Surveillance using trimodal imaging endoscopy after endoscopic submucosal dissection for superficial gastric neoplasia

The effect of incorporating domain knowledge with deep learning in identifying benign and malignant gastric whitish lesions: A retrospective study

BACKGROUND AND AIM

Early whitish gastric neoplasms can be easily misdiagnosed; differential diagnosis of gastric whitish lesions remains a challenge. We aim to build a deep learning (DL) model to diagnose whitish gastric neoplasms and explore the effect of adding domain knowledge in model construction.

METHODS

We collected 4558 images from two institutions to train and test models. We first developed two sole DL models (1 and 2) using supervised and semi-supervised algorithms. Then we selected diagnosis-related features through literature research and developed feature-extraction models to determine features including boundary, surface, roundness, depression, and location. Then predictions of the five feature-extraction models and sole DL model were combined and inputted into seven machine-learning (ML) based fitting-diagnosis models. The optimal model was selected as ENDOANGEL-WD (whitish-diagnosis) and compared with endoscopists.

RESULTS

Sole DL 2 had higher sensitivity (83.12% vs 68.67%, Bonferroni adjusted P = 0.024) than sole DL 1. Adding domain knowledge, the decision tree performed best among the seven ML models, achieving higher specificity than DL 1 (84.38% vs 72.27%, Bonferroni adjusted P < 0.05) and higher accuracy than DL 2 (80.47%, Bonferroni adjusted P < 0.001) and was selected as ENDOANGEL-WD. ENDOANGEL-WD showed better accuracy compared with 10 endoscopists (75.70%, P < 0.001).

CONCLUSIONS

We developed a novel system ENDOANGEL-WD combining domain knowledge and traditional DL to detect gastric whitish neoplasms. Adding domain knowledge improved the performance of traditional DL, which provided a novel solution for establishing diagnostic models for other rare diseases potentially.

Diagnosis of advanced gastric cancer using image enhancement and autofluorescence imaging systems

INTRODUCTION

The indication of laparoscopic gastrectomy for advanced cases such as serosal invasion or peritoneal dissemination is still controversial. We report the clinical experiences of laparoscopic diagnosis with image enhancement and autofluorescence (AF) systems for laparoscopic gastric cancer surgery.

METHODS

The image enhancement system was introduced to 164 patients who underwent laparoscopic gastric cancer surgery as the AF group. The control group was defined as 165 cases who underwent bright light observation without the image enhancement system. The operative findings were recorded during surgery. All image findings were retrospectively compared with the pathological findings. Furthermore, the relationship between the image and pathological findings for peritoneal dissemination and subserosal invasion was evaluated.

RESULTS

This study included 329 patients, divided into the control group of 165 patients (50.2%) and the AF group of 164 patients (49.8%). There was no statistically significant difference in patient background between the two groups. In seven cases of peritoneal dissemination, AF was positive, and macroscopic abnormal vascularization was detected. The sensitivity, specificity, and positive and negative predictive values were 100%, 99.4%, 87.5%, and 100%, respectively. In 29 cases of serosal invasion, AF was positive, and macroscopic abnormal vascularization was detected. In the detection accuracy of subserosal invasion, the sensitivity, specificity, and positive and negative predictive values were 75.7%, 97.6%, 90.3%, and 93.2%, respectively.

CONCLUSION

The image enhancement with the AF system is useful for the laparoscopic intraoperative diagnosis of serosal invasion and peritoneal dissemination for advanced gastric cancer.

Endoscopic Advances for Gastric Neoplasia Detection

This article explores advances in endoscopic neoplasia detection with supporting clinical evidence and future aims. The ability to detect early gastric neoplastic lesions amenable to curative endoscopic submucosal dissection provides the opportunity to decrease gastric cancer mortality rates. Newer imaging techniques offer enhanced views of mucosal and microvascular structures and show promise in differentiating benign from malignant lesions and improving targeted biopsies. Conventional chromoendoscopy is well studied and validated. Narrow band imaging demonstrates superiority over magnified white light. Autofluorescence imaging, i-scan, flexible spectral imaging color enhancement, and bright image enhanced endoscopy show promise but insufficient evidence to change current clinical practice.

Detection of early stage gastric cancers in screening laser endoscopy using linked color imaging for patients with atrophic gastritis

BACKGROUND AND AIMS

Laser endoscopy involves blue laser imaging in bright mode (BLI-bright). Linked color imaging (LCI) is superior to white light imaging (WLI) for detecting gastric cancers. This study aimed to detect gastric cancers on screening endoscopy using not only WLI but also BLI-bright and LCI in patients with atrophic gastritis.

PATIENTS AND METHODS

A total of 500 patients with atrophic gastritis undergoing screening esophagogastroduodenoscopy were included. The gastric lumen was observed in the WLI mode, followed by the LCI and BLI-bright modes. When gastric neoplasms were suspected, the mode was changed to WLI, and we sprayed indigo carmine. Finally, biopsy specimens were taken for those lesions and pathological diagnosis was made. We compared the size, morphology, and color of gastric neoplasms found by the first WLI mode and those detected by only the LCI mode or BLI-bright mode.

RESULTS

We detected 16 gastric neoplasms (3.2%), of which 13 were early gastric cancers (EGCs) and three were gastric adenomas. Ten EGCs and two gastric adenomas (75%) were detected by the first WLI mode; three EGCs and one gastric adenoma (25%) were missed by the first WLI mode and were detected by the LCI mode or BLI-bright mode. All were less than 1 cm in diameter and were reddish. Mean diameter of the lesions was significantly less for LCI-detected or BLI-bright-detected lesions than for WLI-detected lesions (7.8 vs 21.2 mm).

CONCLUSIONS

Laser endoscopy is useful for detecting EGCs by LCI for patients with atrophic gastritis.

NBI utility in oncologic surgery: An organ by organ review

The main aims of the oncologic surgeon should be an early tumor diagnosis, complete surgical resection, and a careful post-treatment follow-up to ensure a prompt diagnosis of recurrence. Radiologic and endoscopic methods have been traditionally used for these purposes, but their accuracy might sometimes be suboptimal. Technological improvements could help the clinician during the diagnostic and therapeutic management of tumors. Narrow band imaging (NBI) belongs to optical image techniques, and uses light characteristics to enhance tissue vascularization. Because neoangiogenesis is a fundamental step during carcinogenesis, NBI could be useful in the diagnostic and therapeutic workup of tumors. Since its introduction in 2001, NBI use has rapidly spread in different oncologic specialties with clear advantages. There is an active interest in this topic as demonstrated by the thriving literature. It is unavoidable for clinicians to gain in-depth knowledge about the application of NBI to their specific field, losing the overall view on the topic. However, by looking at other fields of application, clinicians could find ideas to improve NBI use in their own specialty. The aim of this review is to summarize the existing literature on NBI use in oncology, with the aim of providing the state of the art: we present an overview on NBI fields of application, results, and possible future improvements in the different specialties.

Image-enhanced endoscopy for gastric preneoplastic conditions and neoplastic lesions: a systematic review and meta-analysis

BACKGROUND : Image-enhanced endoscopy (IEE) improves the accuracy of endoscopic diagnosis. We aimed to assess the value of IEE for gastric preneoplastic conditions and neoplastic lesions. METHODS : Medline and Embase were searched until December 2018. Studies allowing calculation of diagnostic measures were included. Risk of bias and applicability were assessed using QUADAS-2. Subgroup analysis was performed to explore heterogeneity. RESULTS : 44 studies met the inclusion criteria. For gastric intestinal metaplasia (GIM), narrow-band imaging (NBI) obtained a pooled sensitivity and specificity of 0.79 (95 %CI 0.72-0.85) and 0.91 (95 %CI 0.88-0.94) on per-patient basis; on per-biopsy basis, it was 0.84 (95 %CI 0.81-0.86) and 0.95 (95 %CI 0.94-0.96), respectively. Tubulovillous pattern was the most accurate marker to detect GIM and it was effectively assessed without high magnification. For dysplasia, NBI showed a pooled sensitivity and specificity of 0.87 (95 %CI 0.84-0.89) and 0.97 (95 %CI 0.97-0.98) on per-biopsy basis. The use of magnification improved the performance of NBI to characterize early gastric cancer (EGC), especially when the vessel plus surface (VS) classification was applied. Regarding other technologies, trimodal imaging also obtained a high accuracy for dysplasia (sensitivity 0.93 [95 %CI 0.85-0.98], specificity 0.98 [95 %CI 0.92-1.00]). For atrophic gastritis, no specific pattern was noted and none of the technologies reached good diagnostic yield. CONCLUSION : NBI is highly accurate for GIM and dysplasia. The presence of tubulovillous pattern and the VS classification seem to be useful to detect GIM and characterize EGC, respectively. These features should be used in current practice and to standardize endoscopic criteria for other technologies.

An Asian consensus on standards of diagnostic upper endoscopy for neoplasia

BACKGROUND

This is a consensus developed by a group of expert endoscopists aiming to standardise the preparation, process and endoscopic procedural steps for diagnosis of early upper gastrointestinal (GI) cancers.

METHOD

The Delphi method was used to develop consensus statements through identification of clinical questions on diagnostic endoscopy. Three consensus meetings were conducted to consolidate the statements and voting. We conducted a systematic literature search on evidence for each statement. The statements were presented in the second consensus meeting and revised according to comments. The final voting was conducted at the third consensus meeting on the level of evidence and agreement.

RESULTS

Risk stratification should be conducted before endoscopy and high risk endoscopic findings should raise an index of suspicion. The presence of premalignant mucosal changes should be documented and use of sedation is recommended to enhance detection of superficial upper GI neoplasms. The use of antispasmodics and mucolytics enhanced visualisation of the upper GI tract, and systematic endoscopic mapping should be conducted to improve detection. Sufficient examination time and structured training on diagnosis improves detection. Image enhanced endoscopy in addition to white light imaging improves detection of superficial upper GI cancer. Magnifying endoscopy with narrow-band imaging is recommended for characterisation of upper GI superficial neoplasms. Endoscopic characterisation can avoid unnecessary biopsy.

CONCLUSION

This consensus provides guidance for the performance of endoscopic diagnosis and characterisation for early gastric and oesophageal neoplasia based on the evidence. This will enhance the quality of endoscopic diagnosis and improve detection of early upper GI cancers.

Blue laser imaging-bright improves the real-time detection rate of early gastric cancer: a randomized controlled study

BACKGROUND AND AIMS

Blue laser imaging-bright (BLI-bright) has shown promise as a more useful tool for detection of early gastric cancer (EGC) than white-light imaging (WLI). However, the diagnostic performance of BLI-bright in the detection of EGC has not been investigated. We aimed to compare real-time detection rates of WLI with that of BLI-bright for EGC.

METHODS

This was a prospective, randomized, controlled study in 2 Japanese academic centers. We investigated 629 patients undergoing follow-up endoscopy for atrophic gastritis with intestinal metaplasia or surveillance after endoscopic resection of EGC. Patients were randomly assigned to receive primary WLI followed by BLI-bright or primary BLI-bright followed by WLI. The real-time detection rates of EGC were compared between primary WLI and primary BLI-bright.

RESULTS

There were 298 patients in each group. The real-time detection rate of EGC with primary BLI-bright was significantly greater than that with primary WLI (93.1% vs 50.0%; P = .001). Primary BLI-bright had a significantly greater ability to detect EGCs in patients with a history of endoscopic resection for EGC, no Helicobacter pylori infection in the stomach after eradication therapy, lesions with an open-type atrophic border, lesions in the lower third of the stomach, depressed-type lesions, small lesions measuring <10 mm and 10 to 20 mm in diameter, reddish lesions, well-differentiated adenocarcinomas, and lesions with a depth of invasion of T1a.

CONCLUSIONS

BLI-bright has a higher real-time detection rate for EGC than WLI. BLI-bright should be performed during surveillance endoscopy in patients at high risk for EGC. (Clinical trial registration number: UMIN000011324.).

Advanced endoscopic imaging in gastric neoplasia and preneoplasia

Conventional white light endoscopy remains the current standard in routine clinical practice for early detection of gastric cancer. However, it may not accurately diagnose preneoplastic gastric lesions. The technological advancements in the field of endoscopic imaging for gastric lesions are fast growing. This article reviews currently available advanced endoscopic imaging modalities, in particular chromoendoscopy, narrow band imaging and confocal laser endomicroscopy, and their corresponding evidence shown to improve diagnosis of preneoplastic gastric lesions. Raman spectrometry and polarimetry are also introduced as promising emerging technologies.