VENet: Variational energy network for gland segmentation of pathological images and early gastric cancer diagnosis of whole slide images

In Japan, the annual number of deaths from gastric cancer is approximately 50,000 and there has been no change over the last 50 years. So far, all efforts have been directed toward improving the detection of early gastric cancer by barium X-ray and endoscopy, since early cancer has a good prognosis, resulting in Japan having the best diagnostic capability for early gastric cancer worldwide. The 5-year survival rate of gastric cancer patients exceeds 60 % in Japan and is much higher than that in Europe and the US (20 %) because of this superior diagnosis of early gastric cancer. In February 2013, national health insurance coverage for Helicobacter pylori eradication therapy to treat H. pylori-associated chronic gastritis became available in Japan. H. pylori-associated gastritis leads to development of gastric and duodenal ulcers and gastric polyps. Therefore, providing treatment for gastritis is likely to substantially decrease the prevalence of both gastric and duodenal ulcers and polyps. Because treatment for H. pylori-associated gastritis, which leads to atrophic gastritis and gastric cancer, is now covered by health insurance in Japan, a strategy to eliminate gastric cancer-related deaths by taking advantage of this innovation was planned. According to this strategy, patients with gastritis will be investigated for H. pylori infection and those who are positive will receive eradication therapy followed by periodic surveillance. If this strategy is implemented, deaths from gastric cancer in Japan will decrease dramatically after 10–20 years.

Magnifying endoscopy for the diagnosis of early gastric cancer: Establishment of technique, diagnostic system, and scientific evidence from Japan

BackgroundGastric carcinoma is the second commonest cause of cancer deaths worldwide. Early detection and diagnosis of gastric cancer in the stomach is important for improving the prognosis of gastric cancer. This retrospective study was designed to investigate the value of magnifying narrow-band imaging (NBI) in the diagnosis of precancerous lesions and early gastric cancer.MethodsThis study included 122 patients who were diagnosed with early gastric cancer or precancerous gastric lesions by endoscopy. The patients underwent an examination with conventional endoscopy, magnifying NBI, and magnifying chromoendoscopy. Images resolution was evaluated, and the morphology, pit patterns and blood capillary forms of lesions were analyzed. The presence of gastric carcinoma and high grade intraepithelial neoplasia in the biopsy samples was considered as a positive pathological result, which is used to assess accuracy of endoscopic diagnosis.ResultsFor image resolution, magnifying NBI and magnifying chromoendoscopy were significantly superior to magnifying conventional endoscopy in morphology, pit pattern and blood capillary form (P < 0.01), and magnifying NBI was significantly superior to magnifying chromoendoscopy in blood capillary form (P < 0.01). IV, V1, and VI type of gastric pit pattern were detected in 14 cases, 43 cases, and 17 cases in patients with high grade intraepithelial neoplasia, respectively. V1 and VI type of gastric pit pattern were detected in 9 cases and 39 cases in patients with early gastric cancer, respectively. The presence of irregular minute vessels and variation in the caliber of vessels was found in 109 cases. The accuracy, sensitivity, specificity, false positive rate and false negative rate for diagnosis of early gastric cancer and precancerous gastric lesions were 68.9%, 95.1%, 63.1%, 24.5%, and 32.4% for conventional endoscopy, 93.6%, 92.7%, 94.5%, 5.7%, and 6.9% for magnifying NBI, and 91.3%, 88.6%, 93.2%, 13.2%, and 21.48% for magnifying chromoendoscopy, respectively.ConclusionsThis study demonstrates that magnifying NBI is superior to conventional endoscopy in the diagnosis of early gastric cancer and precancerous gastric lesions, and can be used for screening early malignancies of the stomach.

A timely diagnosis of early gastric cancer (EGC) can greatly reduce the death rate of patients. However, the manual detection of EGC is a costly and low-accuracy task. The artificial intelligence (AI) method based on deep learning is considered as a potential method to detect EGC. AI methods have outperformed endoscopists in EGC detection, especially with the use of the different region convolutional neural network (RCNN) models recently reported. However, no studies compared the performances of different RCNN series models. This study aimed to compare the performances of different RCNN series models for EGC. Three typical RCNN models were used to detect gastric cancer using 3659 gastroscopic images, including 1434 images of EGC: Faster RCNN, Cascade RCNN, and Mask RCNN. The models were evaluated in terms of specificity, accuracy, precision, recall, and AP. Fast RCNN, Cascade RCNN, and Mask RCNN had similar accuracy (0.935, 0.938, and 0.935). The specificity of Cascade RCNN was 0.946, which was slightly higher than 0.908 for Faster RCNN and 0.908 for Mask RCNN. Faster RCNN and Mask RCNN place more emphasis on positive detection, and Cascade RCNN places more emphasis on negative detection. These methods based on deep learning were conducive to helping in early cancer diagnosis using endoscopic images.

American Gastroenterological Association (AGA) Institute Technology Assessment on Image-Enhanced Endoscopy

AIMTo investigate the value of multiparameter joint analysis in the early diagnosis of gastric cancer (GC) in clinical practice.METHODSConcentrations of CEA, CA724 and three kinds of cytokines (TNF-α, IL-6 and IL-8) in 176 GC patients, 117 atypical hyperplasia patients, and 204 healthy control individuals were used for building the diagnostic model, then 58 GC patients, 41 atypical hyperplasia patients, and 66 healthy control individuals were enrolled independently. The joints of the indicators were analyzed by binary logistic regression analysis method.RESULTSFor discriminating the healthy control group and the GC group, IL-6 had the best diagnostic value, and the area under curve (AUC) of joint analysis was 0.95 (0.93-0.97). For the early stage and advanced stage GC, the AUC were 0.95 (0.92-0.98) and 0.95 (0.92-0.97). For discriminating the atypical hyperplasia group and GC group, CA724 had the best diagnostic value, and the AUC of joint analysis was 0.97 (0.95-0.99). For the early stage and advanced stage GC groups, the AUC were 0.98 (0.96-0.99) and 0.96 (0.94-0.98). After evaluation, for discriminating the GC, early stage GC and advanced cancer group from the healthy control group, the diagnostic sensitivity was 89.66%, 84.21% and 92.31%, respectively, and the specificity was 92.42%, 90.91% and 90.91%. For discriminating the GC, early stage GC and advanced cancer groups from the atypical hyperplasia group, the diagnostic sensitivity was 87.93%, 78.95% and 92.31%, respectively, and the specificity was 87.80%, 85.37% and 90.24%.CONCLUSIONWe have built a diagnostic model including CEA, CA724, IL-6, IL-8, and TNF-α. It may provide potential assistance as a screening method for the early detection of GC.

The key to reducing the mortality of gastric cancer is early detection, early diagnosis, and early treatment of gastric cancer. Early diagnosis of gastric cancer is the key to early detection and diagnosis of gastric cancer. Early diagnosis and treatment of gastric cancer is of great significance for improving the curative effect and reducing mortality of gastric cancer. The purpose of this paper is to study the diagnosis of early gastric cancer based on medical imaging techniques and mathematical modeling. The effect of W-DeepLab network-assisted diagnosis of images under white light was analyzed, and the value of Narrow Band Imaging and Blue Laser Imaging in the diagnosis of early gastric cancer was compared. Because Blue Laser Imaging endoscopy can clearly observe the demarcation line and microvascular morphology; but when using Narrow Band Imaging observation, part of the demarcation line and microvascular morphology is not observed. The results show that Blue Laser Imaging is brighter than Narrow Band Imaging's endoscopic images, and it is easier to observe the microstructure of lesions under endoscopy, so as to accurately determine the nature of lesions.

With the widespread use of advanced endoscopic techniques such as endoscopic submucosal dissection, an increasing number of early colorectal cancer (T1 CRC) and early gastric cancer (EGC) cases are now treated with endoscopic resection as the first-line approach. However, the risk of lymph node metastasis (LNM)-approximately 10% in T1 CRC and 5%-10% in EGC-necessitates additional surgical resection in high-risk cases. Current guideline-based risk stratification depends on pathological evaluation of the resected specimens to determine whether further surgery is needed. Yet both T1 CRC and EGC face shared challenges in LNM risk prediction, particularly in terms of accuracy and reproducibility. This review focuses on the latter. The diagnosis of key pathological risk factors, which serve as predictors of LNM, is subject to considerable interobserver variability among pathologists. One potential solution is the application of artificial intelligence (AI)-assisted whole slide image (WSI) analysis, which has been gaining attention in recent studies. AI-assisted models for LNM prediction in T1 CRC and EGC have shown encouraging results, suggesting that WSI-based AI could offer a pathologist-independent strategy to improve diagnostic consistency. However, the field remains in an early stage, with key limitations including small sample sizes and limited external validation. Additional high-quality evidence will be needed to support clinical implementation. Addressing challenges such as stain standardization and image artifacts will also be critical for achieving regulatory approval and broader clinical adoption.

Response:

Gastric cancer is the fifth most common cancer and in 2018, it was the third most common cause of cancer-related deaths worldwide. Endoscopic advances continue to be made for the diagnosis and management of both early gastric cancer and premalignant gastric conditions. In this review, we discuss the epidemiology and risk factors of gastric cancer and emphasize the differences in early vs late-stage gastric cancer outcomes. We then discuss endoscopic advances in the diagnosis of early gastric cancer and premalignant gastric lesions. This includes the implementation of different imaging modalities such as narrow-band imaging, chromoendoscopy, confocal laser endomicroscopy, and other experimental techniques. We also discuss the use of endoscopic ultrasound in the diagnosis and staging of early gastric cancer. We then discuss the endoscopic advances made in the treatment of these conditions, including endoscopic mucosal resection, endoscopic submucosal dissection, and hybrid techniques such as laparoscopic endoscopic cooperative surgery. Finally, we comment on the current suggested recommendations for surveillance of both gastric cancer and its premalignant conditions.

Objective To evaluate the diagnostic accuracy of four enhanced magnifying endoscopy methods for precancerous lesions and early gastric cancers, i. e., magnifying endoscopy enhanced by epinephrine(ME-EPI), magnifying endoscopy enhanced by acetic acid-indigo carmine mixture(ME-AIM), magnifying endoscopy enhanced by indigo carmine(ME-IDC) and magnifying endoscopy enhanced with narrow-band imaging(ME-NBI). Methods A total of 3 059 patients(3 097 focal gastric lesions) detected with conventional white light endoscopy(WLE) at four hospitals from January 2010 to December 2011 were recruited and divided into three groups according to different chromoendoscopic procedures they underwent: 1 094 ME-EPI cases, 873 ME-AIM cases and 1 130 ME-IDC cases. The pathological results were used as golden standard, and diagnosis accuracy rates of precancerous lesions or early cancers were determined according to the VS criteria. At the mean time, in each group, a paired design was used to compare the accuracy between the above chromoendoscopies and ME-NBI respectively. Results A total of 3 097 lesions of 3 059 patients were evaluated, and 205 lesions were diagnosed as early gastric cancer according to pathological results. The diagnostic accuracy rates of ME-EPI, ME-AIM and ME-IDC for early gastric cancer were 98.2%, 99.0% and 98.4%, respectively. In each group, the diagnostic accuracy rates of paired ME-NBI were 98.0%, 99.0% and 97.4%, respectively. In contrast to ME-NBI, the diagnostic accuracy of ME-EPI and ME-AIM were comparable or lower, but the differences were not statistically significant(P>0.05). However, when it came to ME-IDC, the diagnostic accuracy declined and the differences were statistically significant(P<0.05). Conclusion Chromoendoscopy can be used to identify the early gastric cancer accurately, easily, safely and inexpensively. And the diagnostic effectiveness of ME-NBI is not superior to chromoendoscopy in diagnosing early gastric cancer. Key words: Early gastric cancer; Diagnosis; Chromoendoscopy; Narrow-band imaging

Despite declines in incidence, gastric cancer remains a disease with a poor prognosis and limited treatment options due to its often late stage of diagnosis. In contrast, early gastric cancer...

Objective To evaluate the diagnostic value of VS classification of magnifying endoscopy with blue laser imaging (ME-BLI) for gastric precancerous lesion and early gastric cancer. Methods A retrospective study was performed on the data of 313 patients (322 lesions) with gastric mucosal lesions undergoing ME-BLI in digestive endoscopy center of Renmin Hospital of Wuhan University from January 2014 to January 2017. The accuracy, sensitivity, specificity, positive predictive value and negative predictive value of VS classification by ME-BLI in diagnosis of gastric precancerous lesion and early cancer were analyzed. Results Among the 322 lesions, 57 were pathologically diagnosed as cancerous lesions and 265 were non-cancerous lesions. According to VS classification of ME-BLI, 98.2% (56/57) VS structures of the cancerous lesions were irregular or disappearing, and 100.0% (57/57) cancerous lesions had clear demarcation. Taking the pathological diagnosis as the gold standard, the accuracy of VS classification of ME-BLI was 93.8% (302/322), with a good consistency with pathological diagnosis(Kappa=0.810). The sensitivity, specificity, positive predictive value, and negative predictive value were 98.2% (56/57), 92.8% (246/265), 74.7% (56/75) and 99.6% (246/247), respectively. Conclusion The VS classification of ME-BLI is an effective method with high accuracy, sensitivity and specificity for diagnosis of gastric precancerous lesion and early gastric cancer. Key words: Endoscopy; Magnifying endoscopy with blue laser imaging; VS classification; Gastric precancerous lesions; Early gastric cancer

Chapter 15 - Biosensor-based early diagnosis of gastric cancer

Objective To test the value of miRNA-30c (miR-30c) in early diagnosis of gastric cancer. Methods Serum miR-30c expression levels were detected by using quantitative real-time polymerase chain reaction in 80 patients with advanced gastric cancer, 35 patients with early stage gastric cancer and 35 healthy controls in the Affiliated People's Hospital of Inner Mongolia Medical University from August 2014 to August 2017. The receiver operating characteristic (ROC) curves and the area under the curve (AUC) were analyzed to test the efficacy of the miR-30c in distinguishing advanced gastric cancer, early stage gastric cancer and healthy controls. Results The expression level of serum miR-30c in advanced gastric cancer (0.45±0.11) was lower than that in early stage gastric cancer (0.54±0.15) (t = 5.2, P < 0.05). Compared with the healthy controls (0.61±0.12), miR-30c was down-expressed in the serum of early stage gastric cancer patients (t = 6.7, P < 0.05). Compared with the traditional tumor markers, the AUC of miR-30c was the biggest (0.92±0.03) in early stage and advanced gastric cancer groups, and the sensitivity and specificity in the diagnosis of gastric cancer were 90% and 84%, respectively. The AUC of miR-30c was 0.87±0.04 in early stage gastric cancer and healthy controls, and the sensitivity and specificity in the diagnosis of gastric cancer were 70% and 86%, respectively. Conclusion miR-30c might be used as a potential serum biomarker for the early diagnosis of gastric cancer. Key words: Stomach neoplasms; MicroRNAs; Early diagnosis