A multicenter study developed a machine learning model to predict early recurrence in patients with perihilar cholangiocarcinoma (PCCA) after curative surgery. The model, leveraging five key factors, including carbohydrate antigen 19-9 and tumor size, achieved superior predictive performance, particularly with the random forest algorithm (AUC: 0.983) compared to others. High-risk patients showed significantly different recurrence-free […]
Category: Digital Surgery and Telemedicine
Machine learning predicts early recurrence after liver surgery.
A new machine learning model successfully predicted early extrahepatic recurrence (EEHR) following curative resection of colorectal liver metastases (CRLM). Among 1,410 patients, 131 (9.3%) experienced EEHR, with median overall survival significantly lower for affected patients (35.4 months) versus those without (120.5 months, p < 0.001). The model achieved a c-index of 0.77, highlighting primary tumor […]
Telemedicine is favored by older cancer patients for consultations
A significant majority (77%) of older cancer patients opted for telemedicine consultations over in-person visits at the Cancer and Aging Interdisciplinary Team clinic. Factors influencing in-person visit choices included older age, lower educational status, living in New York City, cognitive impairments, performance measure challenges, and social support issues. The findings highlight the potential of telemedicine […]
New machine learning model accurately predicts cancer treatment response
A novel machine learning model, utilizing systemic inflammation-nutritional index (SINI), successfully predicts pathological complete response (PCR) in patients with locally advanced rectal cancer undergoing neoadjuvant chemoradiotherapy (NCRT). The model achieved a mean area under the curve (AUC) of 0.877 during training and demonstrated consistent performance in both internal (AUC 0.86) and external validation sets (AUC […]
Risk-specific training enhances surgical complication prediction
Utilizing deep learning models on separately defined risk cohorts significantly improved predictive accuracy for postoperative complications. Training on high-risk patients yielded better area under the precision-recall curve for predicting in-hospital mortality, acute kidney injury, and prolonged ICU stays. This tailored approach notably enhanced F1 scores for various complications, suggesting that risk-specific training could address class […]
Machine learning enhances assessment of colorectal liver metastases response
A machine learning model utilizing CT-based radiomics significantly surpassed traditional radiologist assessments in estimating pathologic response of colorectal liver metastases after neoadjuvant therapy. In a study involving 85 patients, the model achieved an area under the curve (AUC) of 0.87, contrasting sharply with the AUCs of 0.53 for RECIST assessments and 0.56 for morphologic evaluation. […]
AI model predicts futility of surgery in cholangiocarcinoma patients
An artificial intelligence model effectively predicted “futile” surgeries in intrahepatic cholangiocarcinoma patients, utilizing ten preoperative factors. In a study of 827 cases, 378 (45.7%) experienced insufficient surgical outcomes, with notable rates of recurrence (78.6%) and mortality (21.4%) within 12 months. The model achieved high accuracy, with a training area under the curve of 0.830 and […]
Artificial intelligence is transforming surgical training.
Advances in artificial intelligence are significantly enhancing surgical education. Researchers highlight its potential for creating realistic simulations, offering personalized feedback, and improving diagnostic capabilities. While AI-assisted surgeries promise greater precision and minimally invasive techniques, challenges such as data security, resistance to adaptation, and ethical issues must be addressed. Additionally, integrating AI into curricula and ensuring […]
Machine learning models outperform traditional risk scores in bariatric surgery
Advanced machine learning techniques, particularly random forest, demonstrated superior performance in predicting 30-day complications in metabolic bariatric surgery compared to the established MBSAQIP risk score. The random forest model achieved an area under the receiver operating characteristic curve (AUROC) of 0.94 in training and 0.88 in validation. Key predictive features included serum alkaline phosphatase, platelet […]
Key Research Areas Identified for Data-Driven Surgery
An expert consensus has highlighted critical research areas in multimodal data-driven surgery to enhance data management in minimally invasive procedures. The study reveals priorities such as digitizing operating room activities, enhancing data streaming technologies, and establishing uniform protocols for multimodal data handling. Additionally, the integration of AI for operational efficiency and patient safety was emphasized. […]