A DEEP LEARNING-BASED SYSTEM FOR IDENTIFYING DIFFERENTIATION STATUS AND DELINEATING MARGINS OF EARLY GASTRIC CANCER IN NARROW-BAND IMAGING ENDOSCOPY

2020 ◽  
Author(s):  
H Yu ◽  
L Wu ◽  
T Ling ◽  
S Hu
Keyword(s):  
Gastric Cancer ◽  
Deep Learning ◽  
Early Gastric Cancer ◽  
Narrow Band ◽  
Narrow Band Imaging ◽  
Differentiation Status

A deep learning-based system for identifying differentiation status and delineating the margins of early gastric cancer in magnifying narrow-band imaging endoscopy

Endoscopy ◽  
2020 ◽  
Author(s):  
Tingsheng Ling ◽  
Lianlian Wu ◽  
Yiwei Fu ◽  
Qinwei Xu ◽  
Ping An ◽  
...  
Keyword(s):  
Gastric Cancer ◽  
Deep Learning ◽  
Early Gastric Cancer ◽  
Real Time ◽  
Narrow Band ◽  
Narrow Band Imaging ◽  
Superior Performance ◽  
Accurate Identification ◽  
Real Time System ◽  
Differentiation Status

Abstract Background Accurate identification of the differentiation status and margins for early gastric cancer (EGC) is critical for determining the surgical strategy and achieving curative resection in EGC patients. The aim of this study was to develop a real-time system to accurately identify differentiation status and delineate the margins of EGC on magnifying narrow-band imaging (ME-NBI) endoscopy. Methods 2217 images from 145 EGC patients and 1870 images from 139 EGC patients were retrospectively collected to train and test the first convolutional neural network (CNN1) to identify EGC differentiation status. The performance of CNN1 was then compared with that of experts using 882 images from 58 EGC patients. Finally, 928 images from 132 EGC patients and 742 images from 87 EGC patients were used to train and test CNN2 to delineate the EGC margins. Results The system correctly predicted the differentiation status of EGCs with an accuracy of 83.3 % (95 % confidence interval [CI] 81.5 % – 84.9 %) in the testing dataset. In the man – machine contest, CNN1 performed significantly better than the five experts (86.2 %, 95 %CI 75.1 % – 92.8 % vs. 69.7 %, 95 %CI 64.1 % – 74.7 %). For delineating EGC margins, the system achieved an accuracy of 82.7 % (95 %CI 78.6 % – 86.1 %) in differentiated EGC and 88.1 % (95 %CI 84.2 % – 91.1 %) in undifferentiated EGC under an overlap ratio of 0.80. In unprocessed EGC videos, the system achieved real-time diagnosis of EGC differentiation status and EGC margin delineation in ME-NBI endoscopy. Conclusion We developed a deep learning-based system to accurately identify differentiation status and delineate the margins of EGC in ME-NBI endoscopy. This system achieved superior performance when compared with experts and was successfully tested in real EGC videos.

Endoscopic submucosal dissection by using Clutch cutter and IT knife 2 for early gastric cancer with severe submucosal fibrosis: a case report

MedPharmRes ◽  
2018 ◽  
Vol 2 (3) ◽  
pp. 17-21
Author(s):  
Nhan Le ◽  
Phuong Vien ◽  
Nghia Le
Keyword(s):  
Gastric Cancer ◽  
Early Gastric Cancer ◽  
Endoscopic Submucosal Dissection ◽  
Narrow Band ◽  
Narrow Band Imaging ◽  
Magnifying Endoscopy ◽  
Gastrointestinal Cancers ◽  
Pathological Result ◽  
Well Differentiated ◽  
Submucosal Fibrosis

Introduction: Gastric cancer is one of the highly malignant gastrointestinal cancers and the third leading cause of cancer death. In the last decade, early gastric cancer (EGC) has been reported by using narrow-band imaging (NBI) magnifying endoscopy. Advances in endoscopic techniques, such as endoscopic submucosal dissection (ESD), have enabled the en bloc resection of these EGC. Although ESD is performed for early gastric cancer, there are still many difficult problems in technique of this procedure. The difficulty of gastric ESD depends on the size and location of a tumor, presence of severe submucosal fibrosis, presence of ulceration... We report a case of our successful ESD by using Clutch cutter and IT knife 2 in treatment of EGC with severe submucosal fibrosis. Case presentation: A 62-year-old man felt an epigastric discomfort two months ago. The narrow-band imaging (NBI) magnifying endoscopy revealed a suspected early gastric cancer type 0 - IIa + IIc (Japanese classification of early gastrointestinal cancers) at the incisura angularis, the size of this lesion was 15 mm in diameter, and pathological result of endoscopic biopsy was a well-differentiated adenocarcinoma. ESD was performed and we found there was severe submucosal fibrosis which was dissected safer and faster by using Clutch cutter and IT knife 2. There were no complications such as severe bleeding and perforation. The size of resected specimen was 60 x 35 mm and the time of procedure was 150 minutes. After ESD, the pathological result was a well differentiated adenocarcinoma, pT1a, UL(-), LY(-), V(-), no cancer cell in vertical and horizontal margins. The healing time of ESD-induced ulcer was 5 weeks without local recurrence. Conclusion: Through this case, we aim to emphasize the importance of using Clutch cutter and IT knife 2 as a modified technique which makes ESD a safe procedure in treatment of EGC with severe submucosal fibrosis.

Magnifying Blue Laser Imaging versus Magnifying Narrow-Band Imaging for the Diagnosis of Early Gastric Cancer: A Prospective, Multicenter, Comparative Study

Digestion ◽  
2017 ◽  
Vol 96 (3) ◽  
pp. 127-134 ◽  
Author(s):  
Osamu Dohi ◽  
Nobuaki Yagi ◽  
Shigeto Yoshida ◽  
Shoko Ono ◽  
Yoji Sanomura ◽  
...  
Keyword(s):  
Gastric Cancer ◽  
Early Gastric Cancer ◽  
Comparative Study ◽  
Narrow Band ◽  
Narrow Band Imaging ◽  
Blue Laser ◽  
Laser Imaging ◽  
Blue Laser Imaging

scholarly journals Diagnostic limitations of magnifying endoscopy with narrow-band imaging in early gastric cancer

2020 ◽  
Vol 08 (10) ◽  
pp. E1233-E1242
Author(s):  
Kohei Matsumoto ◽  
Hiroya Ueyama ◽  
Takashi Yao ◽  
Daiki Abe ◽  
Shotaro Oki ◽  
...  
Keyword(s):  
Gastric Cancer ◽  
Early Gastric Cancer ◽  
Gastric Adenocarcinoma ◽  
Narrow Band ◽  
Narrow Band Imaging ◽  
Magnifying Endoscopy ◽  
Clinicopathological Features ◽  
Fundic Gland ◽  
Undifferentiated Adenocarcinoma ◽  
Histological Characteristics

Abstract Background and study aims Magnifying endoscopy with narrow band imaging (M-NBI) has made a huge contribution to endoscopic diagnosis of early gastric cancer (EGC). However, we sometimes encountered false-negative cases with M-NBI diagnosis (i. e., M-NBI diagnostic limitation lesion: M-NBI-DLL). However, clinicopathological features of M-NBI-DLLs have not been well elucidated. We aimed to clarify the clinicopathological features and histological reasons of M-NBI-DLLs. Patients and methods In this single-center retrospective study, M-NBI-DLLs were extracted from 456 EGCs resected endoscopically at our hospital. We defined histological types of M-NBI-DLLs and analyzed clinicopathologically to clarify histological reasons of M-NBI-DLLs. Results Of 456 EGCs, 48 lesions (10.5 %) of M-NBI-DLLs were enrolled. M-NBI-DLLs was classified into four histological types as follows: gastric adenocarcinoma of fundic-gland type (GA-FG, n = 25), gastric adenocarcinoma of fundic-gland mucosal type (GA-FGM, n = 1), differentiated adenocarcinoma (n = 14), and undifferentiated adenocarcinoma (n = 8). Thirty-nine lesions of M-NBI-DLLs were H. pylori-negative gastric cancers (39/47, 82.9 %). Histological reasons for M-NBI-DLLs were as follows: 1) completely covered with non-neoplastic mucosa (25/25 GA-FG, 8/8 undifferentiated adenocarcinoma); 2) well-differentiated adenocarcinoma with low-grade atypia (1/1 GA-FGM, 14/14 differentiated adenocarcinoma); 3) similarity of surface structure (10/14 differentiated adenocarcinoma); and 4) partially covered and/or mixed with a non-neoplastic mucosa (1/1 GA-FGM, 6/14 differentiated adenocarcinoma). Conclusions Diagnostic limitations of M-NBI depend on four distinct histological characteristics. For accurate diagnosis of M-NBI-DLLs, it may be necessary to fully understand endoscopic features of these lesions using white light imaging and M-NBI based on these histological characteristics and to take a precise biopsy.

Sign in / Sign up

Export Citation Format

Share Document