Facial nerve paralysis rehabilitation detection system based on artificial intelligence

Abstract

The invention provides a facial nerve paralysis rehabilitation detection system based on artificial intelligence. The system comprises: a facial three-dimensional model construction module for constructing a three-dimensional model of the face of a patient according to an acquired facial image of the patient; a face three-dimensional model calibration module which is used for calibrating a three-dimensional model of the face of the patient based on the acquired face thermal imaging image of the patient; a three-dimensional model parameter prediction module which is used for predicting a next frame of three-dimensional model based on the patient face thermal imaging image and the parameters of the three-dimensional model; a facial muscle detection module which is used for obtaining the responsivity of facial muscles based on the obtained muscle current response sequence and obtaining the movement degree of the facial muscles based on the predicted three-dimensional model; and a rehabilitation degree detection module which is used for obtaining the rehabilitation degree of the patient based on the movement degree of the facial muscles and the responsivity of the facial muscles. The system can accurately quantify the rehabilitation degree of a patient and assist the patient in rehabilitation training.

Description

technical field [0001] The invention relates to the fields of medical treatment and artificial intelligence, in particular to an artificial intelligence-based facial nerve paralysis rehabilitation detection system. Background technique [0002] At present, the methods for detecting facial nerve paralysis mostly use facial semantic segmentation and key point detection methods or analysis methods based on facial symmetry, but these methods have not been able to perform complete analysis and feature extraction on the characteristics of the patient's facial muscles, and lack accuracy. and efficiency. For example, the invention patent application document with the publication number CN111553250A proposes a method based on the analysis of the movement of each region of the face to realize facial paralysis detection. This method analyzes all facial key points under each action sequence, but the invention The above key points do not describe the more detailed movement characteristi...

AI Technical Summary

Problems solved by technology

[0002] At present, the methods for detecting facial nerve paralysis mostly use facial semantic segmentation and key point detection methods or analysis methods based on facial symmetry, but these methods have not been able to perform complete analysis and feature extraction on the characteristics of the patient's facial muscles, and lack accuracy. and efficiency

For example, the invention patent application document with the publication number CN111553250A proposes a method based on the analysis of the movement of each region of the face to realize facial paralysis detection. This method analyzes all facial key points under each action sequence, but the invention The above key points cannot describe the more detailed movement characteristics of facial muscles

Images