Toward Automatic Detection of Gastric Lesion for Upper Gastrointestinal Endoscopy with Neural Network

Author(s):  
Sai Wang ◽  
Qi He ◽  
Ping Zhang ◽  
Xin Chen ◽  
Siyang Zuo

In this paper, we compared the performance of several neural networks in the classification of early gastric cancer (EGC) images and proposed a method of converting the output value of the network into a calorific value to locate the lesion. The algorithm was improved using transfer learning and fine-tuning principles. The test set accuracy rate reached 0.72, sensitivity reached 0.67, specificity reached 0.77, and precision rate reached 0.78. The experimental results show the potential to meet clinical demands for automatic detection of gastric lesion.

2021 ◽  
Vol 12 (02) ◽  
pp. 103-106
Author(s):  
Avnish Kumar Seth ◽  
Rinkesh Kumar Bansal

Abstract Background We report three patients with endoscopic insufflation–induced gastric barotrauma (EIGB) during upper gastrointestinal endoscopy (UGIE) for percutaneous endoscopic gastrostomy (PEG). A definition and classification of EIGB is proposed. Materials and Methods Records of patients undergoing UGIE over 7 years (April 2013–March 2020) were reviewed. Patients who developed new onset of bleeding or petechial spots in proximal stomach, in an area previously documented to be normal during the same endoscopic procedure, were studied. Results New onset of bleeding or petechial spots in proximal stomach occurred in 3/286 (0.1%) patients undergoing PEG and in none of the 19,323 other UGIE procedures during the study period. All patients were men with median age 76 years (range 68–80 years), with no coagulopathy. Aspirin and apixaban were discontinued 1 week and 3 days prior to the procedure. Fresh blood was noted in the stomach at a median of 275 seconds (range 130–340) seconds after commencement of endoscopy. At retroflexion, multiple linear mucosal breaks of up to 3 cm, with oozing of blood, were noted in the proximal stomach along the lesser curvature, close to the gastroesophageal junction in two patients. In the third patient, multiple petechial spots were noticed in the fundus. The plan for PEG was abandoned and the stomach deflated by endoscopic suction. There was no subsequent hematemesis, melena, or drop in hemoglobin. One week later, repeat UGIE in the first two patients revealed multiple healing linear ulcers of 1 to 3 cm in the lesser curvature and PEG was performed. Conclusion Overinsufflation over a short duration during UGIE may lead to EIGB. Early detection is key and in the absence gastric perforation, patients can be managed conservatively.


Author(s):  
Po-Hsiang Lin ◽  
Jer-Guang Hsieh ◽  
Hsien-Chung Yu ◽  
Jyh-Horng Jeng ◽  
Chiao-Lin Hsu ◽  
...  

Determining the target population for the screening of Barrett’s esophagus (BE), a precancerous condition of esophageal adenocarcinoma, remains a challenge in Asia. The aim of our study was to develop risk prediction models for BE using logistic regression (LR) and artificial neural network (ANN) methods. Their predictive performances were compared. We retrospectively analyzed 9646 adults aged ≥20 years undergoing upper gastrointestinal endoscopy at a health examinations center in Taiwan. Evaluated by using 10-fold cross-validation, both models exhibited good discriminative power, with comparable area under curve (AUC) for the LR and ANN models (Both AUC were 0.702). Our risk prediction models for BE were developed from individuals with or without clinical indications of upper gastrointestinal endoscopy. The models have the potential to serve as a practical tool for identifying high-risk individuals of BE among the general population for endoscopic screening.


Sign in / Sign up

Export Citation Format

Share Document