Development of Artificial Intelligence for Parathyroid Recognition During Endoscopic Thyroid Surgery.

OBJECTIVE: We aimed to establish an artificial intelligence (AI) model to identify parathyroid glands during endoscopic approaches and compare it with senior and junior surgeons' visual estimation. METHODS: A total of 1,700 images of parathyroid glands from 166 endoscopic thyroidectomy videos were labeled. Data from 20 additional full-length videos were used as an independent external cohort. The YOLO V3, Faster R-CNN, and Cascade algorithms were used for deep learning, and the optimal algorithm was selected for independent external cohort analysis. Finally, the identification rate, initial recognition time, and tracking periods of PTAIR (Artificial Intelligence model for Parathyroid gland Recognition), junior surgeons, and senior surgeons were compared. RESULTS: The Faster R-CNN algorithm showed the best balance after optimizing the hyperparameters of each algorithm and was updated as PTAIR. The precision, recall rate, and F1 score of the PTAIR were 88.7%, 92.3%, and 90.5%, respectively. In the independent external cohort, the parathyroid identification rates of PTAIR, senior surgeons, and junior surgeons were 96.9%, 87.5%, and 71.9%, respectively. In addition, PTAIR recognized parathyroid glands 3.83 s ahead of the senior surgeons (p = 0.008), with a tracking period 62.82 s longer than the senior surgeons (p = 0.006). CONCLUSIONS: PTAIR can achieve earlier identification and full-time tracing under a particular training strategy. The identification rate of PTAIR is higher than that of junior surgeons and similar to that of senior surgeons. Such systems may have utility in improving surgical outcomes and also in accelerating the education of junior surgeons. LEVEL OF EVIDENCE: 3 Laryngoscope, 132:2516-2523, 2022.

Nomogram for prognosis of patients with esophageal squamous cell cancer after minimally invasive esophagectomy established based on non-textbook outcome.

BACKGROUND: Non-textbook outcome (non-TO) represents a new prognostic evaluation index for surgical oncology. The present study aimed to develop new nomograms based on non-TO to predict the mortality and recurrence rate in patients with esophageal squamous cell cancer (ESCC) after minimally invasive esophagectomy (MIE). METHODS: The study involved a retrospective analysis of 613 ESCC patients, from the prospectively maintained database from January 2011 to December 2018. All the included ESCC patients underwent MIE, and they were randomly (1:1) assigned to the training cohort (307 patients) and the validation cohort (306 patients). Kaplan-Meier survival analysis was used to analyze the differences recorded between overall survival (OS) and disease-free survival (DFS). In the case of the training cohort, the nomograms based on non-TO were developed using Cox regression, and the performance of these nomograms was calibrated and evaluated in the validation cohort. RESULTS: Significant differences were recorded for 5-year OS and DFS between non-TO and TO groups (p < 0.05). Multivariate cox analysis revealed that non-TO, intraoperative bleeding, T stage, and N stage acted as independent risk factors that affected OS and DFS (p < 0.05). The results for multivariate regression were used to build non-TO-based nomograms to predict OS and DFS of patients with ESCC, the t-AUC curve analysis showed that the nomograms predicting OS and DFS were more accurate as compared to TNM staging, during the follow-up period in the training cohort and validation cohort. Further, the nomogram score was used to divide ESCC patients into low-, middle-, and high-risk groups and significant differences were recorded for OS and DFS between these three groups (p < 0.001). CONCLUSIONS: Non-TO was identified as an independent prognostic factor for ESCC patients. The nomograms based on non-TO could availably predict OS and DFS in ESCC patients after MIE.

Textbook outcome after minimally invasive esophagectomy is an important prognostic indicator for predicting long-term oncological outcomes with locally advanced esophageal squamous cell carcinoma.

Background: The textbook outcome (TO) emerges as a novel prognostic factor in surgical oncology. The present study aimed to evaluate the effect of TO on the risk of death and recurrence in patients with esophageal squamous cell carcinoma (ESCC) after minimally invasive esophagectomy (MIE). Methods: The study involved retrospective analysis of 528 patients with ESCC who were subjected to MIE from January 2011 to December 2017. TO included 8 parameters: complete resection; microscopically tumor-negative resection margins (R0); ≥15 lymph nodes removed and examined; no serious postoperative complications; no postoperative intervention; no re-admission to the intensive care unit (ICU); hospital stay ≤21 days; and no readmission ≤30 days. The Cox and logistic regression model were used to analyze the prognostic factors of survival and risk factors for TO. Results: Among the 528 patients with ESCC who were subjected to MIE, 53.2% reached TO. In the case of patients with locally advanced ESCC, 5-year overall survival (OS) was 51.1% (41.2-61.2%) for the TO group but 33.7% (23.7-43.7%) for the non-TO group (HR =0.644, 95% CI: 0.449-0.924, P=0.015). Similarly, 5-year disease-free survival (DFS) was 47.6% (38.0-57.2%) for the TO group but 29.1% (20.1-38.1%) for the non-TO group (HR =0.671, 95% CI: 0.479-0.940, P=0.018). In addition, 5-year recurrence-free survival (RFS) was 62.9% (53.7-72.1%) for the TO group but 39.8% (29.4-50.2%) for the non-TO group (HR =0.606, 95% CI: 0.407-0.902, P=0.012). Multivariate logistic regression analysis further showed that age, American Society of Anesthesiology (ASA) score, intraoperative blood loss, and smoking status acted as independent risk factors for TO. The results of the multivariate analysis assisted in the establishment of a nomogram for the prediction of TO occurrence. This nomogram exhibited satisfactory consistency and prediction ability [area under the receiving operator characteristic (AUROC) =0.717]. Conclusions: The present study showed that achieving of TO after MIE improves survival rate and reduce the recurrence rate in patients with locally advanced ESCC. The study further determined the independent factors associated with TO achievement and established a prediction model.