Digestive tract disease auxiliary diagnosis system based on deep learning
A technology for gastrointestinal diseases and auxiliary diagnosis, applied in the field of artificial intelligence, can solve the problems of low accuracy of human eye diagnosis, slow manual inspection, and large number of biopsies, so as to improve the efficiency of auxiliary diagnosis and treatment and scientific research, and improve the level of screening. , the effect of reducing the burden on doctors
Patent Information
- Authority / Receiving Office
- CN · China
- Current Assignee / Owner
- Publication Date
- 2020-05-08
Smart Images

Figure 1 
Figure 2 
Figure 3
Abstract
Description
technical field
[0001] The invention belongs to the technical field of artificial intelligence, and in particular relates to an auxiliary diagnosis system for digestive tract diseases based on deep learning. Background technique
[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.
[0003] According to the 2018 global cancer statistics, 4 of the top 10 tumors with the highest incidence rate come from the digestive tract. Gastrointestinal diseases, including benign, precancerous and malignant diseases of the digestive tract, are seriously threatening the quality of life and life safety of patients, causing a huge health burden. Early diagnosis and treatment can improve the prognosis of patients and save medical resources problem needs to be resolved urgently. With the development and popularization of medical imaging equipment, gastrointestinal diseases can usually be foun...
Examples
Embodiment approach
[0073] As another implementation, the lesion area positioning module further includes:
[0074] Digestive tract lesion area recognition model training module, which is used to acquire multiple images of digestive tract parts containing known digestive tract lesions, such as Figure 3(a)-Figure 3(d) , and label the known gastrointestinal lesion area, the labeling results are as follows Figure 4 As shown; specify the existing network model architecture, or build a custom network module architecture; receive model training parameters, and train the digestive tract lesion area recognition model according to the labeled training images. Identify the digestive tract lesion area and the final output labeling results, such as Figure 5 shown.
[0075] As an implementation, the existing network model architecture adopts the YOLO v3 neural network.
[0076] Utilizing its characteristics of high detection accuracy and fast detection speed, it can meet the needs of electronic gastrosc...