Background And Aims: The Mayo Colonoscopy Skills Assessment Tool (MCSAT) has previously been used to describe learning curves and competency benchmarks for colonoscopy; however, these data were limited to a single training center. The newer Assessment of Competency in Endoscopy (ACE) tool is a refinement of the MCSAT tool put forth by the Training Committee of the American Society for Gastrointestinal Endoscopy, intended to include additional important quality metrics. The goal of this study is to validate the changes made by updating this tool and establish more generalizable and reliable learning curves and competency benchmarks for colonoscopy by examining a larger national cohort of trainees.
Methods: In a prospective, multicenter trial, gastroenterology fellows at all stages of training had their core cognitive and motor skills in colonoscopy assessed by staff. Evaluations occurred at set intervals of every 50 procedures throughout the 2013 to 2014 academic year. Skills were graded by using the ACE tool, which uses a 4-point grading scale defining the continuum from novice to competent. Average learning curves for each skill were established at each interval in training and competency benchmarks for each skill were established using the contrasting groups method.
Results: Ninety-three gastroenterology fellows at 10 U.S. academic institutions had 1061 colonoscopies assessed by using the ACE tool. Average scores of 3.5 were found to be inclusive of all minimal competency thresholds identified for each core skill. Cecal intubation times of less than 15 minutes and independent cecal intubation rates of 90% were also identified as additional competency thresholds during analysis. The average fellow achieved all cognitive and motor skill endpoints by 250 procedures, with >90% surpassing these thresholds by 300 procedures.
Conclusions: Nationally generalizable learning curves for colonoscopy skills in gastroenterology fellows are described. Average ACE scores of 3.5, cecal intubation rates of 90%, and intubation times less than 15 minutes are recommended as minimal competency criteria. On average, it takes 250 procedures to achieve competence in colonoscopy. The thresholds found in this multicenter cohort by using the ACE tool are nearly identical to the previously established MCSAT benchmarks and are consistent with recent gastroenterology training recommendations but far higher than current training requirements in other specialties.
Download full-text PDF |
Source |
---|---|
http://dx.doi.org/10.1016/j.gie.2015.04.041 | DOI Listing |
Nat Commun
January 2025
Lunenfeld-Tanenbaum Research Institute, Sinai Health System, Toronto, ON, Canada.
Spatial protein expression technologies can map cellular content and organization by simultaneously quantifying the expression of >40 proteins at subcellular resolution within intact tissue sections and cell lines. However, necessary image segmentation to single cells is challenging and error prone, easily confounding the interpretation of cellular phenotypes and cell clusters. To address these limitations, we present STARLING, a probabilistic machine learning model designed to quantify cell populations from spatial protein expression data while accounting for segmentation errors.
View Article and Find Full Text PDFComput Biol Med
January 2025
Department of Industrial Engineering, Izmir University of Economics, Izmir, 35330, Türkiye. Electronic address:
Background: The severity of recent Coronavirus (COVID-19) pandemics has revealed the importance of development of inoculation strategies in case of limited vaccine availability. Authorities have implemented inoculation strategies based on perceived risk factors such as age and existence of other chronic health conditions for survivability from the disease. However, various other factors can be considered for identifying the preferred inoculation strategies depending on the vaccine availability and disease spread levels.
View Article and Find Full Text PDFNeural Netw
December 2024
Department of Mechanical Engineering, National University of Singapore, 9 Engineering Drive 1, Singapore 117575, Singapore. Electronic address:
Manual annotation of ultrasound images relies on expert knowledge and requires significant time and financial resources. Semi-supervised learning (SSL) exploits large amounts of unlabeled data to improve model performance under limited labeled data. However, it faces two challenges: fusion of contextual information at multiple scales and bias of spatial information between multiple objects.
View Article and Find Full Text PDFBMC Cancer
January 2025
Department of Medical Oncology, Cancer Centre of Excellence, King Faisal Specialist Hospital and Research Centre, Riyadh, Saudi Arabia.
Background: Pancreatic ductal adenocarcinoma (PDAC) remains one of the most lethal malignancies, with limited treatment options yielding poor outcomes. This study aimed to evaluate the real-world clinical characteristics, treatment patterns, and outcomes of patients with locally advanced unresectable and de-novo metastatic PDAC in Saudi Arabia, providing regional data to compare with international benchmarks.
Methods: This is a retrospective, multicentre study involving 350 patients diagnosed with unresectable locally advanced or de-novo metastatic PDAC between January 2015 and November 2023.
BMC Neurol
January 2025
Department of Diagnostic Radiology, Kaohsiung Chang Gung Memorial Hospital and Chang Gung University College of Medicine, School of Medicine, College of Medicine, National Sun Yat-Sen University, No. 123 Ta-Pei Road, Niao-Sung Dist, Kaohsiung, 83305, Taiwan.
Background And Purpose: White matter hyperintensities in brain MRI are key indicators of various neurological conditions, and their accurate segmentation is essential for assessing disease progression. This study aims to evaluate the performance of a 3D convolutional neural network and a 3D Transformer-based model for white matter hyperintensities segmentation, focusing on their efficacy with limited datasets and similar computational resources.
Materials And Methods: We implemented a convolution-based model (3D ResNet-50 U-Net with spatial and channel squeeze & excitation) and a Transformer-based model (3D Swin Transformer with a convolutional stem).
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!