Classification of needle-EMG resting potentials by machine learning.

https://www.ncbi.nlm.nih.gov/pubmed/30353953

Muscle Nerve. 2018 Oct 24. doi: 10.1002/mus.26363. [Epub ahead of print]

Classification of needle-EMG resting potentials by machine learning.

Nodera H1, Osaki Y1, Yamazaki H1, Mori A1, Izumi Y1, Kaji R1.

Author information

1

Department of Neurology, Tokushima University, Tokushima, Japan.

Abstract

INTRODUCTION:

The diagnostic importance of audio signal characteristics in needle electromyography (EMG) is well established. Given the recent advent of audio-sound identification by artificial intelligence, we hypothesized that the extraction of characteristic resting EMG signals and application of machine learning algorithms could help classify various EMG discharges.

METHODS:

Data files of six classes of resting EMG signals were divided into 2-second segments. Extraction of characteristic features (384 and 4,367 features each) was used to classify the six types of discharges using machine learning algorithms.

RESULTS:

Across 841 audio files, the best overall accuracy of 90.4% was observed for the smaller feature set. Among the feature classes, Mel-Frequency Cepstral Coefficients (MFCC)-related features were useful in correct classification.

DISCUSSION:

We showed that needle EMG resting signals were satisfactorily classifiable by the combination of feature extraction and machine learning, and this can be applied to clinical settings. This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.

KEYWORDS:

Mel-Frequency Cepstral Coefficient; audio feature; classification; machine learning; needle electromyography; resting potential

PMID:

 

30353953

 

DOI:

 

10.1002/mus.26363

재활의학과나 신경과, 신경외과에서 검사하는

근전도 검사에대한 분석에도 
(당연히) 머신러닝을 통한 분석이 시도되고 있는 것 같다.

EEG 나 EKG 나 EMG 나....   모두 주파수 파형과 특정 feature 를 통해서 파악하는 것이니까.

이런건 당연히 인공지능 (deep learning) 이 더 학습을 잘하고, 패턴을 파악하고 분석해 내고, 더 빠르고 잘하지. 

학습도 빠르고, 지치지도 않고, 투정도 안하고....