to perform nonlinear transformations. Long-short-term-memory networks selectively retained and discarded both significant and insignificant changes in frame sets that capture the subject's movements. The class-wise F1 score was measured to assess harmonic performance.</div></div><div><h3>Results</h3><div>Our assessment achieved a mean F1 score of 80.1% in determining the performance level of each subject, outperforming previous models. Additionally, the model classified performance with 90.1% accuracy for the novice group, 65.7% for the intermediate group, and 86.3% for the proficient group. Despite lower accuracy in the intermediate class, this metric outperformed other models in this group by nearly 10%. The present model classified each video into appropriate skill levels at an estimated 10.2 ± 0.4 seconds.</div></div><div><h3>Conclusions</h3><div>Our machine learning model provides a robust framework for skill assessment in open surgery. The application of machine learning in clinical practice should be considered to evaluate surgeons' skills and help improve training and outcomes.</div></div>","PeriodicalId":22152,"journal":{"name":"Surgery","volume":"192 ","pages":"Article 110079"},"PeriodicalIF":2.7,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146078866","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Retrograde intraureteral injections of indocyanine green versus ureteral stents to reduce urinary tract complications during pelvic surgery: Systematic review and meta-analysis","authors":"Patrick E. Goldhawk-White BSc ,&nbsp;Kevin P. White MD, PhD ,&nbsp;Ceana Nezhat MD ,&nbsp;Fernando Dip MD ,&nbsp;Raul J. Rosenthal MD ,&nbsp;Steven D. Wexner MD, PhD","doi":"10.1016/j.surg.2025.110047","DOIUrl":"10.1016/j.surg.2025.110047","url":null,"abstract":"<div><h3>Introduction</h3><div>Failure to successfully delineate ureters during pelvic surgery carries significant risks. The ureters are the most commonly injured retroperitoneal structures, and ureteral injuries are associated with prolonged hospital stays, higher health care costs, and increased morbidity. In this meta-analysis, we compared the retrograde ureteral injection of indocyanine green and ureteral stents to visualize ureters and prevent ureteral injury and other urinary tract complications during pelvic surgery.</div></div><div><h3>Methods</h3><div>Thorough reviews of PubMed, Embase, and Scopus were conducted to identify all studies evaluating the efficacy of retrograde ureteral indocyanine green injections and/or ureteral stents for ureteral injury prevention. Strict inclusion criteria were applied, incorporating a Modified Downs and Black Assessment score. Because no studies comparing these 2 approaches directly were identified, indirect comparisons were performed contrasting rates of successful stent or indocyanine green placement, ureteral injury, and urinary tract complications, including the Pearson χ² or Fisher exact test and odds ratios with 95% confidence intervals.</div></div><div><h3>Results</h3><div>Twenty and 17 studies were eligible for meta-analysis for stents and indocyanine green, encompassing 25,784 and 569 patients, respectively. All 1,009 ureters that were injected with indocyanine green were injected successfully versus the successful insertion of only 99.1% of the 5,202 ureteral stents for which this approach was reported (<em>P</em> = .001). Ureteral injuries and other urinary tract complications occurred in 0.18% and 2.16% of ICG patients versus 1.12% and 11.58% of stent patients (<em>P</em> = .02 and <em>P</em> &lt; .001), respectively.</div></div><div><h3>Conclusion</h3><div>Although direct comparisons remain necessary, evidence suggests that retrograde intraureteral injections of indocyanine green are at least as good at preventing ureteral injury as stents, with a much lower rate of other urinary tract complications.</div></div>","PeriodicalId":22152,"journal":{"name":"Surgery","volume":"192 ","pages":"Article 110047"},"PeriodicalIF":2.7,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146030959","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Preperitoneal enhanced-view totally extraperitoneal (PeTEP) technique in midline and lateral incisional hernia repair: Early multicenter outcomes","authors":"Joaquín M. Munoz-Rodriguez MD, PhD, FACS ,&nbsp;Laura Román García de León MD ,&nbsp;Álvaro Robin Valle De Lersundi MD, PhD ,&nbsp;Luis A. Blazquez-Hernando MD, PhD ,&nbsp;Manuel Medina Pedrique MD ,&nbsp;Celia Fidalgo Martínez MD ,&nbsp;Marcello De Luca MD ,&nbsp;José Luis Lucena de la Poza MD, PhD ,&nbsp;Miguel A. Garcia-Urena MD, PhD, FACS ,&nbsp;Javier Lopez-Monclus MD, PhD, FACS","doi":"10.1016/j.surg.2025.110041","DOIUrl":"10.1016/j.surg.2025.110041","url":null,"abstract":"<div><h3>Background</h3><div>Minimally invasive options for incisional hernia repair have expanded, yet data on preperitoneal enhanced-view totally extraperitoneal approaches for incisional hernias are limited. We aimed to evaluate safety, feasibility, and early outcomes of the preperitoneal enhanced-view totally extraperitoneal approach via cranial, caudal, and midline access in midline and lateral incisional hernias.</div></div><div><h3>Methods</h3><div>We performed a multicenter retrospective cohort study across 2 university hospitals (January 2024–June 2025). Patients with midline or lateral incisional hernias undergoing a preperitoneal enhanced-view totally extraperitoneal approach were included; primary hernias and loss-of-domain cases were excluded. Perioperative management was standardized. The primary end point was recurrence; secondary end points included surgical site occurrences, surgical site occurrences requiring procedural intervention, mesh infection, bulging, chronic pain, operative time, and length of stay.</div></div><div><h3>Results</h3><div>We analyzed 60 incisional hernia repairs (58.3% men; age 65.1 ± 12.2 years; body mass index 28.7 ± 4.2 kg/m²; obesity 35%). Defects were midline in 81.7% and lateral in 18.3%; most were European Hernia Society W2 classification. Access was cranial in 76.7%, midline 13.3%, and caudal in 10.0%. Three procedures required intraoperative conversion to enhanced-view totally extraperitoneal approach due to peritoneal fragility; no conversions to open surgery occurred. Fascial closure was achieved in all cases; a tailored preperitoneal mesh (mean area 544.6 ± 272 cm²) was placed without fixation. Mean operative time was 163 ± 66 minutes; length of stay was 1.28 ± 0.7 days. Surgical site occurrence rate was 6.6% (2 superficial hematomas, 2 asymptomatic seromas), with no surgical site infections or surgical site occurrences requiring procedural interventions. At 8.12 ± 3.92 months' follow-up, there were no recurrences, mesh infections, chronic seromas, or chronic pain.</div></div><div><h3>Conclusions</h3><div>The preperitoneal enhanced-view totally extraperitoneal approach appears to be a feasible and safe minimally invasive option for midline and lateral incisional hernia repair, enabling broad anatomic applicability with low short-term morbidity and no early recurrences. Prospective studies with long-term follow-up are needed to validate its long-term efficacy.</div></div>","PeriodicalId":22152,"journal":{"name":"Surgery","volume":"192 ","pages":"Article 110041"},"PeriodicalIF":2.7,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145941430","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Drivers and composition of hospitalization costs in patients undergoing laparoscopic tension-free hiatal hernia repair: A quantile regression study","authors":"Xiaoli Liu MD,&nbsp;Qiuyue Ma PhD,&nbsp;Haiyun Li BS,&nbsp;Minxian Zhao BS,&nbsp;Yingmo Shen MD,&nbsp;Huiqi Yang MD","doi":"10.1016/j.surg.2025.110036","DOIUrl":"10.1016/j.surg.2025.110036","url":null,"abstract":"<div><h3>Background</h3><div>Laparoscopic tension-free repair has become the standard surgical treatment for hiatal hernia because of its favorable clinical outcomes. However, the associated hospitalization costs vary considerably, driven by both patient complexity and technical factors. This study aimed to analyze the drivers and composition of hospitalization costs in patients undergoing laparoscopic tension-free hiatal hernia repair using a quantile regression approach.</div></div><div><h3>Methods</h3><div>A retrospective observational study was conducted among patients who underwent laparoscopic tension-free hiatal hernia repair at Beijing Chao-Yang Hospital, Capital Medical University, between 2020 and 2023. Clinical, demographic, and procedural data were extracted from electronic medical records. The primary outcome was total hospitalization cost; secondary outcomes included cost composition (diagnosis, treatment, nursing, medication, materials, and others) and postoperative recurrence. Quantile regression was used to identify factors associated with total hospitalization costs across the 10th–90th percentiles. Recurrence was examined using multivariable logistic regression.</div></div><div><h3>Results</h3><div>A total of 197 patients were included. Most were female (60.4%) and aged ≥65 years (54.8%). Material costs represented the largest share of total hospitalization expenses in every year (&gt;58%). In quantile regression, mesh fixation with absorbable sutures versus tackers was associated with lower costs across all quantiles (β = –11,671 to –8,372; all <em>P</em> ≤ .003). Length of stay was positively associated with costs from the 10th to the 70th quantile (β = 623–917; all <em>P</em> ≤ .032). Intensive care unit use increased costs predominantly in the lower-mid quantiles (q10–q40; β = 2,577–4,301). Postoperative recurrence occurred in 9 of 197 patients (4.6%) and had no independent predictors on multivariable analysis.</div></div><div><h3>Conclusion</h3><div>Hospitalization costs for laparoscopic tension-free hiatal hernia repair were largely driven by material expenditures. Absorbable suture fixation reduced costs across all quantiles without prolonging operative time, whereas longer length of stay increased costs, and intensive care unit use affected mainly the lower-mid range. Early recurrence was uncommon and had no independent predictors, supporting cost-conscious strategies that prioritize judicious fixation, discharge efficiency, and selective critical-care use, to be confirmed in prospective studies with standardized long-term outcomes.</div></div>","PeriodicalId":22152,"journal":{"name":"Surgery","volume":"192 ","pages":"Article 110036"},"PeriodicalIF":2.7,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145941447","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"AI-driven prediction of completion time and errors in the Advanced Training in Laparoscopic Suturing (ATLAS) needle handling task: One step closer to automated surgical skill assessment","authors":"Huu Phong Nguyen PhD ,&nbsp;Sofia Garces-Palacios MD ,&nbsp;Darian Hoagland MD ,&nbsp;Nicole Wise MD ,&nbsp;Kailen Wong BS ,&nbsp;Sai Abhinav Pydimarry BS ,&nbsp;Sharanya Vunnava BS ,&nbsp;Daniel J. Scott MD ,&nbsp;Dmitry Nepomnayshy MD ,&nbsp;Ganesh Sankaranarayanan PhD","doi":"10.1016/j.surg.2025.110045","DOIUrl":"10.1016/j.surg.2025.110045","url":null,"abstract":"<div><h3>Background</h3><div>The Advanced Training in Laparoscopic Suturing is a proficiency-based curriculum of 6 structured tasks. In the needle handling task, participants maneuver a needle through 6 standardized holes on a circular platform. Performance (completion time and errors) is currently evaluated in person or through manual video review. This study explored the potential of artificial intelligence models to automate the assessment of this task by predicting task duration and detecting needle drop errors.</div></div><div><h3>Methods</h3><div>A retrospective review was conducted of Advanced Training in Laparoscopic Suturing needle handling task videos collected from 2 tertiary centers. Two complementary artificial intelligence models were developed. First, videos were annotated across 10 distinct phases. A deep expandable three-dimensional convolutional network combined with hybrid adaptive k-nearest neighbors and smoothed moving average and exponential moving average was trained for phase segmentation and duration prediction. Second, a vision transformer model was trained to detect needle drop errors by classifying frame segments based on needle visibility.</div></div><div><h3>Results</h3><div>Phase segmentation accuracy improved from 82.06% ± 0.84% to 89.67% ± 1.27%, with the highest accuracy reaching 90.56% and an F1-score of 86.90% using the hybrid k-nearest neighbors and smoothed moving average model. The predicted task duration error had a mean error of 0.84%. The vision transformer model achieved a 95.16% classification accuracy on validation frames and detected 66.6% of needle drops &gt;2 seconds and 63.6% of needle drops &gt;5 seconds in test videos.</div></div><div><h3>Conclusion</h3><div>Artificial intelligence–based models exhibited high and moderate accuracy for task duration prediction and needle drop error, respectively, offering scalable solutions for objective surgical assessments.</div></div>","PeriodicalId":22152,"journal":{"name":"Surgery","volume":"192 ","pages":"Article 110045"},"PeriodicalIF":2.7,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145980613","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Factors associated with the feasibility and margin quality of sublobar resection for peripheral small-sized non–small cell lung cancer","authors":"Shusheng Zhu MM ,&nbsp;Zhihua Li MM ,&nbsp;Wenzheng Xu MD ,&nbsp;Zhicheng He MD ,&nbsp;Liang Chen MD ,&nbsp;Weibing Wu MD","doi":"10.1016/j.surg.2025.110080","DOIUrl":"10.1016/j.surg.2025.110080","url":null,"abstract":"<div><h3>Objective</h3><div>This study aimed to identify factors associated with the feasibility and margin quality of sublobar resection for peripheral small-sized non–small cell lung cancer (NSCLC).</div></div><div><h3>Methods</h3><div>Patients with peripheral small-sized non–small cell lung cancer (≤2 cm) who underwent sublobar resection or lobectomy between 2020 and 2023 were screened. Surgical procedures were determined through discussion under the guidance of 3-dimensional computed tomography bronchography and angiography. A surgical margin equal to or larger than the maximum tumor diameter was considered sufficient. Logistic regression analyses were used to screen factors associated with the feasibility and margin quality of sublobar resection. Predictive nomograms were developed for segmentectomy and wedge based on independent factors, respectively.</div></div><div><h3>Results</h3><div>There were 383 (51.3%), 286 (38.3%), and 78 patients (10.4%) undergoing wedge resection, segmentectomy, and lobectomy, respectively. The likelihood of achieving sufficient margins was 90.9% for segmentectomy and 74.4% for wedge resection. Overall, 545 of 747 patients (73.0%) received sublobar resection with sufficient margins. The multivariable logistic regression analysis (segmentectomy with insufficient margins or lobectomy = 1, segmentectomy with sufficient margins = 0) showed that tumor size, radiologic types, located lobes, tumor-to-segmental bronchus distance, and subsegmental attribution were significantly associated with the feasibility and margin quality of segmentectomy. For wedge resection, tumor size and Lewis types were significantly associated with its feasibility or margin quality. The predictive models exhibited good performance, with an area under the curve of 0.821 for segmentectomy and 0.765 for wedge resection.</div></div><div><h3>Conclusion</h3><div>Tumor size, radiologic types, located lobe, tumor-to-segmental bronchus distance, Lewis types, and subsegmental attribution types were associated with the feasibility and margin quality of sublobar resection.</div></div>","PeriodicalId":22152,"journal":{"name":"Surgery","volume":"192 ","pages":"Article 110080"},"PeriodicalIF":2.7,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146047059","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}