Artificial intelligence for diagnosing neoplasia on digital cholangioscopy: development and multicenter validation of a convolutional neural network model.

BACKGROUND: We aimed to develop a convolutional neural network (CNN) model for detecting neoplastic lesions during real-time digital single-operator cholangioscopy (DSOC) and to clinically validate the model through comparisons with DSOC expert and nonexpert endoscopists. METHODS: In this two-stage study, we first developed and validated CNN1. Then, we performed a multicenter diagnostic trial to compare four DSOC experts and nonexperts against an improved model (CNN2). Lesions were classified into neoplastic and non-neoplastic in accordance with Carlos Robles-Medranda (CRM) and Mendoza disaggregated criteria. The final diagnosis of neoplasia was based on histopathology and 12-month follow-up outcomes. RESULTS: In stage I, CNN2 achieved a mean average precision of 0.88, an intersection over the union value of 83.24 %, and a total loss of 0.0975. For clinical validation, a total of 170 videos from newly included patients were analyzed with the CNN2. Half of cases (50 %) had neoplastic lesions. This model achieved significant accuracy values for neoplastic diagnosis, with a 90.5 % sensitivity, 68.2 % specificity, and 74.0 % and 87.8 % positive and negative predictive values, respectively. The CNN2 model outperformed nonexpert #2 (area under the receiver operating characteristic curve [AUC]-CRM 0.657 vs. AUC-CNN2 0.794, P < 0.05; AUC-Mendoza 0.582 vs. AUC-CNN2 0.794, P < 0.05), nonexpert #4 (AUC-CRM 0.683 vs. AUC-CNN2 0.791, P < 0.05), and expert #4 (AUC-CRM 0.755 vs. AUC-CNN2 0.848, P < 0.05; AUC-Mendoza 0.753 vs. AUC-CNN2 0.848, P < 0.05). CONCLUSIONS: The proposed CNN model distinguished neoplastic bile duct lesions with good accuracy and outperformed two nonexpert and one expert endoscopist.

The Learning Curve for Peroral Endoscopic Myotomy in Latin America: A Slide to the Right?

BACKGROUND/AIMS: Peroral endoscopic myotomy (POEM) has been increasingly used for achalasia in Latin America, where Chagas disease is prevalent, and this makes POEM more challenging. The aim of this study was to determine the learning curve for POEM in Latin America. METHODS: Patients undergoing POEM in Latin America with a single operator were included from a prospective registry over 4 years. Non-linear regression and cumulative sum control chart (CUSUM) analyses were conducted for the learning curve. RESULTS: A total of 125 patients were included (52% male; mean age, 59 years), of which 80 had type II achalasia (64%), and 38 had Chagas disease (30%). The average pre-procedure and post-procedure Eckardt scores were 6.79 and 1.87, respectively. Technical success was achieved in 93.5% of patients, and clinical success was achieved in 88.8%. Adverse events occurred in 27 patients (22%) and included bleeding (4 patients), pneumothorax (4 patients), mucosal perforation (13 patients), mediastinitis (2 patients), and leakage (4 patients).
The CUSUM chart showed a median procedure time of 97 min (range, 45-196 min), which was achieved at the 61st procedure. Procedure duration progressively decreased, with the last 10 procedures under 50 min approaching a plateau (p-value <0.01). CONCLUSION: Mastering POEM in Latin America requires approximately 61 procedures for both POEM efficiency and to accomplish the procedure within 97 minutes.

How does per-oral endoscopic myotomy compare to Heller myotomy? The Latin American perspective.

Background and study aims Both Heller myotomy (HM) and per-oral endoscopic myotomy (POEM) are efficacious therapies for achalasia. The efficacy and safety of POEM vs HM in Latin America and specifically in patients with Chagas disease is unknown. Patients and methods Consecutive patients undergoing either HM or POEM for achalasia were included from nine Latin American centers in a prospective registry over 5 years. Technical success was defined as undergoing a successful myotomy. Clinical success was defined as achieving an Eckardt score < 3. Data on demographics, procedure info, Eckardt score, and adverse events (AEs) were collected. Student's t test, Chi squared, and logistic regression analyses were conducted. Results One hundred thirty-three patients were included (59 male; 44 %; mean age 47). POEM was performed in 69 patients, HM in 64 patients. A total of 35 patients had Chagas disease, 17 of 69 in the POEM group, 18 of 64 in the HM group. Both groups had significant reduction in Eckardt scores ( P  < 0.00001), but successful initial therapy was significantly higher in the POEM group compared to the HM group ( P  = 0.01304). AEs were similar in both group (17 % vs 14 %) and consisted of pneumothorax (n = 3 vs 2), bleeding requiring transfusion (n = 3 vs 2), and mediastinitis (n = 3 vs 1). Hospital stay was longer in the HM group than in the POEM group ( P  < 0.00001). In the Chagas subgroup, post-procedure Eckardt score in the POEM group was significantly reduced by 5.71 points ( P  < 0.00001) versus 1.56 points in the HM group ( P  = 0.042793). Conclusion Both HM and POEM are efficacious for achalasia, but POEM was associated with higher initial therapy success and shorter hospital stay in Latin America. In Chagas patients with achalasia, POEM was significantly more effective than HM.