Artificial Intelligence in the Diagnosis of Upper Gastrointestinal Diseases

Expanding role and scope of artificial intelligence in the field of gastrointestinal pathology

Digital pathology (DP) and its subsidiaries including artificial intelligence (AI) are rapidly making inroads into the area of diagnostic anatomic pathology (AP) including gastrointestinal (GI) pathology. It is poised to revolutionize the field of diagnostic AP. Historically, AP has been slow to adopt digital technology, but this is changing rapidly, with many centers worldwide transitioning to DP. Coupled with advanced techniques of AI such as deep learning and machine learning, DP is likely to transform histopathology from a subjective field to an objective, efficient, and transparent discipline. AI is increasingly integrated into GI pathology, offering numerous advancements and improvements in overall diagnostic accuracy, efficiency, and patient care. Specifically, AI in GI pathology enhances diagnostic accuracy, streamlines workflows, provides predictive insights, integrates multimodal data, supports research, and aids in education and training, ultimately improving patient care and outcomes. This review summarized the latest developments in the role and scope of AI in AP with a focus on GI pathology. The main aim was to provide updates and create awareness among the pathology community.

Role of Artificial Intelligence in the Diagnosis of Gastroesophageal Reflux Disease

Gastroesophageal reflux disease (GERD) is a disorder that usually presents with heartburn. GERD is diagnosed clinically, but most patients are misdiagnosed due to atypical presentations. The increased use of artificial intelligence (AI) in healthcare has provided multiple ways of diagnosing and treating patients accurately. In this review, multiple studies in which AI models were used to diagnose GERD are discussed. According to the studies, using AI models helped to diagnose GERD in patients accurately. AI, although considered one of the most potent emerging aspects of medicine with its accuracy in patient diagnosis, presents limitations of its own, which explains why healthcare providers may hesitate to use AI in patient care. The challenges and limitations should be addressed before AI is fully incorporated into the healthcare system.

Artificial Intelligence Tools for the Diagnosis of Eosinophilic Esophagitis in Adults Reporting Dysphagia: Development, External Validation, and Software Creation for Point-of-Care Use

BACKGROUND

Despite increased awareness of eosinophilic esophagitis (EoE), the diagnostic delay has remained stable over the past 3 decades. There is a need to improve the diagnostic performance and optimize resources allocation in the setting of EoE.

OBJECTIVE

We developed and validated 2 point-of-care machine learning (ML) tools to predict a diagnosis of EoE before histology results during office visits.

METHODS

We conducted a multicenter study in 3 European tertiary referral centers for EoE. We built predictive ML models using retrospectively extracted clinical and esophagogastroduodenoscopy (EGDS) data collected from 273 EoE and 55 non-EoE dysphagia patients. We validated the models on an independent cohort of 93 consecutive patients with dysphagia undergoing EGDS with biopsies at 2 different centers. Models' performance was assessed by area under the curve (AUC), sensitivity, specificity, and positive and negative predictive values (PPV and NPV). The models were integrated into a point-of-care software package.

RESULTS

The model trained on clinical data alone showed an AUC of 0.90 and a sensitivity, specificity, PPV, and NPV of 0.90, 0.75, 0.80, and 0.87, respectively, for the diagnosis of EoE in the external validation cohort. The model trained on a combination of clinical and endoscopic data showed an AUC of 0.94, and a sensitivity, specificity, PPV, and NPV of 0.94, 0.68, 0.77, and 0.91, respectively, in the external validation cohort.

CONCLUSION

Our software-integrated models (https://webapplicationing.shinyapps.io/PointOfCare-EoE/) can be used at point-of-care to improve the diagnostic workup of EoE and optimize resources allocation.

Artificial Intelligence Tools for Improving Manometric Diagnosis of Esophageal Dysmotility

PURPOSE OF REVIEW

Artificial intelligence (AI) is a broad term that pertains to a computer's ability to mimic and sometimes surpass human intelligence in interpretation of large datasets. The adoption of AI in gastrointestinal motility has been slower compared to other areas such as polyp detection and interpretation of histopathology.

RECENT FINDINGS

Within esophageal physiologic testing, AI can automate interpretation of image-based tests, especially high resolution manometry (HRM) and functional luminal imaging probe (FLIP) studies. Basic tasks such as identification of landmarks, determining adequacy of the HRM study and identification from achalasia from non-achalasia patterns are achieved with good accuracy. However, existing AI systems compare AI interpretation to expert analysis rather than to clinical outcome from management based on AI diagnosis. The use of AI methods is much less advanced within the field of ambulatory reflux monitoring, where challenges exist in assimilation of data from multiple impedance and pH channels. There remains potential for replication of the AI successes within esophageal physiologic testing to HRM of the anorectum, and to innovative and novel methods of evaluating gastric electrical activity and motor function. The use of AI has tremendous potential to improve detection of dysmotility within the esophagus using esophageal physiologic testing, as well as in other regions of the gastrointestinal tract. Eventually, integration of patient presentation, demographics and alternate test results to individual motility test interpretation will improve diagnostic precision and prognostication using AI tools.

Deep learning to predict esophageal variceal bleeding based on endoscopic images

OBJECTIVE

Esophageal varix (EV) bleeding is a particularly serious complications of cirrhosis. Prediction of EV bleeding requires extensive endoscopy experience; it remains unreliable and inefficient. This retrospective cohort study evaluated the feasibility of using deep learning (DL) to predict the 12-month risk of EV bleeding based on endoscopic images.

METHODS

Six DL models were trained to perform binary classification of endoscopic images of EV bleeding. The models were subsequently validated using an external test dataset, then compared with classifications performed by two endoscopists.

RESULTS

In the validation dataset, EfficientNet had the highest accuracy (0.910), followed by ConvMixer (0.898) and Xception (0.875). In the test dataset, EfficientNet maintained the highest accuracy (0.893), which was better than the endoscopists (0.800 and 0.763). Notably, one endoscopist displayed higher recall (0.905), compared with EfficientNet (0.870). When their predictions were assisted by artificial intelligence, the accuracies of the two endoscopists increased by 17.3% and 19.0%. Moreover, statistical agreement among the models was dependent on model architecture.

CONCLUSIONS

This study demonstrated the feasibility of using DL to predict the 12-month risk of EV bleeding based on endoscopic images. The findings suggest that artificial intelligence-aided diagnosis will be a useful addition to cirrhosis management.

Robotic Colonoscopy and Beyond: Insights into Modern Lower Gastrointestinal Endoscopy

Lower gastrointestinal endoscopy is considered the gold standard for the diagnosis and removal of colonic polyps. Delays in colonoscopy following a positive fecal immunochemical test increase the likelihood of advanced adenomas and colorectal cancer (CRC) occurrence. However, patients may refuse to undergo conventional colonoscopy (CC) due to fear of possible risks and pain or discomfort. In this regard, patients undergoing CC frequently require sedation to better tolerate the procedure, increasing the risk of deep sedation or other complications related to sedation. Accordingly, the use of CC as a first-line screening strategy for CRC is hampered by patients' reluctance due to its invasiveness and anxiety about possible discomfort. To overcome the limitations of CC and improve patients' compliance, several studies have investigated the use of robotic colonoscopy (RC) both in experimental models and in vivo. Self-propelling robotic colonoscopes have proven to be promising thanks to their peculiar dexterity and adaptability to the shape of the lower gastrointestinal tract, allowing a virtually painless examination of the colon. In some instances, when alternatives to CC and RC are required, barium enema (BE), computed tomographic colonography (CTC), and colon capsule endoscopy (CCE) may be options. However, BE and CTC are limited by the need for subsequent investigations whenever suspicious lesions are found. In this narrative review, we discussed the current clinical applications of RC, CTC, and CCE, as well as the advantages and disadvantages of different endoscopic procedures, with a particular focus on RC.

Encouraging appropriate use of proton pump inhibitors: existing initiatives and proposals for the future

INTRODUCTION

Proton pump inhibitors (PPIs) have revolutionized the management of acid-related disorders, representing today the mainstay treatment of these conditions. However, despite their large range of indications and usefulness, the remarkable expansion of their use in the last two decades cannot be explained by the increasing prevalence of acid-related diseases only. An inappropriate prescription for clinical conditions in which the pathogenetic role of acid has not been documented has been described, with the natural consequence of increasing the costs and the potential risk of iatrogenic harm due to adverse events and complications recently emerged.

AREAS COVERED

In this review, we summarize current indications of PPIs administration, potential adverse events associated with their chronic utilization, and misuse of PPIs. Moreover, we describe existing and possible initiatives for improving the use of PPIs, and some proposals for the future.

EXPERT OPINION

PPI deprescribing is the preferred and most effective approach to reduce the use of PPIs, rather than adopting sharp discontinuation, probably due to fewer withdrawal symptoms. Nonetheless, large knowledge gaps still exist in clinical practice regarding the optimal approach of PPI deprescribing in various clinical scenarios. Further prospective well-designed international studies are eagerly warranted to improve our perspectives on controlling global PPI inappropriate use.

Endoscopy in Pediatric Eosinophilic Esophagitis

Eosinophilic esophagitis (EoE) is a chronic allergen-mediated clinicopathologic condition that currently requires esophagogastroduodenoscopy with biopsies and histologic evaluation to diagnose and monitor its progress. This state-of-the art review outlines the pathophysiology of EoE, reviews the application of endoscopy as a diagnostic and therapeutic tool, and discusses potential complications related to therapeutic endoscopic interventions. It also introduces recent innovations that can enhance the endoscopist's ability to diagnose and monitor EoE with minimally invasive procedures and perform therapeutic maneuvers more safely and effectively.

Automated Multimodal Machine Learning for Esophageal Variceal Bleeding Prediction Based on Endoscopy and Structured Data

Esophageal variceal (EV) bleeding is a severe medical emergency related to cirrhosis. Early identification of cirrhotic patients with at a high risk of EV bleeding is key to improving outcomes and optimizing medical resources. This study aimed to evaluate the feasibility of automated multimodal machine learning (MMML) for predicting EV bleeding by integrating endoscopic images and clinical structured data. This study mainly includes three steps: step 1, developing deep learning (DL) models using EV images by 12-month bleeding on TensorFlow (backbones include ResNet, Xception, EfficientNet, ViT and ConvMixer); step 2, training and internally validating MMML models integrating clinical structured data and DL model outputs to predict 12-month EV bleeding on an H2O-automated machine learning platform (algorithms include DL, XGBoost, GLM, GBM, RF, and stacking); and step 3, externally testing MMML models. Furthermore, existing clinical indices, e.g., the MELD score, Child‒Pugh score, APRI, and FIB-4, were also examined. Five DL models were transfer learning to the binary classification of EV endoscopic images at admission based on the occurrence or absence of bleeding events during the 12-month follow-up. An EfficientNet model achieved the highest accuracy of 0.868 in the validation set. Then, a series of MMML models, integrating clinical structured data and the output of the EfficientNet model, were automatedly trained to predict 12-month EV bleeding. A stacking model showed the highest accuracy (0.932), sensitivity (0.952), and F1-score (0.879) in the test dataset, which was also better than the existing indices. This study is the first to evaluate the feasibility of automated MMML in predicting 12-month EV bleeding based on endoscopic images and clinical variables.

Ambulatory pH-Impedance Findings Confirm That Grade B Esophagitis Provides Objective Diagnosis of Gastroesophageal Reflux Disease

INTRODUCTION

The Lyon Consensus designates Los Angeles (LA) grade C/D esophagitis or acid exposure time (AET) >6% on impedance-pH monitoring (MII-pH) as conclusive for gastroesophageal reflux disease (GERD). We aimed to evaluate proportions with objective GERD among symptomatic patients with LA grade A, B, and C esophagitis on endoscopy.

METHODS

Demographics, clinical data, endoscopy findings, and objective proton-pump inhibitor response were collected from symptomatic prospectively enrolled patients from 2 referral centers. Off-therapy MII-pH parameters included AET, number of reflux episodes, mean nocturnal baseline impedance, and postreflux swallow-induced peristaltic wave index. Objective GERD evidence was compared between LA grades.

RESULTS

Of 155 patients (LA grade A: 74 patients, B: 61 patients, and C: 20 patients), demographics and presentation were similar across LA grades. AET >6% was seen in 1.4%, 52.5%, and 75%, respectively, in LA grades A, B, and C. Using additional MII-pH metrics, an additional 16.2% with LA grade A and 47.5% with LA grade B esophagitis had AET 4%-6% with low mean nocturnal baseline impedance and postreflux swallow-induced peristaltic wave index; there were no additional gains using the number of reflux episodes or symptom-reflux association metrics. Compared with LA grade C (100% conclusive GERD based on endoscopic findings), 100% of LA grade B esophagitis also had objective GERD but only 17.6% with LA grade A esophagitis ( P < 0.001 compared with each). Proton-pump inhibitor response was comparable between LA grades B and C (74% and 70%, respectively) but low in LA grade A (39%, P < 0.001).

DISCUSSION

Grade B esophagitis indicates an objective diagnosis of GERD.

Achalasia, from diagnosis to treatment

INTRODUCTION

Achalasia is an uncommon esophageal motility disorder and is characterized by alterations of the motility of the esophageal body in conjunction with altered lower esophageal sphincter (LES) relaxation. The clinical presentation of patients with achalasia may be complex; however, the most frequent symptom is dysphagia. The management of patients with achalasia is often challenging, due to the heterogeneous clinical presentation.

AREAS COVERED

The diagnosis and management of achalasia has significantly improved in the last years due to the growing availability of high-resolution manometry (HRM) and the implementation in the therapeutic armamentarium of new therapeutic endoscopic procedures. Traditional therapeutic strategies include botulinum toxin injected to the LES and pneumatic balloon dilation. On the other hand, surgical treatments contemplate laparoscopic Heller myotomy and, less frequently, esophagectomy. Furthermore, in the last few years, per oral endoscopic myotomy (POEM) has been proposed as the main endoscopic therapeutic alternative to the laparoscopic Heller myotomy.

EXPERT OPINION

Diagnosis and treatment of achalasia still represent a challenging area. However, we believe that an accurate up-front evaluation is, nowadays, necessary in addressing patients with achalasia for a more accurate diagnosis as well as for the best treatment options.

Development and Validation of Deep Learning Models for the Multiclassification of Reflux Esophagitis Based on the Los Angeles Classification

This study is to evaluate the feasibility of deep learning (DL) models in the multiclassification of reflux esophagitis (RE) endoscopic images, according to the Los Angeles (LA) classification for the first time. The images were divided into three groups, namely, normal, LA classification A + B, and LA C + D. The images from the HyperKvasir dataset and Suzhou hospital were divided into the training and validation datasets as a ratio of 4 : 1, while the images from Jintan hospital were the independent test set. The CNNs- or Transformer-architectures models (MobileNet, ResNet, Xception, EfficientNet, ViT, and ConvMixer) were transfer learning via Keras. The visualization of the models was proposed using Gradient-weighted Class Activation Mapping (Grad-CAM). Both in the validation set and the test set, the EfficientNet model showed the best performance as follows: accuracy (0.962 and 0.957), recall for LA A + B (0.970 and 0.925) and LA C + D (0.922 and 0.930), Marco-recall (0.946 and 0.928), Matthew's correlation coefficient (0.936 and 0.884), and Cohen's kappa (0.910 and 0.850), which was better than the other models and the endoscopists. According to the EfficientNet model, the Grad-CAM was plotted and highlighted the target lesions on the original images. This study developed a series of DL-based computer vision models with the interpretable Grad-CAM to evaluate the feasibility in the multiclassification of RE endoscopic images. It firstly suggests that DL-based classifiers show promise in the endoscopic diagnosis of esophagitis.

Global research trends of artificial intelligence applied in esophageal carcinoma: A bibliometric analysis (2000-2022) via CiteSpace and VOSviewer

Objective

Using visual bibliometric analysis, the application and development of artificial intelligence in clinical esophageal cancer are summarized, and the research progress, hotspots, and emerging trends of artificial intelligence are elucidated.

Methods

On April 7th, 2022, articles and reviews regarding the application of AI in esophageal cancer, published between 2000 and 2022 were chosen from the Web of Science Core Collection. To conduct co-authorship, co-citation, and co-occurrence analysis of countries, institutions, authors, references, and keywords in this field, VOSviewer (version 1.6.18), CiteSpace (version 5.8.R3), Microsoft Excel 2019, R 4.2, an online bibliometric platform (http://bibliometric.com/) and an online browser plugin (https://www.altmetric.com/) were used.

Results

A total of 918 papers were included, with 23,490 citations. 5,979 authors, 39,962 co-cited authors, and 42,992 co-cited papers were identified in the study. Most publications were from China (317). In terms of the H-index (45) and citations (9925), the United States topped the list. The journal “New England Journal of Medicine” of Medicine, General &amp; Internal (IF = 91.25) published the most studies on this topic. The University of Amsterdam had the largest number of publications among all institutions. The past 22 years of research can be broadly divided into two periods. The 2000 to 2016 research period focused on the classification, identification and comparison of esophageal cancer. Recently (2017-2022), the application of artificial intelligence lies in endoscopy, diagnosis, and precision therapy, which have become the frontiers of this field. It is expected that closely esophageal cancer clinical measures based on big data analysis and related to precision will become the research hotspot in the future.

Conclusions

An increasing number of scholars are devoted to artificial intelligence-related esophageal cancer research. The research field of artificial intelligence in esophageal cancer has entered a new stage. In the future, there is a need to continue to strengthen cooperation between countries and institutions. Improving the diagnostic accuracy of esophageal imaging, big data-based treatment and prognosis prediction through deep learning technology will be the continuing focus of research. The application of AI in esophageal cancer still has many challenges to overcome before it can be utilized.

Applications of Artificial Intelligence to Eosinophilic Esophagitis

Eosinophilic Esophagitis (EoE) is a chronic immune-related inflammation, and challenges to its diagnosis and treatment evaluation persist. This literature review evaluates all AI applications to EOE, including 15 studies using AI algorithms for counting eosinophils in biopsies, as well as newer diagnostics using mRNA transcripts in biopsies, endoscopic photos, blood and urine biomarkers, and an improved scoring system for disease classification. We also discuss the clinical impact of these models, challenges faced in applying AI to EoE, and future applications. In conclusion, AI has the potential to improve diagnostics and clinical evaluation in EoE, improving patient outcomes.

Artificial Intelligence for Upper Gastrointestinal Endoscopy: A Roadmap from Technology Development to Clinical Practice

Stomach cancer is the third deadliest type of cancer in the world (0.86 million deaths in 2017). In 2035, a 20% increase will be observed both in incidence and mortality due to demographic effects if no interventions are foreseen. Upper GI endoscopy (UGIE) plays a paramount role in early diagnosis and, therefore, improved survival rates. On the other hand, human and technical factors can contribute to misdiagnosis while performing UGIE. In this scenario, artificial intelligence (AI) has recently shown its potential in compensating for the pitfalls of UGIE, by leveraging deep learning architectures able to efficiently recognize endoscopic patterns from UGIE video data. This work presents a review of the current state-of-the-art algorithms in the application of AI to gastroscopy. It focuses specifically on the threefold tasks of assuring exam completeness (i.e., detecting the presence of blind spots) and assisting in the detection and characterization of clinical findings, both gastric precancerous conditions and neoplastic lesion changes. Early and promising results have already been obtained using well-known deep learning architectures for computer vision, but many algorithmic challenges remain in achieving the vision of AI-assisted UGIE. Future challenges in the roadmap for the effective integration of AI tools within the UGIE clinical practice are discussed, namely the adoption of more robust deep learning architectures and methods able to embed domain knowledge into image/video classifiers as well as the availability of large, annotated datasets.

Advances on Neurogastroenterology and Motility Disorders: Pathophysiology, Diagnostics and Management

Symptoms related to abnormalities in gastrointestinal tract motility and functions are very common in the general population, affecting both pediatrics and adults, from both sexes [...].

Clinical use of mean nocturnal baseline impedance and post-reflux swallow-induced peristaltic wave index for the diagnosis of gastro-esophageal reflux disease

The clinical diagnosis of gastro-esophageal reflux disease (GERD) is based on the presence of typical esophageal troublesome symptoms. In clinical practice, heartburn relief following a proton pump inhibitor (PPI) trial or endoscopy can confirm a diagnosis of GERD. In cases of diagnostic uncertainty or before anti-reflux interventions, combined impedance-pH monitoring (MII-pH) provides a comprehensive assessment of both physical and chemical properties of the refluxate, allowing to achieve a conclusive diagnosis of GERD. Recently, the Lyon Consensus proposed the use of mean nocturnal baseline impedance (MNBI) and post-reflux swallow-induced peristaltic wave index (PSPW-I) as novel MII-pH metrics to support the diagnosis of GERD. The calculation of MNBI and PSPW-I currently needs to be performed manually, but artificial intelligence systems for the automated analysis of MII-pH tracings are being developed. Several studies demonstrated the increased diagnostic yield MNBI and PSPW-I for the categorization of patients with GERD at both on- and off-PPI MII-pH monitoring. Accordingly, we performed a narrative review on the clinical use and diagnostic yield of MNBI and PSPW-I when the diagnosis of GERD is uncertain. Based on currently available evidence, we strongly support the evaluation of PSPW-I and MNBI as part of the standard assessment of MII-pH tracings for the evaluation of GERD, especially in patients with endoscopy-negative heartburn.