A deep neural network improves endoscopic detection of early gastric cancer without blind spots.
Endoscopy
; 51(6): 522-531, 2019 06.
Article
en En
| MEDLINE
| ID: mdl-30861533
ABSTRACT
BACKGROUND:
Gastric cancer is the third most lethal malignancy worldwide. A novel deep convolution neural network (DCNN) to perform visual tasks has been recently developed. The aim of this study was to build a system using the DCNN to detect early gastric cancer (EGC) without blind spots during esophagogastroduodenoscopy (EGD).METHODS:
3170 gastric cancer and 5981 benign images were collected to train the DCNN to detect EGC. A total of 24549 images from different parts of stomach were collected to train the DCNN to monitor blind spots. Class activation maps were developed to automatically cover suspicious cancerous regions. A grid model for the stomach was used to indicate the existence of blind spots in unprocessed EGD videos.RESULTS:
The DCNN identified EGC from non-malignancy with an accuracy of 92.5â%, a sensitivity of 94.0â%, a specificity of 91.0â%, a positive predictive value of 91.3â%, and a negative predictive value of 93.8â%, outperforming all levels of endoscopists. In the task of classifying gastric locations into 10 or 26 parts, the DCNN achieved an accuracy of 90â% or 65.9â%, on a par with the performance of experts. In real-time unprocessed EGD videos, the DCNN achieved automated performance for detecting EGC and monitoring blind spots.CONCLUSIONS:
We developed a system based on a DCNN to accurately detect EGC and recognize gastric locations better than endoscopists, and proactively track suspicious cancerous lesions and monitor blind spots during EGD.
Texto completo:
1
Bases de datos:
MEDLINE
Asunto principal:
Neoplasias Gástricas
/
Gastroscopía
/
Redes Neurales de la Computación
/
Detección Precoz del Cáncer
Tipo de estudio:
Diagnostic_studies
/
Prognostic_studies
/
Screening_studies
Límite:
Humans
Idioma:
En
Revista:
Endoscopy
Año:
2019
Tipo del documento:
Article
País de afiliación:
China