Two-step artificial intelligence system for endoscopic gastric biopsy improves the diagnostic accuracy of pathologists.

Background: Endoscopic biopsy is the pivotal procedure for the diagnosis of gastric cancer. In this study, we applied whole-slide images (WSIs) of endoscopic gastric biopsy specimens to develop an endoscopic gastric biopsy assistant system (EGBAS).

Methods: The EGBAS was trained using 2373 WSIs expertly annotated and internally validated on 245 WSIs. A large-scale, multicenter test dataset of 2003 WSIs was used to externally evaluate EGBAS. Eight pathologists were compared with the EGBAS using a man-machine comparison test dataset. The fully manual performance of the pathologists was also compared with semi-manual performance using EGBAS assistance.

Results: The average area under the curve of the EGBAS was 0·979 (0·958-0·990). For the diagnosis of all four categories, the overall accuracy of EGBAS was 86·95%, which was significantly higher than pathologists (P< 0·05). The EGBAS achieved a higher κ score (0·880, very good κ) than junior and senior pathologists (0·641 ± 0·088 and 0·729 ± 0·056). With EGBAS assistance, the overall accuracy (four-tier classification) of the pathologists increased from 66·49 ± 7·73% to 73·83 ± 5·73% (P< 0·05). The length of time for pathologists to manually complete the dataset was 461·44 ± 117·96 minutes; this time was reduced to 305·71 ± 82·43 minutes with EGBAS assistance (P = 0·00).

Conclusions: The EGBAS is a promising system for improving the diagnosis ability and reducing the workload of pathologists.