Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Abnormal gait detection method based on long-short-term memory neural network

A long-short-term memory and neural network technology, applied in the field of abnormal gait detection based on long-short-term memory neural network, can solve problems such as lack of diagnostic methods, dependence on clinicians' professional knowledge, and patients unable to receive treatment.

Pending Publication Date: 2021-08-31
NANKAI UNIV
View PDF9 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, due to the unclear pathology and the lack of early diagnosis methods of the disease, many patients cannot receive timely treatment
Gait assessment is challenging, relies heavily on clinician expertise, and is subjective

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Abnormal gait detection method based on long-short-term memory neural network
  • Abnormal gait detection method based on long-short-term memory neural network
  • Abnormal gait detection method based on long-short-term memory neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0032] This is a binary classification experiment, which distinguishes between normal people and patients.

[0033] Using data from a database, the dataset contains plantar pressure measurements from 93 Parkinson's disease patients and 73 healthy controls. The subjects walked at their own pace on a level surface for approximately 2 minutes. In each insole worn by the subject, there are eight sensors, denoted as L1, L2, L3, L4, L5, L6, L7, L8, R1, R2, R3, R4, R5, R6, R7 and R8 sampling location. Measure plantar pressure at 100 Hz / s. That is to collect 8 channels of plantar pressure signals on the left foot and right foot respectively, and calculate the resultant force of the left and right feet, a total of 18 channels of signals.

[0034] After analysis, the gait cycle is roughly 1s, and the data set is divided according to the experimental objects to ensure that the data of the same patient only appear in the training set or test set. Because the gait cycle is about 1s, it...

Embodiment 2

[0051] This is a multi-category experiment. According to the scoring information in the database, the test subjects are divided into 5 levels. The expert score is less than 5 for the first category, the score between 5 and 15 is the second category, and the score between 15 and 25 is the third category. Class, based on 25 and 35 is the fourth class, greater than 35 is the fifth class, and the class is one-hot encoded.

[0052]In each insole worn by the subject, there are eight sensors, denoted as L1, L2, L3, L4, L5, L6, L7, L8, R1, R2, R3, R4, R5, R6, R7 and R8 Sampling position: Using the data in the database, collect 8 channels of plantar pressure signals on the left foot and right foot, and calculate the resultant force of the left and right feet, a total of 18 channels of signals, and the sampling frequency is 100HZ.

[0053] After analysis, the gait cycle is roughly 1s, and the data set is divided according to the experimental objects to ensure that the data of the same p...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention provides an abnormal gait detection method based on a long-short-term memory neural network. The method comprises the following steps of data preprocessing, wherein data preprocessing comprises the steps of performing test object category division based on a database, collecting plantar pressure data in the gait process of a subject, dividing data sets according to test objects, and ensuring that data of the same patient only appears in a training set or a test set; and model building based on the long-short-term memory neural network, performing mean value pooling on the output of each LSTM unit, wherein the output of each LSTM unit represents abstract features learned from all input data before the current moment, splicing the abstract features with features at the last moment, sending the spliced abstract features to a full connection layer, finally outputting a probability value by using an activation function, using a Sigmoid function during binary classification, and using a Softmax function during multi-classification; and model training and result evaluation. According to the method, gait disorder can be quantitatively evaluated, and diagnosis assistance is provided for doctors.

Description

technical field [0001] The invention relates to the fields of digital diagnosis, artificial intelligence, and rehabilitation training, in particular to an abnormal gait detection method based on a long-short-term memory neural network. Background technique [0002] Abnormal gait can be caused by motor or sensory impairment, and its characteristics are related to the lesion site. It can be found in many nervous system diseases, such as Parkinson's disease, normal pressure hydrocephalus and so on. The typical abnormal gait is suggestive of a specific disease, but the current diagnostic method commonly used by doctors is to make a diagnosis through inspection, and it is often difficult to distinguish early stage patients with mild illness. There is an urgent need for quantitative and meticulous detection methods for abnormal gait, which can be helpful for diagnosis. [0003] The ground reaction force is the only external force that a person receives during walking, and the pl...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): A61B5/11A61B5/00G06K9/62G06N3/04G06N3/08G16H50/20
CPCA61B5/112A61B5/6807A61B5/7267G16H50/20G06N3/049G06N3/08G06N3/047G06N3/044G06F18/2415
Inventor 孙玉波刘嘉男于宁波韩建达
Owner NANKAI UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com