Usefulness of an artificial intelligence system for the detection of esophageal squamous cell carcinoma evaluated with videos simulating overlooking situation

Effects of artificial intelligence assistance on endoscopist performance: Comparison of diagnostic performance in superficial esophageal squamous cell carcinoma detection using video-based models

Objectives

Superficial esophageal squamous cell carcinoma (ESCC) detection is crucial. Although narrow-band imaging improves detection, its effectiveness is diminished by inexperienced endoscopists. The effects of artificial intelligence (AI) assistance on ESCC detection by endoscopists remain unclear. Therefore, this study aimed to develop and validate an AI model for ESCC detection using endoscopic video analysis and evaluate diagnostic improvements.

Methods

Endoscopic videos with and without ESCC lesions were collected from May 2020 to January 2022. The AI model trained on annotated videos and 18 endoscopists (eight experts, 10 non-experts) evaluated their diagnostic performance. After 4 weeks, the endoscopists re-evaluated the test data with AI assistance. Sensitivity, specificity, and accuracy were compared between endoscopists with and without AI assistance.

Results

Training data comprised 280 cases (140 with and 140 without lesions), and test data, 115 cases (52 with and 63 without lesions). In the test data, the median lesion size was 14.5 mm (range: 1-100 mm), with pathological depths ranging from high-grade intraepithelial to submucosal neoplasia. The model's sensitivity, specificity, and accuracy were 76.0%, 79.4%, and 77.2%, respectively. With AI assistance, endoscopist sensitivity (57.4% vs. 66.5%) and accuracy (68.6% vs. 75.9%) improved significantly, while specificity increased slightly (87.0% vs. 91.6%). Experts demonstrated substantial improvements in sensitivity (59.1% vs. 70.0%) and accuracy (72.1% vs. 79.3%). Non-expert accuracy increased significantly (65.8% vs. 73.3%), with slight improvements in sensitivity (56.1% vs. 63.7%) and specificity (81.9% vs. 89.2%).

Conclusions

AI assistance enhances ESCC detection and improves endoscopists' diagnostic performance, regardless of experience.

Role of Artificial Intelligence in the Detection and Management of Premalignant and Malignant Lesions of the Esophagus and Stomach

The advent of artificial intelligence (AI) and deep learning algorithms, particularly convolutional neural networks, promises to address pitfalls, bridging the care for patients at high risk with improved detection (computer-aided detection [CADe]) and characterization (computer-aided diagnosis [CADx]) of lesions. This review describes the available artificial intelligence (AI) technology and the current data on AI tools for screening esophageal squamous cell cancer, Barret's esophagus-related neoplasia, and gastric cancer. These tools outperformed endoscopists in many situations. Recent randomized controlled trials have demonstrated the successful application of AI tools in clinical practice with improved outcomes.

Artificial Intelligence Applications in Image-Based Diagnosis of Early Esophageal and Gastric Neoplasms

Artificial intelligence (AI) holds the potential to transform the management of upper gastrointestinal (GI) conditions, such as Barrett's esophagus, esophageal squamous cell cancer, and early gastric cancer. Advancements in deep learning and convolutional neural networks offer improved diagnostic accuracy and reduced diagnostic variability across different clinical settings, particularly where human error or fatigue may impair diagnostic precision. Deep learning models have shown the potential to improve early cancer detection and lesion characterization, predict invasion depth, and delineate lesion margins with remarkable accuracy, all contributing to effective treatment planning. Several challenges, however, limit the broad application of AI in GI endoscopy, particularly in the upper GI tract. Subtle lesion morphology and restricted diversity in training datasets, which are often sourced from specialized centers, may constrain the generalizability of AI models in various clinical settings. Furthermore, the "black box" nature of some AI systems can impede explainability and clinician trust. To address these issues, efforts are underway to incorporate multimodal data, such as combining endoscopic and histopathologic imaging, to bolster model robustness and transparency. In the future, AI promises substantial advancements in automated real-time endoscopic guidance, personalized risk assessment, and optimized biopsy decision making. As it evolves, it would substantially impact not only early diagnosis and prognosis, but also the cost-effectiveness of managing upper GI diseases, ultimately leading to improved patient outcomes and more efficient health care delivery.

Artificial intelligence in endoscopic diagnosis of esophageal squamous cell carcinoma and precancerous lesions

Esophageal squamous cell carcinoma (ESCC) poses a significant global health challenge, necessitating early detection, timely diagnosis, and prompt treatment to improve patient outcomes. Endoscopic examination plays a pivotal role in this regard. However, despite the availability of various endoscopic techniques, certain limitations can result in missed or misdiagnosed ESCCs. Currently, artificial intelligence (AI)-assisted endoscopic diagnosis has made significant strides in addressing these limitations and improving the diagnosis of ESCC and precancerous lesions. In this review, we provide an overview of the current state of AI applications for endoscopic diagnosis of ESCC and precancerous lesions in aspects including lesion characterization, margin delineation, invasion depth estimation, and microvascular subtype classification. Furthermore, we offer insights into the future direction of this field, highlighting potential advancements that can lead to more accurate diagnoses and ultimately better prognoses for patients.

Optimizing AI models to predict esophageal squamous cell carcinoma risk by incorporating small datasets of soft palate images

There is a currently an unmet need for non-invasive methods to predict the risk of esophageal squamous cell carcinoma (ESCC). Previously, we found that specific soft palate morphologies are strongly associated with increased ESCC risk. However, there is currently no artificial intelligence (AI) system that utilizes oral images for ESCC risk assessment. Here, we evaluated three AI models and three fine-tuning approaches with regard to their ESCC predictive power. Our dataset contained 539 cases, which were subdivided into 221 high-risk cases (2491 images) and 318 non-high-risk cases (2524 images). We used 480 cases (4295 images) for the training dataset, and the rest for validation. The Bilinear convolutional neural network (CNN) model (especially when pre-trained on fractal images) demonstrated diagnostic precision that was comparable to or better than other models for distinguishing between high-risk and non-high-risk groups. In addition, when tested with a small number of images containing soft palate data, the model showed high precision: the best AUC model had 0.91 (sensitivity 0.86, specificity 0.79). This study presents a significant advance in the development of an AI-based non-invasive screening tool for the identification of high-risk ESCC patients. The approach may be particularly suitable for institutes with limited medical imaging resources.

Computer-aided endoscopic diagnostic system modified with hyperspectral imaging for the classification of esophageal neoplasms

INTRODUCTION

The early detection of esophageal cancer is crucial to enhancing patient survival rates, and endoscopy remains the gold standard for identifying esophageal neoplasms. Despite this fact, accurately diagnosing superficial esophageal neoplasms poses a challenge, even for seasoned endoscopists. Recent advancements in computer-aided diagnostic systems, empowered by artificial intelligence (AI), have shown promising results in elevating the diagnostic precision for early-stage esophageal cancer.

METHODS

In this study, we expanded upon traditional red-green-blue (RGB) imaging by integrating the YOLO neural network algorithm with hyperspectral imaging (HSI) to evaluate the diagnostic efficacy of this innovative AI system for superficial esophageal neoplasms. A total of 1836 endoscopic images were utilized for model training, which included 858 white-light imaging (WLI) and 978 narrow-band imaging (NBI) samples. These images were categorized into three groups, namely, normal esophagus, esophageal squamous dysplasia, and esophageal squamous cell carcinoma (SCC).

RESULTS

An additional set comprising 257 WLI and 267 NBI images served as the validation dataset to assess diagnostic accuracy. Within the RGB dataset, the diagnostic accuracies of the WLI and NBI systems for classifying images into normal, dysplasia, and SCC categories were 0.83 and 0.82, respectively. Conversely, the HSI dataset yielded higher diagnostic accuracies for the WLI and NBI systems, with scores of 0.90 and 0.89, respectively.

CONCLUSION

The HSI dataset outperformed the RGB dataset, demonstrating an overall diagnostic accuracy improvement of 8%. Our findings underscored the advantageous impact of incorporating the HSI dataset in model training. Furthermore, the application of HSI in AI-driven image recognition algorithms significantly enhanced the diagnostic accuracy for early esophageal cancer.

Surveillance for metachronous cancers after endoscopic resection of esophageal squamous cell carcinoma

The literature pertaining to surveillance following treatment for esophageal squamous cell carcinoma (SCC) was reviewed and summarized, encompassing the current status and future perspectives. Analysis of the standardized mortality and incidence ratios for these cancers indicates an elevated risk of cancer in the oral cavity, pharynx, larynx, and lungs among patients with esophageal SCC compared to the general population. To enhance the efficacy of surveillance for these metachronous cancers, risk stratification is needed. Various factors, including multiple Lugol-voiding lesions, multiple foci of dilated vascular areas, young age, and high mean corpuscular volume, have been identified as predictors of metachronous SCCs. Current practice involves stratifying the risk of metachronous esophageal and head/neck SCCs based on the presence of multiple Lugol-voiding lesions. Endoscopic surveillance, scheduled 6-12 months post-endoscopic resection, has demonstrated effectiveness, with over 90% of metachronous esophageal SCCs treatable through minimally invasive modalities. Narrow-band imaging emerges as the preferred surveillance method for esophageal and head/neck SCC based on comparative studies of various imaging techniques. Innovative approaches, such as artificial intelligence-assisted detection systems and radiofrequency ablation of high-risk background mucosa, may improve outcomes in patients following endoscopic resection.

Endoscopic Artificial Intelligence for Image Analysis in Gastrointestinal Neoplasms

BACKGROUND

Artificial intelligence (AI) using deep learning systems has recently been utilized in various medical fields. In the field of gastroenterology, AI is primarily implemented in image recognition and utilized in the realm of gastrointestinal (GI) endoscopy. In GI endoscopy, computer-aided detection/diagnosis (CAD) systems assist endoscopists in GI neoplasm detection or differentiation of cancerous or noncancerous lesions. Several AI systems for colorectal polyps have already been applied in colonoscopy clinical practices. In esophagogastroduodenoscopy, a few CAD systems for upper GI neoplasms have been launched in Asian countries. The usefulness of these CAD systems in GI endoscopy has been gradually elucidated.

SUMMARY

In this review, we outline recent articles on several studies of endoscopic AI systems for GI neoplasms, focusing on esophageal squamous cell carcinoma (ESCC), esophageal adenocarcinoma (EAC), gastric cancer (GC), and colorectal polyps. In ESCC and EAC, computer-aided detection (CADe) systems were mainly developed, and a recent meta-analysis study showed sensitivities of 91.2% and 93.1% and specificities of 80% and 86.9%, respectively. In GC, a recent meta-analysis study on CADe systems demonstrated that their sensitivity and specificity were as high as 90%. A randomized controlled trial (RCT) also showed that the use of the CADe system reduced the miss rate. Regarding computer-aided diagnosis (CADx) systems for GC, although RCTs have not yet been conducted, most studies have demonstrated expert-level performance. In colorectal polyps, multiple RCTs have shown the usefulness of the CADe system for improving the polyp detection rate, and several CADx systems have been shown to have high accuracy in colorectal polyp differentiation.

KEY MESSAGES

Most analyses of endoscopic AI systems suggested that their performance was better than that of nonexpert endoscopists and equivalent to that of expert endoscopists. Thus, endoscopic AI systems may be useful for reducing the risk of overlooking lesions and improving the diagnostic ability of endoscopists.

The role of artificial intelligence in the endoscopic diagnosis of esophageal cancer: a systematic review and meta-analysis

Early detection of esophageal cancer is limited by accurate endoscopic diagnosis of subtle macroscopic lesions. Endoscopic interpretation is subject to expertise, diagnostic skill, and thus human error. Artificial intelligence (AI) in endoscopy is increasingly bridging this gap. This systematic review and meta-analysis consolidate the evidence on the use of AI in the endoscopic diagnosis of esophageal cancer. The systematic review was carried out using Pubmed, MEDLINE and Ovid EMBASE databases and articles on the role of AI in the endoscopic diagnosis of esophageal cancer management were included. A meta-analysis was also performed. Fourteen studies (1590 patients) assessed the use of AI in endoscopic diagnosis of esophageal squamous cell carcinoma-the pooled sensitivity and specificity were 91.2% (84.3-95.2%) and 80% (64.3-89.9%). Nine studies (478 patients) assessed AI capabilities of diagnosing esophageal adenocarcinoma with the pooled sensitivity and specificity of 93.1% (86.8-96.4) and 86.9% (81.7-90.7). The remaining studies formed the qualitative summary. AI technology, as an adjunct to endoscopy, can assist in accurate, early detection of esophageal malignancy. It has shown superior results to endoscopists alone in identifying early cancer and assessing depth of tumor invasion, with the added benefit of not requiring a specialized skill set. Despite promising results, the application in real-time endoscopy is limited, and further multicenter trials are required to accurately assess its use in routine practice.

Application of convolutional neural network-based endoscopic imaging in esophageal cancer or high-grade dysplasia: A systematic review and meta-analysis

BACKGROUND

Esophageal cancer is the seventh-most common cancer type worldwide, accounting for 5% of death from malignancy. Development of novel diagnostic techniques has facilitated screening, early detection, and improved prognosis. Convolutional neural network (CNN)-based image analysis promises great potential for diagnosing and determining the prognosis of esophageal cancer, enabling even early detection of dysplasia.

AIM

To conduct a meta-analysis of the diagnostic accuracy of CNN models for the diagnosis of esophageal cancer and high-grade dysplasia (HGD).

METHODS

PubMed, EMBASE, Web of Science and Cochrane Library databases were searched for articles published up to November 30, 2022. We evaluated the diagnostic accuracy of using the CNN model with still image-based analysis and with video-based analysis for esophageal cancer or HGD, as well as for the invasion depth of esophageal cancer. The pooled sensitivity, pooled specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR), diagnostic odds ratio (DOR) and area under the curve (AUC) were estimated, together with the 95% confidence intervals (CI). A bivariate method and hierarchical summary receiver operating characteristic method were used to calculate the diagnostic test accuracy of the CNN model. Meta-regression and subgroup analyses were used to identify sources of heterogeneity.

RESULTS

A total of 28 studies were included in this systematic review and meta-analysis. Using still image-based analysis for the diagnosis of esophageal cancer or HGD provided a pooled sensitivity of 0.95 (95%CI: 0.92-0.97), pooled specificity of 0.92 (0.89-0.94), PLR of 11.5 (8.3-16.0), NLR of 0.06 (0.04-0.09), DOR of 205 (115-365), and AUC of 0.98 (0.96-0.99). When video-based analysis was used, a pooled sensitivity of 0.85 (0.77-0.91), pooled specificity of 0.73 (0.59-0.83), PLR of 3.1 (1.9-5.0), NLR of 0.20 (0.12-0.34), DOR of 15 (6-38) and AUC of 0.87 (0.84-0.90) were found. Prediction of invasion depth resulted in a pooled sensitivity of 0.90 (0.87-0.92), pooled specificity of 0.83 (95%CI: 0.76-0.88), PLR of 7.8 (1.9-32.0), NLR of 0.10 (0.41-0.25), DOR of 118 (11-1305), and AUC of 0.95 (0.92-0.96).

CONCLUSION

CNN-based image analysis in diagnosing esophageal cancer and HGD is an excellent diagnostic method with high sensitivity and specificity that merits further investigation in large, multicenter clinical trials.

A single-center prospective study evaluating the usefulness of artificial intelligence for the diagnosis of esophageal squamous cell carcinoma in a real-time setting

BACKGROUND

Several pre-clinical studies have reported the usefulness of artificial intelligence (AI) systems in the diagnosis of esophageal squamous cell carcinoma (ESCC). We conducted this study to evaluate the usefulness of an AI system for real-time diagnosis of ESCC in a clinical setting.

METHODS

This study followed a single-center prospective single-arm non-inferiority design. Patients at high risk for ESCC were recruited and real-time diagnosis by the AI system was compared with that of endoscopists for lesions suspected to be ESCC. The primary outcomes were the diagnostic accuracy of the AI system and endoscopists. The secondary outcomes were sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and adverse events.

RESULTS

A total of 237 lesions were evaluated. The accuracy, sensitivity, and specificity of the AI system were 80.6%, 68.2%, and 83.4%, respectively. The accuracy, sensitivity, and specificity of endoscopists were 85.7%, 61.4%, and 91.2%, respectively. The difference between the accuracy of the AI system and that of the endoscopists was - 5.1%, and the lower limit of the 90% confidence interval was less than the non-inferiority margin.

CONCLUSIONS

The non-inferiority of the AI system in comparison with endoscopists in the real-time diagnosis of ESCC in a clinical setting was not proven.

TRIAL REGISTRATION

Japan Registry of Clinical Trials (jRCTs052200015, 18/05/2020).

The current state of artificial intelligence in endoscopic diagnosis of early esophageal squamous cell carcinoma

Esophageal squamous cell carcinoma (ESCC) is a common malignant tumor of the digestive tract. The most effective method of reducing the disease burden in areas with a high incidence of esophageal cancer is to prevent the disease from developing into invasive cancer through screening. Endoscopic screening is key for the early diagnosis and treatment of ESCC. However, due to the uneven professional level of endoscopists, there are still many missed cases because of failure to recognize lesions. In recent years, along with remarkable progress in medical imaging and video evaluation technology based on deep machine learning, the development of artificial intelligence (AI) is expected to provide new auxiliary methods of endoscopic diagnosis and the treatment of early ESCC. The convolution neural network (CNN) in the deep learning model extracts the key features of the input image data using continuous convolution layers and then classifies images through full-layer connections. The CNN is widely used in medical image classification, and greatly improves the accuracy of endoscopic image classification. This review focuses on the AI-assisted diagnosis of early ESCC and prediction of early ESCC invasion depth under multiple imaging modalities. The excellent image recognition ability of AI is suitable for the detection and diagnosis of ESCC and can reduce missed diagnoses and help endoscopists better complete endoscopic examinations. However, the selective bias used in the training dataset of the AI system affects its general utility.

Artificial intelligence for detecting and delineating the extent of superficial esophageal squamous cell carcinoma and precancerous lesions under narrow-band imaging (with video)

BACKGROUND AND AIMS

Although narrow-band imaging (NBI) is a useful modality for detecting and delineating esophageal squamous cell carcinoma (ESCC), there is a risk of incorrectly determining the margins of some lesions even with NBI. This study aimed to develop an artificial intelligence (AI) system for detecting superficial ESCC and precancerous lesions and delineating the extent of lesions under NBI.

METHODS

Nonmagnified NBI images from 4 hospitals were collected and annotated. Internal and external image test datasets were used to evaluate the detection and delineation performance of the system. The delineation performance of the system was compared with that of endoscopists. Furthermore, the system was directly integrated into the endoscopy equipment, and its real-time diagnostic capability was prospectively estimated.

RESULTS

The system was trained and tested using 10,047 still images and 140 videos from 1112 patients and 1183 lesions. In the image testing, the accuracy of the system in detecting lesions in internal and external tests was 92.4% and 89.9%, respectively. The accuracy of the system in delineating extents in internal and external tests was 88.9% and 87.0%, respectively. The delineation performance of the system was superior to that of junior endoscopists and similar to that of senior endoscopists. In the prospective clinical evaluation, the system exhibited satisfactory performance, with an accuracy of 91.4% in detecting lesions and an accuracy of 85.9% in delineating extents.

CONCLUSIONS

The proposed AI system could accurately detect superficial ESCC and precancerous lesions and delineate the extent of lesions under NBI.

Unassisted Clinicians Versus Deep Learning-Assisted Clinicians in Image-Based Cancer Diagnostics: Systematic Review With Meta-analysis

BACKGROUND

A number of publications have demonstrated that deep learning (DL) algorithms matched or outperformed clinicians in image-based cancer diagnostics, but these algorithms are frequently considered as opponents rather than partners. Despite the clinicians-in-the-loop DL approach having great potential, no study has systematically quantified the diagnostic accuracy of clinicians with and without the assistance of DL in image-based cancer identification.

OBJECTIVE

We systematically quantified the diagnostic accuracy of clinicians with and without the assistance of DL in image-based cancer identification.

METHODS

PubMed, Embase, IEEEXplore, and the Cochrane Library were searched for studies published between January 1, 2012, and December 7, 2021. Any type of study design was permitted that focused on comparing unassisted clinicians and DL-assisted clinicians in cancer identification using medical imaging. Studies using medical waveform-data graphics material and those investigating image segmentation rather than classification were excluded. Studies providing binary diagnostic accuracy data and contingency tables were included for further meta-analysis. Two subgroups were defined and analyzed, including cancer type and imaging modality.

RESULTS

In total, 9796 studies were identified, of which 48 were deemed eligible for systematic review. Twenty-five of these studies made comparisons between unassisted clinicians and DL-assisted clinicians and provided sufficient data for statistical synthesis. We found a pooled sensitivity of 83% (95% CI 80%-86%) for unassisted clinicians and 88% (95% CI 86%-90%) for DL-assisted clinicians. Pooled specificity was 86% (95% CI 83%-88%) for unassisted clinicians and 88% (95% CI 85%-90%) for DL-assisted clinicians. The pooled sensitivity and specificity values for DL-assisted clinicians were higher than for unassisted clinicians, at ratios of 1.07 (95% CI 1.05-1.09) and 1.03 (95% CI 1.02-1.05), respectively. Similar diagnostic performance by DL-assisted clinicians was also observed across the predefined subgroups.

CONCLUSIONS

The diagnostic performance of DL-assisted clinicians appears better than unassisted clinicians in image-based cancer identification. However, caution should be exercised, because the evidence provided in the reviewed studies does not cover all the minutiae involved in real-world clinical practice. Combining qualitative insights from clinical practice with data-science approaches may improve DL-assisted practice, although further research is required.

TRIAL REGISTRATION

PROSPERO CRD42021281372; https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=281372.

Cooperation between artificial intelligence and endoscopists for diagnosing invasion depth of early gastric cancer

BACKGROUND AND STUDY AIMS

The diagnostic ability of endoscopists to determine invasion depth of early gastric cancer is not favorable. We designed an artificial intelligence (AI) classifier for differentiating intramucosal and submucosal gastric cancers and examined it to establish a diagnostic method based on cooperation between AI and endoscopists.

PATIENTS AND METHODS

We prepared 500 training images using cases of mainly depressed-type early gastric cancer from 250 intramucosal cancers and 250 submucosal cancers. We also prepared 200 test images each of 100 cancers from another institution. We designed an AI classifier to differentiate between intramucosal and submucosal cancers by deep learning. We examined the performance of the AI classifier and the majority vote of the endoscopists as high confidence and low confidence diagnostic probability, respectively, and cooperatively combined them to establish a diagnostic method providing high accuracy.

RESULTS

Internal evaluation of the training images showed that accuracy, sensitivity, specificity, and F1 measure by the AI classifier were 77%, 76%, 78%, and 0.768, and those of the majority vote of the endoscopists were 72.6%, 53.6%, 91.6%, and 0.662, respectively. A diagnostic method based on cooperation between AI and the endoscopists showed that the respective values were 78.0%, 76.0%, 80.0%, and 0.776 for the test images. The value of F1 measure was especially higher than those by AI or the endoscopists alone.

CONCLUSIONS

Cooperation between AI and endoscopists improved the diagnostic ability to determine invasion depth of early gastric cancer.

Deep Learning for the Diagnosis of Esophageal Cancer in Endoscopic Images: A Systematic Review and Meta-Analysis

Esophageal cancer, one of the most common cancers with a poor prognosis, is the sixth leading cause of cancer-related mortality worldwide. Early and accurate diagnosis of esophageal cancer, thus, plays a vital role in choosing the appropriate treatment plan for patients and increasing their survival rate. However, an accurate diagnosis of esophageal cancer requires substantial expertise and experience. Nowadays, the deep learning (DL) model for the diagnosis of esophageal cancer has shown promising performance. Therefore, we conducted an updated meta-analysis to determine the diagnostic accuracy of the DL model for the diagnosis of esophageal cancer. A search of PubMed, EMBASE, Scopus, and Web of Science, between 1 January 2012 and 1 August 2022, was conducted to identify potential studies evaluating the diagnostic performance of the DL model for esophageal cancer using endoscopic images. The study was performed in accordance with PRISMA guidelines. Two reviewers independently assessed potential studies for inclusion and extracted data from retrieved studies. Methodological quality was assessed by using the QUADAS-2 guidelines. The pooled accuracy, sensitivity, specificity, positive and negative predictive value, and the area under the receiver operating curve (AUROC) were calculated using a random effect model. A total of 28 potential studies involving a total of 703,006 images were included. The pooled accuracy, sensitivity, specificity, and positive and negative predictive value of DL for the diagnosis of esophageal cancer were 92.90%, 93.80%, 91.73%, 93.62%, and 91.97%, respectively. The pooled AUROC of DL for the diagnosis of esophageal cancer was 0.96. Furthermore, there was no publication bias among the studies. The findings of our study show that the DL model has great potential to accurately and quickly diagnose esophageal cancer. However, most studies developed their model using endoscopic data from the Asian population. Therefore, we recommend further validation through studies of other populations as well.

Artificial intelligence for diagnosing gastric lesions under white-light endoscopy

BACKGROUND

The ability of endoscopists to identify gastric lesions is uneven. Even experienced endoscopists may miss or misdiagnose lesions due to heavy workload or fatigue or subtle changes in lesions under white-light endoscopy (WLE). This study aimed to develop an artificial intelligence (AI) system that could diagnose six common gastric lesions under WLE and to explore its role in assisting endoscopists in diagnosis.

METHODS

Images of early gastric cancer, advanced gastric cancer, submucosal tumor, polyp, peptic ulcer, erosion, and lesion-free gastric mucosa were retrospectively collected to train and test the system. The performance of the system was compared with that of 12 endoscopists. The performance of endoscopists with or without referring to the system was also evaluated.

RESULTS

A total of 29,809 images from 8947 patients and 1579 images from 496 patients were used to train and test the system, respectively. For per-lesion analysis, the overall accuracy of the system was 85.7%, which was comparable to that of senior endoscopists (85.1%, P = 0.729) and significantly higher than that of junior endoscopists (78.8%, P < 0.001). With system assistance, the overall accuracies of senior and junior endoscopists increased to 89.3% (4.2%, P < 0.001) and 86.2% (7.4%, P < 0.001), respectively. Senior and junior endoscopists achieved varying degrees of improvement in the diagnostic performance of other types of lesions except for polyp. The diagnostic times of senior (3.8 vs 3.2 s per image, P = 0.500) and junior endoscopists (6.2 vs 4.6 s per image, P = 0.144) assisted by the system were both slightly shortened, despite no significant differences.

CONCLUSIONS

The proposed AI system could be applied as an auxiliary tool to reduce the workload of endoscopists and improve the diagnostic accuracy of gastric lesions.

Utility of an artificial intelligence system for classification of esophageal lesions when simulating its clinical use

Previous reports have shown favorable performance of artificial intelligence (AI) systems for diagnosing esophageal squamous cell carcinoma (ESCC) compared with endoscopists. However, these findings don't reflect performance in clinical situations, as endoscopists classify lesions based on both magnified and non-magnified videos, while AI systems often use only a few magnified narrow band imaging (NBI) still images. We evaluated the performance of the AI system in simulated clinical situations. We used 25,048 images from 1433 superficial ESCC and 4746 images from 410 noncancerous esophagi to construct our AI system. For the validation dataset, we took NBI videos of suspected superficial ESCCs. The AI system diagnosis used one magnified still image taken from each video, while 19 endoscopists used whole videos. We used 147 videos and still images including 83 superficial ESCC and 64 non-ESCC lesions. The accuracy, sensitivity and specificity for the classification of ESCC were, respectively, 80.9% [95% CI 73.6-87.0], 85.5% [76.1-92.3], and 75.0% [62.6-85.0] for the AI system and 69.2% [66.4-72.1], 67.5% [61.4-73.6], and 71.5% [61.9-81.0] for the endoscopists. The AI system correctly classified all ESCCs invading the muscularis mucosa or submucosa and 96.8% of lesions ≥ 20 mm, whereas even the experts diagnosed some of them as non-ESCCs. Our AI system showed higher accuracy for classifying ESCC and non-ESCC than endoscopists. It may provide valuable diagnostic support to endoscopists.

Implementation of artificial intelligence in upper gastrointestinal endoscopy

The application of artificial intelligence (AI) using deep learning has significantly expanded in the field of esophagogastric endoscopy. Recent studies have shown promising results in detecting and differentiating early gastric cancer using AI tools built using white light, magnified, or image-enhanced endoscopic images. Some studies have reported the use of AI tools to predict the depth of early gastric cancer based on endoscopic images. Similarly, studies based on using AI for detecting early esophageal cancer have also been reported, with an accuracy comparable to that of endoscopy specialists. Moreover, an AI system, developed to diagnose pharyngeal cancer, has shown promising performance with high sensitivity. These reports suggest that, if introduced for regular use in clinical settings, AI systems can significantly reduce the burden on physicians. This review summarizes the current status of AI applications in the upper gastrointestinal tract and presents directions for clinical practice implementation and future research.

Artificial Intelligence and Deep Learning for Upper Gastrointestinal Neoplasia

Upper gastrointestinal (GI) neoplasia account for 35% of GI cancers and 1.5 million cancer-related deaths every year. Despite its efficacy in preventing cancer mortality, diagnostic upper GI endoscopy is affected by a substantial miss rate of neoplastic lesions due to failure to recognize a visible lesion or imperfect navigation. This may be offset by the real-time application of artificial intelligence (AI) for detection (computer-aided detection [CADe]) and characterization (computer-aided diagnosis [CADx]) of upper GI neoplasia. Stand-alone performance of CADe for esophageal squamous cell neoplasia, Barrett's esophagus-related neoplasia, and gastric cancer showed promising accuracy, sensitivity ranging between 83% and 93%. However, incorporation of CADe/CADx in clinical practice depends on several factors, such as possible bias in the training or validation phases of these algorithms, its interaction with human endoscopists, and clinical implications of false-positive results. The aim of this review is to guide the clinician across the multiple steps of AI development in clinical practice.

Systematic review with meta-analysis: artificial intelligence in the diagnosis of oesophageal diseases

BACKGROUND

Artificial intelligence (AI) has recently been applied to endoscopy and questionnaires for the evaluation of oesophageal diseases (ODs).

AIM

We performed a systematic review with meta-analysis to evaluate the performance of AI in the diagnosis of malignant and benign OD.

METHODS

We searched MEDLINE, EMBASE, EMBASE Classic and the Cochrane Library. A bivariate random-effect model was used to calculate pooled diagnostic efficacy of AI models and endoscopists. The reference tests were histology for neoplasms and the clinical and instrumental diagnosis for gastro-oesophageal reflux disease (GERD). The pooled area under the summary receiver operating characteristic (AUROC), sensitivity, specificity, positive and negative likelihood ratio (PLR and NLR) and diagnostic odds ratio (DOR) were estimated.

RESULTS

For the diagnosis of Barrett's neoplasia, AI had AUROC of 0.90, sensitivity 0.89, specificity 0.86, PLR 6.50, NLR 0.13 and DOR 50.53. AI models' performance was comparable with that of endoscopists (P = 0.35). For the diagnosis of oesophageal squamous cell carcinoma, the AUROC, sensitivity, specificity, PLR, NLR and DOR were 0.97, 0.95, 0.92, 12.65, 0.05 and DOR 258.36, respectively. In this task, AI performed better than endoscopists although without statistically significant differences. In the detection of abnormal intrapapillary capillary loops, the performance of AI was: AUROC 0.98, sensitivity 0.94, specificity 0.94, PLR 14.75, NLR 0.07 and DOR 225.83. For the diagnosis of GERD based on questionnaires, the AUROC, sensitivity, specificity, PLR, NLR and DOR were 0.99, 0.97, 0.97, 38.26, 0.03 and 1159.6, respectively.

CONCLUSIONS

AI demonstrated high performance in the clinical and endoscopic diagnosis of OD.

Artificial intelligence for detecting superficial esophageal squamous cell carcinoma under multiple endoscopic imaging modalities: A multicenter study

BACKGROUND AND AIM

Diagnosis of esophageal squamous cell carcinoma (ESCC) is complicated and requires substantial expertise and experience. This study aimed to develop an artificial intelligence (AI) system for detecting superficial ESCC under multiple endoscopic imaging modalities.

METHODS

Endoscopic images were retrospectively collected from West China Hospital, Sichuan University as a training dataset and an independent internal validation dataset. Images from other four hospitals were used as an external validation dataset. The AI system was compared with 11 experienced endoscopists. Furthermore, videos were collected to assess the performance of the AI system.

RESULTS

A total of 53 933 images from 2621 patients and 142 videos from 19 patients were used to develop and validate the AI system. In the internal and external validation datasets, the performance of the AI system under all or different endoscopic imaging modalities was satisfactory, with sensitivity of 92.5-99.7%, specificity of 78.5-89.0%, and area under the receiver operating characteristic curves of 0.906-0.989. The AI system achieved comparable performance with experienced endoscopists. Regarding superficial ESCC confined to the epithelium, the AI system was more sensitive than experienced endoscopists on white-light imaging (90.8% vs 82.5%, P = 0.022). Moreover, the AI system exhibited good performance in videos, with sensitivity of 89.5-100% and specificity of 73.7-89.5%.

CONCLUSIONS

We developed an AI system that showed comparable performance with experienced endoscopists in detecting superficial ESCC under multiple endoscopic imaging modalities and might provide valuable support for inexperienced endoscopists, despite requiring further evaluation.

Artificial Intelligence in the Diagnosis of Upper Gastrointestinal Diseases

Artificial intelligence (AI) has enormous potential to support clinical routine workflows and therefore is gaining increasing popularity among medical professionals. In the field of gastroenterology, investigations on AI and computer-aided diagnosis (CAD) systems have mainly focused on the lower gastrointestinal (GI) tract. However, numerous CAD tools have been tested also in upper GI disorders showing encouraging results. The main application of AI in the upper GI tract is endoscopy; however, the need to analyze increasing loads of numerical and categorical data in short times has pushed researchers to investigate applications of AI systems in other upper GI settings, including gastroesophageal reflux disease, eosinophilic esophagitis, and motility disorders. AI and CAD systems will be increasingly incorporated into daily clinical practice in the coming years, thus at least basic notions will be soon required among physicians. For noninsiders, the working principles and potential of AI may be as fascinating as obscure. Accordingly, we reviewed systematic reviews, meta-analyses, randomized controlled trials, and original research articles regarding the performance of AI in the diagnosis of both malignant and benign esophageal and gastric diseases, also discussing essential characteristics of AI.

Artificial intelligence and early esophageal cancer

The development of esophageal cancer (EC) from early to advanced stage results in a high mortality rate and poor prognosis. Advanced EC not only poses a serious threat to the life and health of patients but also places a heavy economic burden on their families and society. Endoscopy is of great value for the diagnosis of EC, especially in the screening of Barrett’s esophagus and early EC. However, at present, endoscopy has a low diagnostic rate for early tumors. In recent years, artificial intelligence (AI) has made remarkable progress in the diagnosis of digestive system tumors, providing a new model for clinicians to diagnose and treat these tumors. In this review, we aim to provide a comprehensive overview of how AI can help doctors diagnose early EC and precancerous lesions and make clinical decisions based on the predicted results. We analyze and summarize the recent research on AI and early EC. We find that based on deep learning (DL) and convolutional neural network methods, the current computer-aided diagnosis system has gradually developed from in vitro image analysis to real-time detection and diagnosis. Based on powerful computing and DL capabilities, the diagnostic accuracy of AI is close to or better than that of endoscopy specialists. We also analyze the shortcomings in the current AI research and corresponding improvement strategies. We believe that the application of AI-assisted endoscopy in the diagnosis of early EC and precancerous lesions will become possible after the further advancement of AI-related research.