16 April 2022 in Journal of the Canadian Association of Gastroenterology
Adequate bowel preparation is integral to effective colonoscopy. Inadequate bowel preparation has been associated with reduced adenoma detection rate and increased post-colonoscopy colorectal cancer (PCCRC). As a result, the USMSTF recommends early interval reevaluation for colonoscopies with inadequate bowel preparation. However, bowel preparation documentation is highly variable with subjective interpretation. In this study, we developed deep convolutional neural networks (DCNN) to objectively ascertain bowel preparation. Bowel preparation scores were assigned using the Boston Bowel Preparation Scale (BBPS). Bowel preparation adequacy and inadequacy were defined as BBPS ≥2 and BBPS <2, respectively. A total of 38523 images were extracted from 28 colonoscopy videos and split into 26966 images for training, 7704 for validation, and 3853 for testing. Two DCNNs were created using a Densenet-169 backbone in PyTorch library evaluating BBPS score and bowel preparation adequacy. We used Adam optimiser with an initial learning rate of 3 × 10−4 and a scheduler to decay the learning rate of each parameter group by 0.1 every 7 epochs along with focal loss as our criterion for both classifiers. The overall accuracy for BBPS subclassification and determination of adequacy was 91% and 98%, respectively. The accuracy for BBPS 0, BBPS 1, BBPS 2, and BBPS 3 was 84%, 91%, 85%, and 96%, respectively. We developed DCCNs capable of assessing bowel preparation adequacy and scoring with a high degree of accuracy. However, this algorithm will require further research to assess its efficacy in real-time colonoscopy.
3 June 2022 in Expert Review of Gastroenterology & Hepatology
Artificial intelligence has been rapidly deployed in gastroenterology and endoscopy. The acceleration of deep convolutional neural networks along with hardware development has allowed implementatio...
11 November 2024 in Endoscopy International Open
Background and study aims Competent endoscopic ultrasound (EUS) performance requires a combination of technical, cognitive, and non-technical skills. Direct observation assessment tools can be employed to enhance learning and ascertain clinical competence; however, there is a need to systematically evaluate validity evidence supporting their use. We aimed to evaluate the validity evidence of competency assessment tools for EUS and examine their educational utility. Methods We systematically searched five databases and gray literature for studies investigating EUS competency assessment tools from inception to May 2023. Data on validity evidence across five domains (content, response process, internal structure, relations to other variables, and consequences) were extracted and graded (maximum score 15). We evaluated educational utility using the Accreditation Council for Graduate Medical Education framework and methodological quality using the Medical Education Research Quality Instrument (MERSQI). Results From 2081 records, we identified five EUS assessment tools from 10 studies. All tools are formative assessments intended to guide learning, with four employed in clinical settings. Validity evidence scores ranged from 3 to 12. The EUS and ERCP Skills Assessment Tool (TEESAT), Global Assessment of Performance and Skills in EUS (GAPS-EUS), and the EUS Assessment Tool (EUSAT) had the strongest validity evidence with scores of 12, 10, and 10, respectively. Overall educational utility was high given ease of tool use. MERSQI scores ranged from 9.5 to 12 (maximum score 13.5). Conclusions The TEESAT, GAPS-EUS, and EUSAT demonstrate strong validity evidence for formative assessment of EUS and are easily implemented in educational settings to monitor progress and support learning.
31 July 2023 in International Journal of Gastrointestinal Intervention
Daniel J. Low, Abraham Yu, Cynthia Liu, Faisal Shaukat Ali, Ahmad Al Nakshabandi, Emmanuel Coronel, Phillip Ge, and Jeffrey H. Lee. Int J Gastrointest Interv 2023;12:115-22. https://doi.org/10.18528/ijgii230006
26 June 2025 in Journal of the Canadian Association of Gastroenterology
Introduction: Patients with surgically altered gastrointestinal anatomy undergoing endoscopic retrograde cholangiopancreatography (ERCP) pose challenges due to anatomical distortions. Various patient and endoscopic factors, such as sex and positioning, may impact procedural success. It is unclear how these factors may impact the technical success of ERCP among patients with altered anatomy. Objective: We aimed to determine the patient and endoscopic factors that were associated with technical success of ERCP. Methods: We conducted a retrospective single-centre study using data from 2010 to 2020 that included patients with hepaticojejunostomy, Roux-en-Y anastomosis, Billroth-1, or Billroth-2 anatomy at a single tertiary care centre in Toronto, Canada. We extracted data from a database. The primary outcome was technical success of the ERCP, defined as successful navigation to the papilla or surgical anastomosis, selective cannulation and cholangiography or pancreatography. Penalized logistic regression with elastic net regularization was used to identify significant predictors of technical success. Effect size was odds ratio with 95% confidence interval. The model was evaluated using the area under the curve (AUC) metric. Results: Overall, there were 205 patients included in the analysis. In the multivariate analysis, the most significant contributors to predicting technical success of ERCP were expert endoscopic experience and non-Roux-en Y anatomy. The elastic net model demonstrated moderate predictive performance, with an AUC of 0.656. Conclusions: The findings emphasize the importance of tailored procedural planning to optimize ERCP success in patients with altered anatomy.
14 April 2023 in Endoscopy
Aims Cancer patients with large volume hepatic tumor burden (LVHTB), hyperbilirubinemia and bile duct obstruction are frequently referred for ERCP with uncertain benefit. In most cases a bilirubin of < 2mg/dL is needed for systemic chemotherapy. The aim of this study was to evaluate the outcomes of patients with LVHTB referred for ERCP. Methods A retrospective review was performed on all patients with LVHTB referred for ERCP from 7/2021-9/2022. LVHTB was defined as estimated >50% liver involvement as assessed by cross-sectional imaging. Appropriate candidates for ERCP were selected based on the presence of dilated bile ducts within the remnant liver amenable to endoscopic drainage. The primary outcomes included technical success (successful biliary stent placement), clinical success (bilirubin normalization <2mg/dL) and subsequent treatment eligibility. Results Of 170 patients with LVHTB referred, 81 (47.6%) underwent ERCP and 89 (52.4%) were declined. Technical success was 93.8% (76/81). Clinical success was 27.2% (22/81). Percutaneous transhepatic biliary drainage was subsequently performed in 14 with a clinical success of 14.3% (2/14). Only 21.0% (17/81) received subsequent chemotherapy treatment. The remainder went to hospice 72.8% (59/81) or did not follow-up 6.2% (5/81). To date, 76.5% (62/81) passed away within a median of 32days (range 8-383) after ERCP ([Table 1]). Conclusions Based on our experience, less than half of referrals with LVHTB are candidates for ERCP. Despite high technical success, only one in four patients will have adequate response to allow treatment. This information should prove useful to setting expectations of patients and referring providers. Article published online: 14 April 2023 © 2023. European Society of Gastrointestinal Endoscopy. All rights reserved. Georg Thieme Verlag KG Rüdigerstraße 14, 70469 Stuttgart, Germany