Parkinson’s Disease EMG Data Augmentation and Simulation with DCGANs and Style Transfer (2020-05-01T00:00:00.000000Z)

TL;DR

Two new data augmentation approaches based on Deep Convolutional Generative Adversarial Networks (DCGANs) and Style Transfer for augmenting Parkinson’s Disease electromyography (EMG) signals are proposed.

Abstract

This paper proposes two new data augmentation approaches based on Deep Convolutional Generative Adversarial Networks (DCGANs) and Style Transfer for augmenting Parkinson’s Disease (PD) electromyography (EMG) signals. The experimental results indicate that the proposed models can adapt to different frequencies and amplitudes of tremor, simulating each patient’s tremor patterns and extending them to different sets of movement protocols. Therefore, one could use these models for extending the existing patient dataset and generating tremor simulations for validating treatment approaches on different movement scenarios.

Authors

R. A. Zanini

2 papers

E. Colombini

1 papers

TL;DR

Abstract

Authors

References47 items

Generative Adversarial Networks

Improving Style Transfer with Calibrated Metrics

Parkinson’s Disease EMG Signal Prediction Using Neural Networks

Synthesis of Realistic ECG using Generative Adversarial Networks

Electrocardiogram generation with a bidirectional LSTM-CNN generative adversarial network

Time Series Simulation by Conditional Generative Adversarial Net

A package for the computational analysis of complex biophysical signals

A Comprehensive Mathematical Model of Motor Unit Pool Organization, Surface Electromyography, and Force Generation

GANSynth: Adversarial Neural Audio Synthesis

EMG-based hand gesture control system for robotics

An empirical study on evaluation metrics of generative adversarial networks

EEG-GAN: Generative adversarial networks for electroencephalograhic (EEG) brain signals

Are GANs Created Equal? A Large-Scale Study

Progressive Growing of GANs for Improved Quality, Stability, and Variation

GANs Trained by a Two Time-Scale Update Rule Converge to a Nash Equilibrium

MidiNet: A Convolutional Generative Adversarial Network for Symbolic-Domain Music Generation

Mode Regularized Generative Adversarial Networks

Deep Learning with Convolutional Neural Networks Applied to Electromyography Data: A Resource for the Classification of Movements for Prosthetic Hands

Improved Techniques for Training GANs

Perceptual Losses for Real-Time Style Transfer and Super-Resolution

The Pathophysiology of Fatigue in Parkinson's Disease and its Pragmatic Management

Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks

A Neural Algorithm of Artistic Style

Electromyography data for non-invasive naturally-controlled robotic hand prostheses

Very Deep Convolutional Networks for Large-Scale Image Recognition

Movement Error Rate for Evaluation of Machine Learning Methods for sEMG-Based Hand Movement Classification

Compensation Active de Tremblements Pathologiques des Membres Supérieurs via la Stimulation Electrique Fonctionnelle

On-Line Signature Verification by Dynamic Time Warping and Gaussian Mixture Models

Toward accurate dynamic time warping in linear time and space

A Kernel Method for the Two-Sample-Problem

Physiologically based simulation of clinical EMG signals

Speaker identification using dynamic time warping with stress compensation technique

Accuracy of clinical diagnosis of idiopathic Parkinson's disease: a clinico-pathological study of 100 cases.

Parkinson's Disease Tremor Suppression - A Double Approach Study - Part 1

Keras - GAN

Available online: https://github.com/eriklindernoren/Keras-GAN/ tree/master/dcgan (accessed on 11 June 2019)

Hand Motion Recognition from Single Channel Surface EMG Using Wavelet & Artificial Neural Network☆

System and Methods for Emg-Triggered Neuromuscular Electrical Stimulation

Analysis of Simulated Electromyography (EMG) Signals Using Integrated Computer Muscle Model

Multi-Dimensional Dynamic Time Warping for Gesture Recognition

Cross-correlation as a method for comparing dynamic electromyography signals during gait.

World Health Organization. Neurological Disorders: Public Health Challenges; World Health Organization

Biosignal and Medical Image Processing

While trying to replicate complex output shapes, it is essential to add different convolutional pipelines that focus on speciﬁc features from the signal

We have evaluated different architecture parameters for the Generator model, using a different number of convolutional layers, filters, and different input sizes for the latent space

Mini-batch discriminator is vital if we wish to create a generator that can create a wide range of variations on the generated signals. However

metrics that

Field of Study

Journal Information

Name

Volume

Venue Information

Name

Type

URL

Alternate Names