Intestinal nerve plexus auxiliary identification method based on improved progressive residual network

Abstract

The invention discloses an intestinal nerve plexus auxiliary identification method based on an improved progressive residual network, which comprises the following steps of: 1, collecting data, and preprocessing the data to obtain a training data set; 2, constructing a restoration model based on an improved progressive residual restoration network, and training the restoration model based on the training data set and MSEloss; 3, inputting an original to-be-detected image into the trained restoration model to obtain a restoration result; 4, performing pixel-level difference comparison on the restoration result and the original to-be-detected image to obtain the positions of enteric nerve plexuses; and 5, classifying normal and abnormal nerve plexuses by adopting a classification model. According to the method, nerve plexuses can be accurately and quickly identified, normal nerve plexuses and abnormal nerve plexuses are classified firstly, the nerve plexuses in a normal form and an abnormal form are evaluated, subsequent ganglion cell identification is performed after a nerve plexus is identified as a normal nerve plexus, and the identification efficiency and accuracy are greatly improved.

Description

technical field [0001] The invention relates to the technical field of image recognition, in particular to an enteric plexus-assisted recognition method based on an improved progressive residual network. Background technique [0002] Intestinal aganglionosis (aganglionosis) is one of the common developmental malformations of the digestive tract in children. The gold standard for judging intestinal aganglionosis is pathological examination. Its characteristic histological pathological change is the absence of nerve plexus and ganglion cells in the intestinal wall of the stenosis (lesion segment), and the enteric plexus and ganglion cells of normal shape are usually not seen in the myenterium and submucosal. On the contrary, exogenous cholinergic nerve fibers proliferate in a large number, increase in number, become thick, and tightly interweave into bundles, which are wavy or vortex, and the so-called "hypertrophic plexus" appears. The routine diagnostic process for this di...

AI Technical Summary

Problems solved by technology

[0006] Purpose of the invention: In order to overcome the deficiencies in the prior art, the present invention provides an enteric plexus-assisted identification method based on an improved progressive residual network, which can fully mine the inherent information of pathological pictures, and through special Preprocessing to solve the potential problem of insufficient data

Images