Artificial Intelligence for the Prediction of Helicobacter Pylori Infection in Endoscopic Images: Systematic Review and Meta-Analysis Of Diagnostic Test Accuracy

Abstract

Background Helicobacter pylori plays a central role in the development of gastric cancer, and prediction of H pylori infection by visual inspection of the gastric mucosa is an important function of endoscopy. However, there are currently no established methods of optical diagnosis of H pylori infection using endoscopic images. Definitive diagnosis requires endoscopic biopsy. Artificial intelligence (AI) has been increasingly adopted in clinical practice, especially for image recognition and classification. Objective This study aimed to evaluate the diagnostic test accuracy of AI for the prediction of H pylori infection using endoscopic images. Methods Two independent evaluators searched core databases. The inclusion criteria included studies with endoscopic images of H pylori infection and with application of AI for the prediction of H pylori infection presenting diagnostic performance. Systematic review and diagnostic test accuracy meta-analysis were performed. Results Ultimately, 8 studies were identified. Pooled sensitivity, specificity, diagnostic odds ratio, and area under the curve of AI for the prediction of H pylori infection were 0.87 (95% CI 0.72-0.94), 0.86 (95% CI 0.77-0.92), 40 (95% CI 15-112), and 0.92 (95% CI 0.90-0.94), respectively, in the 1719 patients (385 patients with H pylori infection vs 1334 controls). Meta-regression showed methodological quality and included the number of patients in each study for the purpose of heterogeneity. There was no evidence of publication bias. The accuracy of the AI algorithm reached 82% for discrimination between noninfected images and posteradication images. Conclusions An AI algorithm is a reliable tool for endoscopic diagnosis of H pylori infection. The limitations of lacking external validation performance and being conducted only in Asia should be overcome. Trial Registration PROSPERO CRD42020175957; https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=175957

What Is New in Helicobacter pylori Diagnosis. An Overview

“…This method demonstrated to improve the accuracy and productivity of endoscopic examination, with respect to WLI, with a sensitivity for BLI-bright and for LCI of 96.7% [ 33 ]. In a recent meta-analysis, the artificial intelligence algorithm demonstrated to be an accurate tool for the prediction of H. pylori infection during endoscopic procedures, although, the authors concluded that the real application needs to be evaluated in clinical studies [ 34 ]. Figure 1 shows some gastric mucosa features associated with H. pylori infection in different studies.…”

Use of Endoscopic Images in the Prediction of Submucosal Invasion of Gastric Neoplasms: Automated Deep Learning Model Development and Usability Study

“…This highlights not only the importance of robust answers provided by the AI but also the importance of the baseline level of experience of AI users. Therefore, the conclusion from the previous study [6] that inexperienced doctors would benefit the most from AI support was not reproduced in this study. Rather, endoscopists having at least a certain level of expertise benefited from AI support.…”

“…Artificial intelligence (AI) using deep learning (DL), which mimics the intellectual function of humans, has been increasingly adopted in clinical medicine, especially for cognitive function in computer vision [1][2][3], including automated image recognition, classification, and segmentation tasks [4][5][6]. Application of AI to endoscopic examination is noninvasive and can further help in detecting hidden or hard-to-detect lesions in real time.…”

Diagnosis of Helicobacter pylori Infection and Recent Advances

“…One meta-analysis including 8 studies and 1719 patients (385 patients with H. pylori infection vs. 1334 controls) diagnosed by WLI, BLI, or LCI reported that the sensitivity, specificity, DOR, and AUC of AI for the prediction of H. pylori infection were 0.87, 0.86, 40, and 0.92, respectively. The accuracy of the AI algorithm reached 82% for discrimination between noninfected images and posteradication images [87]. Regarding WLI, a DCNN model trained and verified by WLI of gastric antrum showed a power in diagnosing atrophic gastritis with 94% accuracy, 0.95 sensitivity, and 0.94 specificity, which were higher than those of experts [88], and AI diagnosis could be done in a considerably shorter time less than 200 s [89,90].…”