Xerophthalmia grading evaluation system based on regional adaptive multi-task neural network

Abstract

The invention discloses a xerophthalmia grading evaluation system based on a regional adaptive multi-task neural network. The system comprises a computer memory, a computer processor and a computer program which is stored in the computer memory and can be executed on the computer processor, wherein a trained xerophthalmia grading evaluation model is stored in the computer memory, and the xerophthalmia grading evaluation model comprises a VGG16 network, a regional recommendation network and a U-shaped full convolutional network. When the computer processor executes a computer program, the following steps are achievd by comprising steps of obtaining a to-be-tested original eyelid plate infrared image to carry out gray scale preprocessing, and carrying out bilateral filtering processing on asingle-channel gray scale image obtained after preprocessing; and inputting the processed image into a xerophthalmia grading evaluation model to obtain a xerophthalmia grading result, an eyelid platepositioning result and an alveolar segmentation result. According to the invention, automatic analysis of the eyelid plate infrared photograph can be realized, and auxiliary diagnosis of xerophthalmiagrading can be effectively carried out.

Description

technical field [0001] The invention belongs to the field of medical image analysis and machine learning, and in particular relates to a dry eye classification evaluation system based on a region-adaptive multi-task neural network. Background technique [0002] Dry eye is a multifactorial disease of the ocular surface, characterized by loss of tear film homeostasis, accompanied by ocular surface symptoms, and its pathogenesis includes tear film instability, tear hyperosmolarity, ocular surface inflammation and loss, and neurosensory abnormalities. Many studies have shown that the incidence of dry eye fluctuates from 5 to 50%. The traditional way of diagnosing and evaluating dry eye is to use ocular surface disease index (Ocular Surface Disease Index, OSDI), NIBUT, tear river height, meibomian gland morphology and other indicators, which can objectively diagnose and classify dry eye. However, the shortcomings of this evaluation method are: there are many relevant evaluation ...

AI Technical Summary

Problems solved by technology

[0007] In order to overcome the shortcomings of the existing dry eye classification that requires multiple detection aids, low efficiency, and low precision, the present invention discloses a dry eye classification evaluation system based on a region-adaptive multi-task neural network, which can realize the detection of tarsus Automatic analysis of infrared photos, effective auxiliary diagnosis for dry eye classification

Images