Skip to content

GETTINGSTARTED.md

Latest commit

 

History

History
150 lines (112 loc) · 4.42 KB

GETTINGSTARTED.md

File metadata and controls

150 lines (112 loc) · 4.42 KB

Getting Started

Preprocessing

Create object data. NB: data is a struct as defined in the Data Requirements documentation page.

Use the getInfo function to extract info about the experimental protocol of the data collected. Those info need to be stored in the data structure. Edit the getInfo function to make the data structure compatible with the toolbox. Check the Data Requirements documentation page for more details.

load('data.mat');
info = getInfo(data);

Define the options to build the synergy analyzer object.

par.type = 'EmgKinData';
par.chlabels = [emgchannels(:)', kinchannels(:)'];
par.order = 2; %determines the order of the kinematic derivatives (0->position, 1->velocity, 2->acceleration) 
par.delay = -0.05; %50ms -> KIN data were anticipated of 50 ms to account for an estimated electromechanical delay between EMG and KIN (Scano et al. 2022)

Create the synergy analyzer object. With opt.verbose = 1 details of the processing will be printed in command window.

sa = SynergyAnalyzer(data,info,par);
sa.opt.verbose = 1;

Filter and resample EMG and KIN data.

sa.opt.emgFilter.type = 'fir1';
sa.opt.emgFilter.par = [50 20/(1000/2)];  % 20 Hz @ 1KHz EMG sampling rate
sa.opt.emgFilter.resample = 1;
sa.opt.emgFilter.resample_period = .01; % resampling period [s]
sa.opt.kinFilter.type = 'butter';
sa.opt.kinFilter.par  = [2 10/(100/2)]; %2 order 10Hz @ 100Hz sampling rate
saf = sa.dataFilter;

Plot sample of filtered EMG and KIN data

opt.fill = 1;
opt.figure = figure('NumberTitle','off','Name','Filtered EMG and KIN data',"Position",[0 0 700 650]);
plot(saf.data(1:8),opt) % frontal plane, 8 directions

filtered_data

Get phasic activity by subtracting tonic activity of the EMG signals.

saf = saf.emgPhasic;

Average data in groups of trials within same experimental condition. Define the common time range.

saf.opt.average.gr = saf.groupTrials('type3',[1:8]' );  % type3 is the target number  
saf.opt.average.trange = [-.3 .7]; % time interval in s before and after onset for averaging
sav = saf.average;

Normalize EMG and KIN data to the max of each channel. EMG are normalized between 0 and 2. KIN are normalized between -1 and 1.

sav.opt.normalize.type = 32;
sav.opt.normalize.nonnegch = sav.data(1).nonnegch;
sav = sav.normalize;

Plot a sample of averaged and normalizes EMG and KIN data

opt.figure = figure('NumberTitle','off','Name','Normalized EMG and KIN data',"Position",[0 0 700 650]);
plot(sav.data(1:8),opt)

normalized_data

Synergy Extraction

Define parameters for the extraction

sav.opt.find.algo = 'mmf';
sav.opt.find.N = [1:14];
sav.opt.find.nrep = 10;
sav.opt.find.niter = [100 10 10^-6 10000];
sav.opt.find.plot = 0;

Extract synergies

s1 = sav.find;

Plot $R^2$ curve for the best repetition of each set of synergies

s1.opt.plot.type = 'rsq';
plot(s1)

rsq

Select the number of synergies according to the default method, which chooses the first set that has an $R^2$ larger than 0.8.

Nsel = numsel(s1.syn);

Plot the selected set of synergies

s1.opt.plot.N = Nsel;
s1.opt.plot.type = 'W';
plot(s1)

syn

Plot original data and reconstructed data based on the synergy set chosen. The reconstructed signals are the black thick lines.

s1.opt.plot.type = 'rec';
s1.opt.plot.isect = [1:8];
plot(s1)

rec