Factors Determining the Inter-observer Variability and Diagnostic Accuracy of High-resolution Manometry for Esophageal Motility Disorders

Diagnostic differences in high-resolution esophageal motility in a large Mexican cohort based on geographic distribution

High-resolution esophageal manometry [HRM] has become the gold standard for the evaluation of esophageal motility disorders. It is unclear whether there are HRM differences in diagnostic outcome based on regional or geographic distribution. The diagnostic outcome of HRM in a diverse geographical population of Mexico was compared and determined if there is variability in diagnostic results among referral centers. Consecutive patients referred for HRM during 2016-2020 were included. Four major referral centers in Mexico participated in the study: northeastern, southeastern, and central (Mexico City, two centers). All studies were interpreted by experienced investigators using Chicago Classification 3 and the same technology. A total of 2293 consecutive patients were included. More abnormal studies were found in the center (61.3%) versus south (45.8%) or north (45.2%) P < 0.001. Higher prevalence of achalasia was noted in the south (21.5%) versus center (12.4%) versus north (9.5%) P < 0.001. Hypercontractile disorders were more common in the north (11.0%) versus the south (5.2%) or the center (3.6%) P.001. A higher frequency of weak peristalsis occurred in the center (76.8%) versus the north (74.2%) or the south (69.2%) P < 0.033. Gastroesophageal junction obstruction was diagnosed in (7.2%) in the center versus the (5.3%) in the north and (4.2%) in the south p.141 (ns). This is the first study to address the diagnostic outcome of HRM in diverse geographical regions of Mexico. We identified several significant diagnostic differences across geographical centers. Our study provides the basis for further analysis of the causes contributing to these differences.

Gemini-Assisted Deep Learning Classification Model for Automated Diagnosis of High-Resolution Esophageal Manometry Images

Background/Objectives: To develop a deep learning model for esophageal motility disorder diagnosis using high-resolution manometry images with the aid of Gemini. Methods: Gemini assisted in developing this model by aiding in code writing, preprocessing, model optimization, and troubleshooting. Results: The model demonstrated an overall precision of 0.89 on the testing set, with an accuracy of 0.88, a recall of 0.88, and an F1-score of 0.885. It presented better results for multiple categories, particularly in the panesophageal pressurization category, with precision = 0.99 and recall = 0.99, yielding a balanced F1-score of 0.99. Conclusions: This study demonstrates the potential of artificial intelligence, particularly Gemini, in aiding the creation of robust deep learning models for medical image analysis, solving not just simple binary classification problems but more complex, multi-class image classification tasks.

Prevalence of Different Types of Primary Esophageal Motility Disorders and Their Associated Factors in Patients Referring to Shariati Hospital during 2018-2019

BACKGROUND: Esophageal motility disorders (EMDs) are common in patients with dysphagia and are effectively diagnosed with high-resolution manometry (HREM). In this study, we aimed to evaluate the prevalence of different types of primary EMDs in patients referred for HREM and to further investigate the factors associated with EMDs. METHODS: In this cross-sectional study, all patients referred to the endoscopy section of Shariati Hospital during 2018-2019 (279 patients) were subjected to HREM and were evaluated according to their diagnosis, and the effect of each factor and each symptom on motility disorders was investigated. RESULTS: 84.5% (235) of the participants were diagnosed with at least one esophageal motility disorder; of them, achalasia was the most common form (52.6%). None of the predictive factors showed a statistically significant correlation with EMDs. However, regarding the symptoms, regurgitation and nocturnal cough were significantly more common in patients with EMD (P=0.001 and 0.009, respectively). CONCLUSION: This study demonstrates the high prevalence of EMDs in patients undergoing manometry. None of the factors studied, such as age, sex, diabetes, hypothyroidism, smoking, and alcohol and opium consumption, had a statistically significant correlation with EMDs.

Esophageal Manometry Competency Program Improves Gastroenterology Fellow Performance in Motility Interpretation

INTRODUCTION

Competency-based medical education (CBME) for interpretation of esophageal manometry is lacking; therefore, motility experts and instructional designers developed the esophageal manometry competency (EMC) program: a personalized, adaptive learning program for interpretation of esophageal manometry. The aim of this study was to implement EMC among Gastroenterology (GI) trainees and assess the impact of EMC on competency in manometry interpretation.

METHODS

GI fellows across 14 fellowship programs were invited to complete EMC from February 2018 to October 2018. EMC includes an introductory video, baseline assessment of manometry interpretation, individualized learning pathways, and final assessment of manometry interpretation. The primary outcome was competency for interpretation in 7 individual skill sets.

RESULTS

Forty-four GI trainees completed EMC. Participants completed 30 cases, each including 7 skill sets. At baseline, 4 (9%) participants achieved competency for all 7 skills compared with 24 (55%) at final assessment (P < 0.001). Competency in individual skills increased from a median of 4 skills at baseline to 7 at final assessment (P < 0.001). The greatest increase in skill competency was for diagnosis (Baseline: 11% vs Final: 68%; P < 0.001). Accuracy improved for distinguishing between 5 diagnostic groups and was highest for the Outflow obstructive motility disorder (Baseline: 49% vs Final: 76%; P < 0.001) and Normal motor function (50% vs 80%; P < 0.001).

DISCUSSION

This prospective multicenter implementation study highlights that an adaptive web-based training platform is an effective tool to promote CBME. EMC completion was associated with significant improvement in identifying clinically relevant diagnoses, providing a model for integrating CBME into subspecialized areas of training.