Medical Engineering & Physics 36 (2014) 17161720

Contents lists available at ScienceDirect

Medical Engineering & Physics

journal homepage: www.elsevier.com/locate/medengphy

Technical note

Feature dimensionality reduction for myoelectric pattern recognition:

A comparison study of feature selection and feature projection
methods
Jie Liu
Sensory Motor Performance Program, Rehabilitation Institute of Chicago, 345 E. Superior St, Suite 1443, Chicago, IL 60611, USA

a r t i c l e

i n f o

Article history:
Received 27 August 2014
Accepted 13 September 2014
Keywords:
Myoelectric pattern recognition
Surface EMG
Feature selection
Feature projection

a b s t r a c t
This study investigates the effect of the feature dimensionality reduction strategies on the classication
of surface electromyography (EMG) signals toward developing a practical myoelectric control system.
Two dimensionality reduction strategies, feature selection and feature projection, were tested on both
EMG feature sets, respectively. A feature selection based myoelectric pattern recognition system was
introduced to select the features by eliminating the redundant features of EMG recordings instead of
directly choosing a subset of EMG channels. The Markov random eld (MRF) method and a forward
orthogonal search algorithm were employed to evaluate the contribution of each individual feature to
the classication, respectively. Our results from 15 healthy subjects indicate that, with a feature selection analysis, independent of the type of feature set, across all subjects high overall accuracies can be
achieved in classication of seven different forearm motions with a small number of top ranked original
EMG features obtained from the forearm muscles (average overall classication accuracy >95% with 12
selected EMG features). Compared to various feature dimensionality reduction techniques in myoelectric
pattern recognition, the proposed lter-based feature selection approach is independent of the type of
classication algorithms and features, which can effectively reduce the redundant information not only
across different channels, but also cross different features in the same channel. This may enable robust
EMG feature dimensionality reduction without needing to change ongoing, practical use of classication
algorithms, an important step toward clinical utility.
2014 IPEM. Published by Elsevier Ltd. All rights reserved.

1. Introduction
Robotic devices are considered as the leading interactive rehabilitation systems available. Robotic rehabilitation is an effective
platform for sensorimotor training in people with different neurological injuries such as hemiparetic stroke, cerebral palsy, multiple
sclerosis, Parkinsons disease or spinal cord injury [17]. In contrast, advances have been made to build lighter, stronger and more
versatile upper-limb powered robotic devices [813], relatively
little progress has been made on improving the intuitive control of robotic exoskeletons. A neural control interface is crucial
to providing accurate, intuitive control of upper limb exoskeletons. Among the potential biological signals for humanmachine
interaction (brain, nerve, and muscle signals), electromyography
(EMG), directly reect the human motion intention, may be the only

Tel.: +1 312 238 1474.

E-mail address: jie.liu2009@gmail.com
http://dx.doi.org/10.1016/j.medengphy.2014.09.011
1350-4533/ 2014 IPEM. Published by Elsevier Ltd. All rights reserved.

experimentally non-invasive record of the motor commands to the

muscles that allows applications in routine clinical use [14,15].
Myoelectric control strategies based on pattern recognition have
recently attracted much attention in the development of neural
control systems [1622], which infers great promise for multifunctional exoskeleton control.
To increase dexterity of myoelectric control, multiple features
extracted from each EMG channel are concatenated into a feature
vector and are used for pattern recognition [5,16,1820,2328].
Dimensionality reduction of the feature vector is then needed to
improve computational efciency. To date, previous studies have
tended to focus on channel reduction for reducing the complexity
of myoelectric pattern recognition [17,2931]. However, an evaluation study is necessary to determine the most adequate technique
to be employed in dimensionality reduction.
The existing dimensionality reduction methods are roughly
categorized as either feature selection or feature projection.
Generally, feature projection attempts to determine signicant
transformed features in a new space after the original feature

J. Liu / Medical Engineering & Physics 36 (2014) 17161720

1717

in the measurement space are transformed into the dimensionreduced space utilizing specied transformation, such as principal
components analysis (PCA) [25] and uncorrelated linear discriminant analysis (ULDA) [32]. In contrast, feature selection attempts
to determine the best subset of the original feature set. Feature selection reduces the dimension of features, computation
load and improves classication performance. In general, feature
selection approaches can be grouped into two categories: lter
methods and wrapper methods. The lter methods select features from the original feature space with a given evaluation
criterion which is independent of the type of classier, such as distance measures which reect how well the classes separate from
each other. While, the wrapper methods choose a feature subset with high classication performance estimated by a specied
classier.
The objective of this study was to preliminarily assess the
effect of dimensionality reduction strategies on the surface EMG
classication performance. Two dimensionality reduction strategies, feature selection and feature projection, were preliminarily
tested on classical EMG feature sets, respectively. This study
focused on the exploration of various feature dimensionality
reduction techniques with an emphasis on feature selection
techniques.

The surface EMG recording resulted in n-dimensional feature

vectors in one analysis window (n = 32 for TD feature set; n = 56 for
AR + RMS feature set, TD feature set resulting in a four-dimension
feature per EMG channel, while AR + RMS feature set resulting in
a seven-dimension feature per EMG channel, there are eight channels in total for this experiment) for pattern classication. Two well
known feature projection methods, principal components analysis
(PCA) [25] and uncorrelated linear discriminant analysis (ULDA)
[32], were employed to reduce the feature dimensionality respectively in this study. PCA and ULDA have shown their effectiveness
to reduce the feature dimensionality in myoelectric pattern classication [24,30,33].

2. Methods

2.4. Feature selection

2.1. Data description

In contrast to previous studies [17,2931], except for using feature projection for dimension reduction, feature selection was also
applied to reduce the complexity of the signal processing. Previous
studies have shown that classication accuracy is affected more by
the choice of feature set than by the choice of classier [27,29].
To solve the general feature subset selection problem in myoelectric control, two relatively new lter methods the Markov random
eld (MRF) based FisherMarkov selector [38] and the forward
orthogonal search algorithm, FOS-MOD [39] were examined in this
study to select the most appropriate original surface EMG features,
respectively. Because the lter methods estimate the classication
performance by some indirect assessments which reect how well
the classes separate from each other, making the approach more
exible due to the fact that the selected features are useful for
general classiers.
The mathematical details of the MRF and the FOS-MOD feature selection methods were described in [38,39]. In brief, the
MRF method aims to select the best subset of candidate features that maximizes the between-class distance while minimizing
the within-class distance in a higher dimensional kernel space.
The MRF method efciently selects the globally optimal subset
of features, which are the most useful in characterizing differences among the possible classes [38]. The FOS-MOD algorithm
[39] is a forward orthogonal search feature selection algorithm by
maximizing the overall dependency to nd signicant variables,
which also provides a rank list of selected features ordered according to the percentage contribution for representing the overall
structures.

The data used in this work are identical to the data in a previous
study [33]. Eight-channel EMG data were collected from forearm
muscles and the biceps muscle of 15 healthy subjects using AgAgCl electrodes with a reference electrode placed on the wrist. The
surface EMG signals were sampled at 3000 Hz per channel. Subjects
performed seven forearm motions: hand open, hand close, supination, pronation, wrist exion, wrist extension and rest. EMG data
used in this study were collected in ve trials nished by each subject. Each motion was repeated four times with duration of three
seconds within each data collection trial. The order of these motions
was randomized.
2.2. Feature extraction and pattern recognition
The EMG recordings for each distinct motion were composed
of four active segments corresponding to four repetitions of the
motion for each data collection trial. By visual inspection, active
segments were extracted from EMG data in between manually
located onset and offset times, which were the same for all the
eight channels. A sliding window scheme was used to facilitate
both utilization of limited data stream and continuity of decision output by a classier. Specically, for each active segment,
eight-channel EMG data were further divided into a series of sliding analysis windows (window length: 256 ms, overlapping step:
32 ms).
For each analysis window, a set of features was extracted to
characterize the EMG data for classication of the forearm motions.
The feature set was extracted on each of the eight EMG channels
and then concatenated into a feature vector. In order to examine
if the performance of the feature selection technique would not
depend on the feature set, two classical feature sets were investigated in this study including the time domain (TD) feature set [28],
which comprised of four time domain statistics including the mean
absolute value (MAV), the number of zero crossings (ZC), the waveform length (WL), and the number of slope sign changes (SSC); the
combination of autoregressive (AR) model (six-order as suggested
by Farina and Merletti [34]) coefcients and the root mean square

(RMS) of the signal as a feature set (AR + RMS) [35]. These feature
sets have been shown to be effective signal representations for EMG
pattern recognition with relatively low computational complexity
[1619,30,31,36].
Three classiersthe linear discriminant analysis (LDA) classier [16], the k-nearest neighbor (KNN) classier [37] and the
support vector machine (SVM) [18] were used respectively in the
current study.
2.3. Feature projection

2.5. Performance evaluation

A post-processing method, namely majority vote [16,23] was
used to produce consistent improvements in classication accuracies. A ve-fold cross-validation scheme was used to evaluate
classication performance. Training with a different combination
of the EMG data within four trials were assigned as training dataset,
and sequentially the EMG data of the remaining trial were used as
testing dataset. The overall classication accuracy (CA) was then
calculated as the percentage of correctly classied windows over

J. Liu / Medical Engineering & Physics 36 (2014) 17161720

100

100

90

90

Classification accuracy(%)

Classification accuracy(%)

1718

80
70
ULDA TD
PCA TD
ULDA AR+RMS
PCA AR+RMS

60
50
40
30

80
70
MRF TD
FOS-MOD TD
MRF AR+RMS
FOS-MOD AR+RMS

60
50
40
30

5
10
15
Number of feature dimensions

20
0

20

Fig. 1. The effect of feature set dimension upon the classication accuracy, when
using ULDA and PCA dimensionality reduction, respectively. The accuracy, averaged
across 15 subjects, is shown for the TD feature set and AR + RMS feature set. Results
are given for the LDA classier. Note that ULDA can yield six projected features for
seven classes at most in this study.

all the analysis windows in the testing datasets across all motion
patterns.

5
10
15
Number of selected features

20

Fig. 2. The change in average classication performance with the number of applied
original features in a feature set (TD and AR + RMS feature set) selected using MRF
and FOS-MOD feature selection, respectively.

3.3. Preliminary comparison of feature projection and feature

selection

Pre =

TP
(TP+FP)

(2)

A comparison was made between feature projection and selection on the dataset. Because ULDA is more efcient than PCA for
feature projection (Fig. 1) and the MRF method achieved similar
performance in term of feature selection compared to the FOS-MOD
method (Fig. 2), but the FOS-MOD algorithm only ranks features
that are most representative of the feature data. The MRF algorithm
was tested against ULDA. The comparison between the two methods reveals that the MRF method achieves better performance than
ULDA, as shown in Fig. 3. It is noted that ULDA has higher accuracies
when the dimension of feature subset is less than eight, because the
selected projected features are obtained by ULDA from all the original features. It cannot be argued that projected features by ULDA
are more concentrated with relevant information than the original
features obtained from the MRF selector.

Sen =

TP
(TP+FN)

(3)

3.4. Performance of different features and classiers

number of accurately classied testing data

100%
total number of applied testing data

(1)

The use of precision (Pre) and sensitivity (Sen), which mainly

take into account the false alarm and false negative errors for each
class, respectively, was suggested by a previous study [40]. Thus,
the performance of feature selection based classication was also
assessed by computing the Pre and Sen for each class based on the
confusion matrix. The confusion matrix quanties the number of
instances in the test dataset classied as false positive (FP), true
positive (TP), false negative (FN) and true negative (TN). The Pre
and Sen are dened as

3. Results
3.1. The effect of feature projection on performance
Fig. 1 shows the effect of the number of feature dimensions on
the classication performance averaged across all subjects, using
a combination of the AR + RMS or TD feature set and the LDA classier, where the feature dimensions were reduced via the ULDA
and PCA methods, respectively. The comparison between the two
methods indicates that ULDA is more efcient than PCA in reducing
feature dimensions. The projected features obtained by ULDA are
more concentrated with relevant information than the projected
features obtained via PCA.
3.2. The effect of feature selection on performance
The MRF and FOS-MOD feature selection algorithms were tested
on the AR + RMS and TD feature sets, respectively. Fig. 2 demonstrates the average classication accuracy as a function of the
number of selected original features. It was observed that as features are added, the classication accuracy increases, independent
of the type of feature set. However, the increase is not linear, which
quickly reaches a high value as the rst few features are added then
slowly approaches a maximum. The results indicated that it is feasible to greatly reduce the number of features while maintaining high
classication accuracy. The two methods achieved similar results.

To evaluate the classication performance resulting from using

different selected feature sets and classiers, a preliminary comparison of the LDA, KNN and SVM classiers in combination with
the TD and AR + RMS feature sets was undertaken. Table 1 shows the
results of a comparison of classication performance obtained by
using various feature sets and classiers. Across all subjects, the 12
original TD features selected from 32 available TD features using

100
90

Classification accuracy(%)

CA =

80
70
MRF TD
ULDA TD
MRF AR+RMS
ULDA AR+RMS

60
50
40
30

10
15
Number of features

20

Fig. 3. Comparison of the classication performance using the original features

selected by MRF and the transformed features via ULDA averaged across 15 subjects,
respectively.

J. Liu / Medical Engineering & Physics 36 (2014) 17161720

1719

Fig. 4. Class-to-class confusion matrices derived from the EMG data after using the MRF criterion for feature selection. Class 1 (hand open), Class 2 (hand close), Class 3
(supination), Class 4 (pronation), Class 5 (wrist exion), Class 6 (wrist extension), Class 7 (rest). The precision (Pre), the sensitivity (Sen) for each class and the overall accuracy
for each class are also reported, respectively.

the MRF criterion achieved approximately above 95.7% average

overall classication accuracy. In contrast, the 12 original AR + RMS
features selected from 56 available AR + RMS features using the
MRF criterion achieved approximately above 95.6% average overall
classication accuracy. The MRF feature selection method enables
robust EMG feature dimensionality reduction without needing to
change ongoing, practical use of classication algorithms.
Fig. 4 shows class-to-class classication results in the form of
confusion matrices derived from one subject using TD feature set
and AR + RMS feature set, respectively when an LDA classier was
used. This gure also illustrates three statistical indices, namely
Pre, Sen and overall accuracy, based on the confusion matrices. The
overall accuracy was slightly different from the averaged Pre over
all movement patterns, because of slight variation in the number
of testing windows among movement patterns [26].
4. Discussion and conclusion
The current study focused on the feasibility analysis of selecting the most discriminative features from all extracted features
of the EMG signals for implementing a practical myoelectric control system. This was achieved by selecting the most informative
features using feature selection instead of feature projection. The
EMG feature selection approaches based on the MRF and FOS-MOD
algorithms were developed to select the features by eliminating
redundant features for EMG pattern recognition. The approach
evaluates the contribution of individual features to predicting forearm motions by computing the classication accuracies using the
selected features. The approach has three advantages compared
to the classical channel selection approach [17,2931]. First, it is
independent of the type of classication algorithm used in classication stage. Second, instead of focusing on individual channels, the
method examined each original feature and therefore eliminated
redundant information for different features in the same channel
or across different channels. Third, in contrast to transforming all
Table 1
Overall all pattern recognition results across ve-fold cross-validation (mean SD),
averaged across 15 subjects using MRF selected features (12 out 56 AR + RMS features, 12 out of 32 TD features, Unit: %).
Feature set

LDA

KNN

SVM

TD
AR + RMS

95.7 6.7
95.6 4.1

99.2 1.9
97.2 3.4

98.1 2.5
97.5 4.2

original features into the dimension-reduced feature space using

ULDA or PCA, the method selected the original features directly.
While there are many studies of feature extraction and feature evaluation for myoelectric control [18,2022,25,28,34,35], but
there are few studies stress the exploration of various feature
dimensionality reduction techniques. A previous study compared
feature selection and feature projection (employing PCA) on EMG
patterns and proved that PCA outperformed the feature selection
method based on Euclidean distance class separability criterion
in EMG pattern recognition [24]. But the recent development of
feature selection facilitates more powerful tools to select most discriminative features from all extracted features of signals. Thus, it
is necessary to reevaluate the feature selection approach for EMG
pattern recognition. The effect of feature selection/projection techniques on EMG pattern recognition needs further investigation.
The wrapper methods for feature selection is classier dependent. Furthermore, interpreting weights from trained classiers like
LDA can yield high coefcients to features that are not correlated
to the classes of interest. Therefore, some features may be weighed
higher in order to remove uncorrelated components from other features which may contain relevant information [41]. In contrast, the
lter methods were examined in this study. As the lter methods
work without needing to obtain the feedback from ongoing classication algorithms such as LDA, the lter methods are robust against
such a weakness.
The use of the MRF and FOS-MOD-based lters for EMG feature selection was evaluated in this study. The performance of each
EMG feature set (TD or AR + RMS) was determined in the context
of feature selection and feature projection based dimensionality
reduction, respectively. In contrast to earlier ndings, the feature
selection methods MRF and FOS-MOD outperformed feature projection by ULDA and PCA in reducing dimensionality (Figs. 13).
There are many feature selection/projection techniques available
now. Advanced feature projection techniques, such as iPCA [25],
may be superior to the feature selection techniques presented in
this paper in the context of EMG pattern recognition.
Our results from 15 healthy subjects indicate that with a feature selection analysis, across all subjects high overall accuracies
can be achieved in classication of seven different forearm motions
with a small number of top ranked original EMG features obtained
from the forearm muscles (average overall classication accuracy
>95% with 12 selected EMG features, across all classiers and feature sets), suggesting that a small number of highly discriminative
features contribute the most information in discriminating among

1720

J. Liu / Medical Engineering & Physics 36 (2014) 17161720

the motions of interest can maintain classication performance

comparable to that from all the EMG features (Table 1).
In conclusion, this study introduces an approach utilizing the lter methods based on relatively new feature selection techniques,
the MRF and FOS-MOD algorithms, to select the most informative
features to eliminate redundant features in the same channel or
across different channels and the selected features can represent
neuromuscular control information. The encouraging results indicate that compared to feature projection, it is feasible to greatly
reduce the number of original features while maintaining high
classication accuracy, thereby reduce the complexity of the signal processing and facilitate real-time application. This study also
demonstrates that the performance of the feature selection techniques used in this study does not depend on the type of feature set.
These lter-based feature selection methods enable robust EMG
feature dimensionality reduction without needing to change ongoing, practical use of classication algorithms, an important step
toward clinical utility. In short, except for feature projection, effective feature selection allows for a possible solution to the feature
dimensionality reduction in myoelectric pattern recognition.
Funding
None declared.
Ethical approval
None declared.
Acknowledgments
The authors would like to thank Dr. Adrian Chan at Carleton
University, Canada, for providing the EMG data set.
Conict of interest
None declared.
References
[1] Krebs HI, Dipietro L, Levy-Tzedek S, Fasoli SE, Rykman-Berland A, Zipse J,
et al. A paradigm shift for rehabilitation robotics. IEEE Eng Med Biol Mag
2008;27:6170.
[2] Marchal-Crespo L, Reinkensmeyer DJ. Review of control strategies for robotic
movement training after neurologic injury. J NeuroEng Rehabil 2009:6.
[3] Van Hedel HJA, Dietz V. Rehabilitation of locomotion after spinal cord injury.
Restor Neurol Neurosci 2010;28:12334.
[4] Reinkensmeyer DJ, Boninger ML. Technologies and combination therapies for
enhancing movement training for people with a disability. J NeuroEng Rehabil
2012:9.
[5] Zariffa J, Kapadia N, Kramer JLK, Taylor P, Alizadeh-Meghrazi M, Zivanovic V,
et al. Feasibility and efcacy of upper limb robotic rehabilitation in a subacute
cervical spinal cord injury population. Spinal Cord 2012;50:2206.
[6] Gordon AL, di Maggio A. Rehabilitation for children after acquired brain injury:
current and emerging approaches. Pediatr Neurol 2012;46:33944.
[7] Frisoli A, Procopio C, Chisari C, Creatini I, Bonglio L, Bergamasco M, et al. Positive effects of robotic exoskeleton training of upper limb reaching movements
after stroke. J NeuroEng Rehabil 2012:9.
[8] Nef T, Mihelj M, Riener R. ARMin: a robot for patient-cooperative arm therapy.
Med Biol Eng Comput 2007;45:887900.
[9] Perry JC, Rosen J, Bums S. Upper-limb powered exoskeleton design. IEEEASME
Trans Mechatron 2007;12:40817.
[10] Gupta A, OMalley MK, Patoglu V, Burgar C. Design, control and performance
of ricewrist: a force feedback wrist exoskeleton for rehabilitation and training.
Int J Robot Res 2008;27:23351.
[11] Mao Y, Agrawal SK. Design of a cable-driven arm exoskeleton (CAREX) for
neural rehabilitation. IEEE Trans Robot 2012;28:92231.

[12] Kim H, Miller LM, Byl N, Abrams GM, Rosen J. Redundancy resolution of
the human arm and an upper limb exoskeleton. IEEE Trans Biomed Eng
2012;59:17709.
[13] Ren YP, Kang SH, Park HS, Wu YN, Zhang LQ. Developing a multi-joint upper
limb exoskeleton robot for diagnosis, therapy, and outcome evaluation in neurorehabilitation. IEEE Trans Neural Syst Rehabil Eng 2013;21:4909.
[14] Jiang N, Falla D, dAvella A, Graimann B, Farina D. Myoelectric control in neurorehabilitation. Crit Rev Biomed Eng 2010;38:38191.
[15] Scheme E, Englehart K. Electromyogram pattern recognition for control of powered upper-limb prostheses: state of the art and challenges for clinical use. J
Rehabil Res Dev 2011;48:64359.
[16] Englehart K, Hudgins B. A robust, real-time control scheme for multifunction
myoelectric control. IEEE Trans Biomed Eng 2003;50:84854.
[17] Zhou P, Lowery MM, Englehart KB, Huang H, Li G, Hargrove L, et al. Decoding
a new neural-machine interface for control of articial limbs. J Neurophysiol
2007;98:297482.
[18] Oskoei MA, Hu H. Support vector machine-based classication scheme
for myoelectric control applied to upper limb. IEEE Trans Biomed Eng
2008;55:195665.
[19] Hargrove LJ, Scheme EJ, Englehart KB, Hudgins BS. Multiple binary classications via linear discriminant analysis for improved controllability of a powered
prosthesis. IEEE Trans Neural Syst Rehabil Eng 2010;18:4957.
[20] Boostani R, Moradi MH. Evaluation of the forearm EMG signal features for the
control of a prosthetic hand. Physiol Meas 2003;24:30919.
[21] Delis AL, Carvalho JLA, da Rocha AF, Ferreira RU, Rodrigues SS, Borges GA. Estimation of the knee joint angle from surface electromyographic signals for active
control of leg prostheses. Physiol Meas 2009;30:93146.
[22] Nazarpour K, Sharafat AR, Firoozabadi SM. Application of higher order statistics to surface electromyogram signal classication. IEEE Trans Biomed Eng
2007;54:17629.
[23] Chan ADC, Englehart KB. Continuous myoelectric control for powered prostheses using hidden Markov models. IEEE Trans Biomed Eng 2005;52:1214.
[24] Englehart K, Hudgins B, Parker PA, Stevenson M. Classication of the myoelectric signal using time-frequency based representations. Med Eng Phys
1999;21:4318.
[25] Hargrove LJ, Li G, Englehart KB, Hudgins BS. Principal components analysis
preprocessing for improved classication accuracies in pattern-recognitionbased myoelectric control. IEEE Trans Biomed Eng 2009;56:140714.
[26] Huang H, Kuiken TA, Lipschutz RD. A strategy for identifying locomotion modes
using surface electromyography. IEEE Trans Biomed Eng 2009;56:6573.
[27] Huang Y, Englehart KB, Hudgins B, Chan ADC. A Gaussian mixture model based
classication scheme for myoelectric control of powered upper limb prostheses. IEEE Trans Biomed Eng 2005;52:180111.
[28] Hudgins B, Parker P, Scott RN. A new strategy for multifunction myoelectric
control. IEEE Trans Biomed Eng 1993;40:8294.
[29] Hargrove LJ, Englehart K, Hudgins B. A comparison of surface and intramuscular
myoelectric signal classication. IEEE Trans Biomed Eng 2007;54:84753.
[30] Liu J, Zhou P. A novel myoelectric pattern recognition strategy for hand function
restoration after incomplete cervical spinal cord injury. IEEE Trans Neural Syst
Rehabil Eng 2013;21:96103.
[31] Zhang X, Zhou P. High-density myoelectric pattern recognition toward
improved stroke rehabilitation. IEEE Trans Biomed Eng 2012;59:164957.
[32] Ye J, Li T, Xiong T, Janardan R. Using uncorrelated discriminant analysis for
tissue classication with gene expression data. IEEE/ACM Trans Comput Biol
Bioinform 2004;1:18190.
[33] Chan ADC, Green GC. Myoelectric control development toolbox. In: Proceedings
of the 30th conference of the Canadian medical & biological engineering society; 2007.
[34] Farina D, Merletti R. Comparison of algorithms for estimation of EMG
variables during voluntary isometric contractions. J Electromyogr Kinesiol
2000;10:33749.
[35] Graupe D, Salahi J, Kohn KH. Multifunctional prosthesis and orthosis control
via microcomputer identication of temporal pattern differences in single-site
myoelectric signals. J Biomed Eng 1982;4:1722.
[36] Lee SW, Wilson KM, Lock BA, Kamper DG. Subject-specic myoelectric pattern
classication of functional hand movements for stroke survivors. IEEE Trans
Neural Syst Rehabil Eng 2011;19:55866.
[37] Cover TM, Hart PE. Nearest neighbor pattern classication. IEEE Trans Inf Theory
1967;13:217.
[38] Cheng Q, Zhou H, Cheng J. The FisherMarkov selector: fast selecting maximally
separable feature subset for multiclass classication with applications to highdimensional data. IEEE Trans Pattern Anal Mach Intell 2011;33:121733.
[39] Wei HL, Billings SA. Feature subset selection and ranking for data dimensionality reduction. IEEE Trans Pattern Anal Mach Intell 2007;29:1626.
[40] Rojas-Martinez M, Mananas MA, Alonso JF, Merletti R. Identication of isometric contractions based on high density EMG maps. J Electromyogr Kinesiol
2013;23:3342.
[41] Haufe S, Meinecke F, Gorgen K, Dahne S, Haynes JD, Blankertz B, et al. On the
interpretation of weight vectors of linear models in multivariate neuroimaging.
Neuroimage 2014;87:96110.