Abstract:BackgroundFacilitating accurate location of marking points and circumferential incision in the mucosa at a certain distance outside the boundary of the lesion is important to ensure the quality of endoscopic submucosal dissection (ESD). However, the decision-making of different endoscopists has a large variance due to subjective experience, leading to difficulty in quality control of ESD. To tackle this problem, this study aims at utilizing an intelligent automatic system to give real-time guidance during the operation.MethodsWe built an Artificial Intelligence (AI) assistant system that can detect the lesion boundary and marked points in real-time to quantify the location of marking points and cutting path during ESD of early esophageal cancer(EEC). This system was composed of two major components, that were the marking points suggestion module and the cutting line suggestion module. We collected 499 endoscopic images with Lugol’s staining of EEC from 20 cases and 3397 detailed images of ESD from 12 cases. With these images, we built datasets for the evaluation of lesion boundary detection and marked points detection. Then, the algorithm was built to display the marking position and the mucosal incision position.ResultsWe conducted the experiments, including two parts. The first one was to assess the real-time boundary detection performance and marked points detection performance. Specifically, we achieved a satisfactory real-time lesion segmentation performance with Dice score of 96.1%, and a good marked points detection performance with average precision (AP) 50 of 72.1%. On the basis of boundary detection, our algorithm can automatically display the marking position at a certain distance (such as 0.3cm, which can be adjusted as needed) outside the boundary of the lesion. After the guidance of marking points, the automatic system can detect the marked points and display the mucosal cutting line at a certain distance outside the marked points.ConclusionsSuggesting position of marking and incision were successfully achieved during ESD with our automatic system, which can improve consistency of standard ESD procedure among different level hospitals.

Design of the Image-Guided Biopsy Marking System for Gastroscopy

A Non-Invasive Navigation System for Retargeting Gastroscopic Lesions

Computer-Simulated Biopsy Marking System for Endoscopic Surveillance of Gastric Lesions: A Pilot Study

Gastroscopic Image Graph: Application to Noninvasive Multitarget Tracking under Gastroscopy

A Computer Simulated Biopsy Marking System for Gastroscopy

A Merging Model Reconstruction Method for Image-Guided Gastroscopic Biopsy.

Trial of A Novel Endoscopic Tattooing Biopsy Forceps on Animal Model

Incorporation of 3D Model and Panoramic View for Gastroscopic Lesion Surveillance.

The Establishment and Clinical Appliance of Technique of Mucosa Marking Targeting Biopsy

Global and Local Panoramic Views for Gastroscopy: an Assisted Method of Gastroscopic Lesion Surveillance.

Development of an automatic system to fast quantify marking and incision during endoscopic submucosal dissection

Automated Fiducial Marker Detection and Fiducial Point Localization in CT Images for Lung Biopsy Image-Guided Surgery Systems.

Online Tracking and Retargeting with Applications to Optical Biopsy in Gastrointestinal Endoscopic Examinations.

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

An Image-Guided Sinus Surgery System

A Co-Axial Projection Surgical Navigation System for Breast Cancer Sentinel Lymph Node Mapping: System Design and Clinical Trial

Mixed Reality Biopsy Navigation System Utilizing Markerless Needle Tracking and Imaging Data Superimposition

Accuracy and Efficiency of an Infrared Based Positioning and Tracking System for Image-Guided Intervention

Modular Robotic Scanning Device for Real-Time Gastric Endomicroscopy

Optical biopsy mapping for minimally invasive cancer screening.

A method and a device for diagnosing an upper gastrointestinal endoscopic image