Artificial Intelligence Driven Diagnosis of Motility Patterns in High-Resolution Esophageal Manometry: A Multicentric Multidevice Study.

How to Effectively Use High-Resolution Esophageal Manometry

Introduction: NCCP is defined as recurring, angina-like, retrosternal chest pain of non-cardiac origin. It has been estimated that up to 30% of the non-GERD related NCCP patients demonstrate an esophageal motor disorder using conventional manometry (CM). Thus, the aim of the study was to determine if high resolution esophageal manometry (HREM), which has become the standard of care, improved the diagnosis of esophageal motor disorders, as compared with CM, in patients with NCCP. Methods: We evaluated 300 consecutive non-GERD related NCCP patients who underwent either HREM or CM. A total of 150 patients had CM and the other 150 patients HREM. The Chicago 3 classification and the Castell and Spechler classification were used to determine the esophageal motor disorder of patients undergoing HREM and CM, respectively. Results: In both HREM and CM groups, a normal esophageal motility study was the most frequent finding (47% vs. 36%, respectively). In patients who underwent CM, the most commonly demonstrated esophageal motility disorder was hypotensive lower esophageal sphincter. Hypotensive LES was a significantly more common finding in patients undergoing CM versus HREM (27.3% vs. 4.7%, p<0.001). Other less common motility disorders seen in the CM group were nonspecific esophageal motility disorders (NEMD) (12%), nutcracker esophagus (9.3%), ineffective esophageal motility (IEM) (6.7%), achalasia (2.7%), distal esophageal spasm (DES) (5.3%), and absent contractility (0.67%). In general, hypertensive motility disorders, like DES and nutcracker esophagus, were less common than hypotensive motility disorders (15% vs. 35%, p=0.9) in NCCP patients. There was no significant difference in the number of hypertensive or hypotensive motility disorders diagnosed with CM versus those diagnosed with HREM. In the HREM group, IEM was the most commonly diagnosed motility abnormality. The likelihood of finding an IEM was significantly more common in the HREM group as compared with the CM group (25% vs. 7%, p<0.001). Other motility abnormalities that were less commonly diagnosed with HREM included: achalasia (7.3%), esophagogastric junction outflow obstruction (4%), absent contractility (4%), jackhammer/nutcracker esophagus (3.3%), DES (2.7%) and NEMD (1.3%). Conclusion: HREM did not improve the percentage and type of esophageal motor disorders diagnosed in NCCP patients as compared with CM. This is likely due to limitations of Chicago 3 classification. Normal esophageal motility remains the most common finding.

Objective To analyze the effects of different food bolus on esophageal motility in patients with non-obstructive esophageal dyshagia by high-resolution esophageal manometry. Methods From March 2014 to June 2015, 48 patients with non-obstructive esophageal dysphagia and 12 healthy volunteers (healthy control group) were enrolled. High-resolution manometry was tested when swallowing liquid food, semisolid food and solid food. The lower esophageal sphincter pressure (LESP), 4 second integrated relaxation pressure (4 s IRP), distal contractile integral (DCI), distal latency (DL), and breaks were analyzed. T test was performed for statistical analysis. Results According to the 2014 Chicago classification standard, among 48 patients with dysphagia, esophageal dysmotility was diagnosed in 35 patients (72.9%), while 13 patients (27.1%) had normal esophageal motility, and the most common type of esophageal motility disorder was ineffective esophageal motility (31.2%, 15/48). The LESP of the healthy control group was (10.85±3.75) mmHg (1 mmHg=0.133 kPa) and 4 s IRP was (1.90±0.84) mmHg. The LESP of dysphagia group was (12.20±8.93) mmHg and 4 s IRP was (3.25±1.02) mmHg. There was no significant difference in LESP and 4 s IRP between two groups (both P>0.05). The DCIs of liquid swallows, semisolid swallows and solid swallows of healthy control group were (589.00±292.90), (690.17±52.41) and (808.00±448.53) mmHg·s·cm, respectively, which were significantly lower than those of complete normal group in Chicago classification ((1 346.62±244.83), (1 542.46±231.19) and (1 890.31±363.26) mmHg·s·cm; t=4.76, 4.68 and 3.79; all P=0.001). The DL of solid swallows of healthy control group was (7.72±1.15) s, which was significantly lower than that of complete normal group in Chicago classification ((9.00±1.23) s; t=2.61, P=0.021). The breaks of liquid swallows, semisolid swallows and solid swallows of healthy control group were (2.33±1.74), (2.37±1.72) and (1.53±1.22) cm, respectively, which were higher than those of complete normal group in Chicago classification ((0.58±0.48), (0.52±0.47) and (0.85±0.53) cm), and the differences were statistically significant (t=3.02, 3.68 and 2.54, all P<0.05). Conclusions The most common type of esophageal motility disorder in patients with non-obstructive esophageal dysphagia is ineffective esophageal molitity. When swallowing food, the patients with dysphagia but normal results of esophageal manometry according to Chicago classification require more strength of the esophagus, more complete contraction and longer peristaltic time to swallow food bolus. Key words: High-resolution esophageal manometry; Non-obstructive esophageal dysphagia; Different food bolus swallow

Introduction: While esophageal motor dysfunction is known to be associated with systemic sclerosis, the prevalence of esophageal dysmotility in other rheumatic conditions is not well known. The aim of this study was to examine the prevalence of esophageal motility disorders in rheumatic diseases. Methods: In this retrospective cohort study, adults with a diagnosis of a rheumatic disease [rheumatoid arthritis (RA), Sjogren’s syndrome, systemic sclerosis (SS), systemic lupus erythematous (SLE), mixed connective tissue disease (MCTD), idiopathic Raynaud’s phenomenon, and other], who underwent high-resolution esophageal manometry (HREM) were included. Demographics, clinical characteristics, rheumatic diseases and associated therapies, HREM findings, motility disorders per Chicago classification (CCv3.0) criteria and swallow characteristics, were extracted from the medical record. We analyzed the data using descriptive statistics to assess the prevalence and types of esophageal motor disorders amongst patients with known rheumatic conditions. Results: Of 118 patients, 80% were women, and 73% were white (Table). The most common esophageal symptoms were dysphagia (88%), heartburn (81%), and regurgitation (41%). Rheumatic diseases included RA (38%), Raynaud’s (30%), Sjogren’s (22%), SS (19%) and MCTD (11%). On HREM, 53% had an esophageal motility disorder per Chicago classification. Of note, 38% of the sample with a normal barium swallow had an esophageal dysmotility disorder on HREM. Esophageal hypocontractility (including ineffective esophageal motility, absent contractility, and fragmented peristalsis) was diagnosed in 29 patients (25%), achalasia and/or esophagogastric junction outflow obstruction were found in 25 (21%), and esophageal hypercontractility (including diffuse esophageal spasm and jackhammer esophagus) in 8 (7%). These esophageal disorders were found amongst rheumatic diseases including SLE, RA, Sjogren’s, SS, MCTD, psoriatic arthritis and Raynaud’s (Figure). Conclusion: Esophageal motor disorders were found in more than half of patients with rheumatic diseases who underwent HREM, even though many patients previously had a normal barium swallow. HREM should be considered in patients with rheumatic conditions who present with esophageal symptoms.Figure 1.: Disorders of Esophageal Motility by Rheumatologic DiseaseTable 1.: Patient Characteristics and Manometry Findings

Systemic sclerosis is an autoimmune disease characterized by vasculopathy and tissue fibrosis. Esophageal disorders are often associated. The diagnosis is confirmed and characterized by high-resolution esophageal manometry (HRM). This study aims to describe the esophageal motility disorders found by high-resolution esophageal manometry, in patients followed for systemic sclerosis. An observational study carried out between April 2018 and January 2020, including 23 patients with systemic sclerosis according to EULAR/ACR 2013 criteria, referred for high-resolution esophageal manometry. We used the Chicago V3.0 classification to order esophageal motor disorders into two categories: major and minor disorders of peristalsis. Four parameters were analyzed: lower esophageal sphincter resting pressure (LES), distal latency (DL) and distal contractile integral (DCI) and integrated relaxation pressure (IRP). IRP was normal in all cases. The digestive symptoms were: gastroesophageal reflux in 10 cases (43.4%) and dysphagia in 6 cases (26.2%). Seven patients were asymptomatic (30.4%). HRM revealed a major disorder of esophageal peristalsis in 14 cases (60.9%), minor disorders in 4 patients (17.4%) and was normal in 5 cases (21.7%). There was; hypotonia of the lower esophageal sphincter (LES) with absent contractility in 8 cases (44.5%), only absence of contractility in 6 cases (33.3%), hypotonia of the lower esophageal sphincter (LES) with ineffective motility in 4 cases (22.2%). High-resolution esophageal manometry should be considered in all patients with systemic sclerosis even in the absence of digestive symptoms. In our series, motility disorders were present in 78.3% of cases.

We study the performance of classical and quantum machine learning (ML) models in predicting outcomes of physical experiments. The experiments depend on an input parameter x and involve execution of a (possibly unknown) quantum process E. Our figure of merit is the number of runs of E required to achieve a desired prediction performance. We consider classical ML models that perform a measurement and record the classical outcome after each run of E, and quantum ML models that can access E coherently to acquire quantum data; the classical or quantum data are then used to predict the outcomes of future experiments. We prove that for any input distribution D(x), a classical ML model can provide accurate predictions on average by accessing E a number of times comparable to the optimal quantum ML model. In contrast, for achieving an accurate prediction on all inputs, we prove that the exponential quantum advantage is possible. For example, to predict the expectations of all Pauli observables in an n-qubit system ρ, classical ML models require 2^{Ω(n)} copies of ρ, but we present a quantum ML model using only O(n) copies. Our results clarify where the quantum advantage is possible and highlight the potential for classical ML models to address challenging quantum problems in physics and chemistry.

High-resolution esophageal manometry (HREM) has become a leading tool in the assessment of esophageal motor disorders, replacing conventional manometry. However, there is limited data about the contribution of HREM as compared with conventional manometry to the assessment of esophageal motor disorders in patients with non-cardiac chest pain (NCCP). The aim of the study was to compare the distribution of esophageal motor disorders in patients with NCCP using HREM as compared with conventional manometry and to determine if HREM improved diagnosis of these disorders. In this study, we included 300 consecutive patients with NCCP who underwent either HREM or conventional manometry over a period of 10years. A total of 150 patients had conventional manometry and the other 150 patients HREM. The Chicago 3.0 classification and the Castell and Spechler classification were used to determine the esophageal motor disorder of NCCP patients undergoing HREM and conventional manometry, respectively. In both HREM and the conventional manometry groups, normal esophageal motility was the most frequent finding (47% and 36%; respectively, P=.054). Hypotensive lower esophageal sphincter was the most common motility disorder identified by conventional manometry (27.3%), while ineffective esophageal motility was the most common esophageal motor disorder identified by HREM (25.3%). There is a discrepancy in the type of esophageal motor disorders identified by HREM as compared with conventional manometry in NCCP patients. Hypotensive motility disorders are the most commonly diagnosed by both manometric techniques.

Loss of Peristaltic Reserve, Determined by Multiple Rapid Swallows, Is the Most Frequent Esophageal Motility Abnormality in Patients With Systemic Sclerosis.

The classic manometric findings in systemic sclerosis are aperistalsis of the esophageal body with hypotensive lower esophageal sphincter. These changes contribute to gastroesophageal reflux disease in these patients. With widespread use of high-resolution esophageal manometry, diverse abnormalities are seen. The aim of this study is to characterize esophageal dysmotility in patients with systemic sclerosis undergoing high-resolution esophageal manometry and compare demographic features and diagnostic test results among patients with varying degrees of esophageal dysmotility. Patients with systemic sclerosis who underwent high-resolution esophageal manometry between January 2008 and October 2014 at our institution were identified. High-resolution esophageal manometry studies were reinterpreted using the Chicago Classification, v3.0 criteria. We also reviewed the patient charts for demographic data, indications for manometry, esophagogastroduodenoscopy findings, pH studies, medication use, and autoantibody panel. The cohort consisted of 122 patients with a mean age of 53.3±15.3 years. High-resolution esophageal manometry was normal in 23, showed ineffective esophageal motility in 22, absent contractility in 73, and one case each of type II achalasia, esophagogastric junction outflow obstruction, hypercontractile esophagus, and distal esophageal spasm. Patients with absent contractility were younger and more likely to have erosive esophagitis, hiatal hernia, and esophageal strictures than patients with ineffective esophageal motility or normal manometry. There were no statistically significant differences in the groups based on autoantibodies or indications for manometry. Diverse esophageal motility abnormalities were noted in systemic sclerosis with ineffective esophageal motility or absent contractility observed in over three-fourth of the patients. Patients with absent contractility were younger and had more severe reflux. The severity of gastroesophageal reflux disease related endoscopic findings correlated with the degree of esophageal dysmotility on high-resolution esophageal manometry.

The development of the high-resolution esophageal manometry (HRM) and the Chicago classification have improved the diagnosis and management of esophageal motility disorders. However, some conditions have yet to be addressed by this classification. This review describes findings in HRM which are not included in the current Chicago classification based on the experience in our center. This includes the analysis of the upper esophageal sphincter, proximal esophagus, longitudinal muscle contraction, disorders related to gastroesophageal reflux disease and respiratory symptoms. The utility of provocative tests and the use of HRM in the evaluation of rumination syndrome and post-surgical patients will also be discussed. We believe that characterization of the manometric findings in these areas will eventually lead to incorporation of new criteria into the existing classification.

In patients with dysphagia that is not explained by upper endoscopy, high-resolution esophageal manometry (HRM) is the next logical step in diagnostic testing. This study investigated predictors of failure to refer for HRM after an upper endoscopy that was performed for but did not explain dysphagia. This was a retrospective cohort study of patients >18 years of age who underwent esophagogastroduodenoscopy (EGD) for dysphagia from 2015 to 2021. Patients with EGD findings that explained dysphagia (e.g. esophageal mass, eosinophilic esophagitis, Schatzki ring, etc.) were excluded from the main analyses. The primary outcome was failure to refer for HRM within 1 year of the index non-diagnostic EGD. We also investigated delayed referral for HRM, defined as HRM performed after the median. Multivariable logistic regression modeling was used to identify risk factors that independently predicted failure to refer for HRM, conditioned on the providing endoscopist. Among 2132 patients who underwent EGD for dysphagia, 1240 (58.2%) did not have findings to explain dysphagia on the index EGD. Of these 1240 patients, 148 (11.9%) underwent HRM within 1 year of index EGD. Endoscopic findings (e.g. hiatal hernia, tortuous esophagus, Barrett's esophagus, surgically altered anatomy not involving the gastroesophageal junction, and esophageal varices) perceived to explain dysphagia were independently associated with failure to refer for HRM (adjusted odds ratio 0.45, 95% confidence interval 0.25-0.80). Of the 148 patients who underwent HRM within 1 year of index EGD, 29.7% were diagnosed with a disorder of esophagogastric junction outflow, 17.6% with a disorder of peristalsis, and 2.0% with both disorders of esophagogastric outflow and peristalsis. The diagnosis made by HRM was similar among those who had incidental EGD findings that were non-diagnostic for dysphagia compared with those who had completely normal EGD findings. Demographic factors including race/ethnicity, insurance type, and income were not associated with failure to refer for HRM or delayed HRM. Patients with dysphagia and endoscopic findings unrelated to dysphagia have a similar prevalence of esophageal motility disorders to those with normal endoscopic examinations, yet these patients are less likely to undergo HRM. Provider education is indicated to increase HRM referral in these patients.

High resolution esophageal manometry (HREM) has been interpreted all along by visual interpretation of color plots until the recent introduction of Chicago classification which categorises HREM using objective measurements. It compares HREM diagnosis of esophageal motor disorders by visual interpretation and Chicago classification. Using software Trace 1.2v, 77 consecutive tracings diagnosed by visual interpretation were re-analyzed by Chicago classification and findings compared for concordance between the two systems of interpretation. Kappa agreement rate between the two observations was determined. There were 57 males (74 %) and cohort median age was 41 years (range: 14-83 years). Majority of the referrals were for gastroesophageal reflux disease, dysphagia and achalasia. By "intuitive" visual interpretation, the tracing were reported as normal in 45 (58.4 %), achalasia 14 (18.2 %), ineffective esophageal motility 3 (3.9 %), nutcracker esophagus 11 (14.3 %) and nonspecific motility changes 4 (5.2 %). By Chicago classification, there was 100 % agreement (Kappa 1) for achalasia (type 1: 9; type 2: 5) and ineffective esophageal motility ("failed peristalsis" on visual interpretation). Normal esophageal motility, nutcracker esophagus and nonspecific motility disorder on visual interpretation were reclassified as rapid contraction and esophagogastric junction (EGJ) outflow obstruction by Chicago classification. Chicago classification identified distinct clinical phenotypes including EGJ outflow obstruction not identified by visual interpretation. A significant number of unclassified HREM by visual interpretation were also classified by it.

IntroductionAlthough high resolution oesophageal manometry (HROM) is the gold standard for evaluating oesophageal dysmotility, barium swallow (BS) is widely available and is commonly used. In this study we aim to...

Esophageal motility disorders (EMDs) can significantly impact patients' quality of life. The Chicago Classification (CC) was developed as a robust framework to enable clinicians to better understand and classify the nature of motility disorders. Previous studies have primarily focused on the CC version 3.0 (CCv3.0), and data regarding the correlation between symptoms and CC version 4.0 (CCv4.0) in the Saudi Arabian population are lacking. This study aimed to assess the correlation between symptoms and CCv3.0 and CCv4.0 using high-resolution esophageal manometry (HRM) in Saudi Arabia, to evaluate the diagnostic performance of both classifications. A total of 182 patients presenting with esophageal symptoms were included in this study. HRM was performed to assess esophageal motility, and patients' reported symptoms were recorded. The association between HRM findings and symptomatic variables was analyzed using sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). Variability was observed in the diagnostic performance of symptomatic variables for major EMDs. CCv4.0 demonstrated a higher sensitivity for dysphagia than CCv3.0; however, it exhibited lower sensitivity to atypical gastroesophageal reflux disease (GERD) symptoms. Noncardiac chest pain (NCCP) exhibited the highest specificity and PPV, whereas typical GERD symptoms showed lower specificity. CCv4.0 demonstrated potential improvements in sensitivity for dysphagia, but lower sensitivity for atypical GERD symptoms, compared with CCv3.0. These insights provide guidance for clinicians in Saudi Arabia and contribute to understanding the diagnostic performance of CCv3.0 and CCv4.0.

Deep neural networks (DNNs) have shown success in image classification, with high accuracy in recognition of everyday objects. Performance of DNNs has traditionally been measured assuming human accuracy is perfect. In specific problem domains, however, human accuracy is less than perfect and a comparison between humans and machine learning (ML) models can be performed. In recognising everyday objects, humans have the advantage of a lifetime of experience, whereas DNN models are trained only with a limited image dataset. We have tried to compare performance of human learners and two DNN models on an image dataset which is novel to both, i.e. histological images. We thus aim to eliminate the advantage of prior experience that humans have over DNN models in image classification. Ten classes of tissues were randomly selected from the undergraduate first year histology curriculum of a Medical School in North India. Two machine learning (ML) models were developed based on the VGG16 (VML) and Inception V2 (IML) DNNs, using transfer learning, to produce a 10-class classifier. One thousand (1000) images belonging to the ten classes (i.e. 100 images from each class) were split into training (700) and validation (300) sets. After training, the VML and IML model achieved 85.67 and 89% accuracy on the validation set, respectively. The training set was also circulated to medical students (MS) of the college for a week. An online quiz, consisting of a random selection of 100 images from the validation set, was conducted on students (after obtaining informed consent) who volunteered for the study. 66 students participated in the quiz, providing 6557 responses. In addition, we prepared a set of 10 images which belonged to different classes of tissue, not present in training set (i.e. out of training scope or OTS images). A second quiz was conducted on medical students with OTS images, and the ML models were also run on these OTS images. The overall accuracy of MS in the first quiz was 55.14%. The two ML models were also run on the first quiz questionnaire, producing accuracy between 91 and 93%. The ML models scored more than 80% of medical students. Analysis of confusion matrices of both ML models and all medical students showed dissimilar error profiles. However, when comparing the subset of students who achieved similar accuracy as the ML models, the error profile was also similar. Recognition of ‘stomach’ proved difficult for both humans and ML models. In 04 images in the first quiz set, both VML model and medical students produced highly equivocal responses. Within these images, a pattern of bias was uncovered–the tendency of medical students to misclassify ‘liver’ tissue. The ‘stomach’ class proved most difficult for both MS and VML, producing 34.84% of all errors of MS, and 41.17% of all errors of VML model; however, the IML model committed most errors in recognising the ‘skin’ class (27.5% of all errors). Analysis of the convolution layers of the DNN outlined features in the original image which might have led to misclassification by the VML model. In OTS images, however, the medical students produced better overall score than both ML models, i.e. they successfully recognised patterns of similarity between tissues and could generalise their training to a novel dataset. Our findings suggest that within the scope of training, ML models perform better than 80% medical students with a distinct error profile. However, students who have reached accuracy close to the ML models, tend to replicate the error profile as that of the ML models. This suggests a degree of similarity between how machines and humans extract features from an image. If asked to recognise images outside the scope of training, humans perform better at recognising patterns and likeness between tissues. This suggests that ‘training’ is not the same as ‘learning’, and humans can extend their pattern-based learning to different domains outside of the training set.