US 2022/0189015 A1 · DEEP LEARNING BASED AUXILIARY DIAGNOSIS SYSTEM FOR EARLY GASTROINTESTINAL CANCER AND INSPECTION DEVICE

Abstract

A deep learning-based examination and diagnosis assistance system and apparatus for early digestive tract cancer comprising a feature extraction network, an image classification model, an endoscope classifier, and an early cancer recognition model. The feature extraction network is used for performing initial feature extraction on endoscope images based on a neural network model; the image classification model is used for performing extraction on the initial features to acquire image classification features; the endoscope classifier is used for performing feature extraction on the initial features to acquire endoscope classification features and classify gastroscope/colonoscope images; the early cancer recognition model is used for splicing the initial features, the endoscope classification features, and the image classification features to acquire the probability of early cancer lesions in white light images, electronic dye images or chemical dye images of a corresponding site or acquire a flushing prompt or position recognition prompt for the corresponding site.

Claims18

Dependent claims

A deep learning based inspection and diagnosis assistance system for early digestive tract cancer, comprising a feature extraction network, an image classification model, an endoscopic classifier and an early cancer recognition model,

wherein the feature extraction network is configured to extract a preliminary feature from an endoscopic image according to a neural network model;
the image classification model is configured to perform an extraction on the preliminary feature, acquire an image modality feature and acquire an image classification feature corresponding to a gastroscopic or colonoscopy image;
the endoscopic classifier is configured to extract the preliminary feature, and obtain an endoscope classification feature;
and
the early cancer recognition model is configured to stitch the preliminary feature, the endoscope classification feature, the image modality feature and the image classification feature, and obtain a probability of early cancer lesion in a white light image, an electronic staining image or a chemical staining image of a corresponding part or obtain a washing prompt or a position recognition prompt of the corresponding part.

2 more elements

A deep learning based inspection and diagnosis assistance device for early digestive tract cancer, comprising an AI display, an AI processor, a functional module and a control switch, the functional module being a rinsing and staining operation module,

wherein the AI processor is connected with the functional module through an electrical signal line, and is configured to control, according to a judgment result of the AI processor, the functional module based on a signal of the control switch;
the functional module and a gastrointestinal endoscope are connected with each other through a rising pipeline and a staining pipeline;
and
the functional module and the control switch are connected with each other through a line of a foot switch.

1 more element

Dependencies are inferred from the claim text.

Deep learning based auxiliary diagnosis system for early gastrointestinal cancer and inspection device

Life of this patent

Deep learning based auxiliary diagnosis system for early gastrointestinal cancer and inspection device

Life of this patent

Abstract

Claims18