A study established a deep convolutional neural network (DCNN) system that significantly improved diagnostic accuracy for early gastric cancer (EGC). In independent tests, the model achieved an area under the curve (AUC) of 0.917 and sensitivity of 93.38% for image datasets, while video tests showed an AUC of 0.930 and 96.92% sensitivity. Novice endoscopists reached […]
Category: Digital Surgery and Telemedicine
AI Models Accurately Predict Lymph Node Metastasis in Colorectal Cancer
Artificial intelligence models demonstrated significant predictive capability for lymph node metastasis (LNM) in patients with T1 colorectal cancer, outperforming guideline-recommended metrics. Among 1,386 analyzed patients, 12.5% exhibited LNM, with AUROC values of models including regularized logistic regression classifier (0.673) and catboost classifier (0.679) surpassing the expected 0.525. These findings suggest AI integration could enhance surgical […]
AI and robotics enhance surgical precision and outcomes.
AI and robotic technologies are transforming surgical practices by enhancing precision, reducing errors, and improving patient outcomes. The systematic review synthesizes recent studies across various specialties, showcasing breakthroughs in imaging, data analysis, and automated instruments. Results indicate significant advancements in decision-making and personalized treatment strategies. However, challenges such as costs, ethical issues, and training requirements […]
AI Model Enhances Prediction of Perineural Invasion in Pancreatic Cancer
An advanced AI model utilizing computed tomography can predict extrapancreatic perineural invasion (epni) in pancreatic ductal adenocarcinoma (PDAC) patients. Analyzing data from 1,065 patients, the model achieved an area under the curve of 0.87 to 0.83 across different validation sets. Furthermore, survival analysis confirmed significantly better outcomes for patients predicted to be epni-negative. With its […]
AI coaching enhances performance in laparoscopic cholecystectomy
An AI-assisted coaching program significantly improved novice surgeons’ performance in laparoscopic cholecystectomy, raising evaluation scores from 31 to 40 (p=0.008) and surpassing self-learning group scores by the study’s end (40 vs. 38, p=0.032). The coaching group also achieved a notable increase in the critical view of safety, from 11% to 78% (p=0.021). Participants expressed high […]
Automated Deep Learning Framework Enhances Pelvimetry Accuracy
A novel automated pelvimetry approach using deep learning shows promise in analyzing pelvic dimensions from MRI for total mesorectal excision (TME) difficulty assessment. The framework employs a 3D U-Net to precisely localize anatomical landmarks, achieving a mean localization error of 5.6 mm and demonstrating strong correlations (0.7 to 0.87) between automated and manual measurements. This […]
Predictive model enhances early diagnosis of gangrenous cholecystitis
A novel machine learning model has been developed to predict gangrenous cholecystitis, utilizing clinical data from 1006 patients. Employing the XGBoost algorithm and SHAP for interpretability, key clinical features were identified, including white blood cell count and d-dimer levels. This explainable model aims to assist clinicians in making timely surgical decisions, thereby potentially improving patient […]
Predictive Models Enhance Surgical Decision-Making in Hepatic Echinococcosis
A study developed predictive models to aid surgical decision-making for hepatic cystic echinococcosis (CE). Analyzing data from 406 patients, the Cox regression model achieved a concordance index of 0.94 and AUC of 0.96. The decision tree model highlighted imaging findings, cyst stage, and symptoms as key predictors, with a mean AUC of 0.950. These validated […]
Machine learning effectively predicts bile leaks post-hepatectomy.
A machine learning model using extreme gradient boosting (XGBoost) was developed to predict bile leaks after hepatectomy for liver cancer. The model analyzed data from over 22,000 patients, achieving areas under receiver operating characteristic curves (AUROC) of 0.748, 0.719, and 0.711 in various patient cohorts. Factors such as serum alkaline phosphatase, surgical approach, and cancer […]
AI enhances anatomical recognition in hernia surgery
An artificial intelligence system, Eureka, was validated for its ability to accurately identify dissection layers, nerves, vas deferens, and microvessels during transabdominal preperitoneal inguinal hernia repair. Key metrics included mean intersection over union scores of 0.33 for connective tissue, 0.24 for nerves, 0.50 for vas deferens, and 0.30 for microvessels. The results indicated improved visualization […]