DOC

Fuzzy Systems and Neural Networks Methods to Identify Hand and

By Stacy Kelly,2014-11-26 13:08
6 views 0
Fuzzy Systems and Neural Networks Methods to Identify Hand and

    Fuzzy Systems and Neural Networks Methods to Identify Hand

    and Finger Movements Using Surface EMG signals

     1123ABDELHAFID ZEGHBIB; FRANK PALIS; GEORGI TSENOV; NICOLA SHOYLEV; 2VALERI MLADENOV 1 Otto-von-Guericke-Universität Magdeburg, Institut für Elektrische Energiesysteme

    39016 Magdeburg, GERMANY

     2 Department of Theoretical Electrical Engineering, Technical University of Sofia,

    8 Kl. Ohridski St., BG-1000 Sofia, BULGARIA 3 Department of Electrical Engineering, University of Chemical Technology and Metallurgy,

    8 Kl. Ohridski St., BG-1756 Sofia, BULGARIA

    Abstract: - With help of exploitation of myoelectric signals, amputee persons can have a chance to improve their life with myoelectric prosthesis which are able to function with the amputee’s muscle movements. The

    myoelectric signal (MES) is the electrical manifestation of muscular contraction. This signal recorded at the surface of the skin of the forearm has been exploited to provide the recognition of the movement of the hand and finger Movements of healthy subject. The objective of the paper is first to describe the identification procedure, based on EMG patterns of forearm activity using Fuzzy logic and Neural Networks methods. Second to show the advantage of using features in Time-frequency domain, in comparison to those in time domain. Suitable features in time-frequency domain give high classification rates. Third is to compare between different intelligent computational methods of identification, which are used in this work: Multi-Layer Perceptron (MLP), Radial Basis Function Networks (RBF) and Learning Vector Quantization network (LVQ) as supervised methods and fuzzy Subtractive Clustering (FSC) as unsupervised method.

    Key-Words: - EMG signals, Hand and finger Movements Identification, Neural Networks, Fuzzy Logic, Subtractive clustering, Feature extraction, Short Time Fourier Transform.

    unsupervised methods, like Self Organizing Map, 1 Introduction

    Fuzzy Subtractive Clustering and Competitive MES classification is one of the most difficult

    Layer. Some features in time domain and time-pattern recognition systems because there usually

    frequency domain are extracted from raw EMG exist large variations in Electromyograph (EMG)

    signal and used for identification of movements features. The EMG signal has been used as a tool to

    with help of the above mentioned intelligent provide advanced man-machine interfaces [1],

    computational methods. In practice, determination rehabilitation of the handicapped people, functional

    of relevant features is very difficult. The aim of this electrical stimulation devices (FES) [2] and control

    paper is to distinguish between three finger commands for limb prostheses [3,4].

    movements (thumb, index and middle) and hand The classification problem may be divided into

    closing. three steps: signal presentation, feature extraction

     and pattern recognition. It is shown in this paper that classification performance of hand and finger

    2 EMG Signal Preprocessing movements depends significantly upon feature

    Surface muscle activity signals cannot be analysed extraction, which is very important to improve

    using classical methods, since they are non-considerably the accuracy of classification. Many

    stationary and have complex time-frequency researches proposed several EMG features for

    characteristics. EMG signals, fig 1, which are classification that showed good performance [5] [6]

    evolving in time in an unpredictable way (like a [7]. These identification methods belong to two

    speech signal or an EMG signal) require the notion categories, first are supervised methods, like Multi-

    of frequency analysis for each local time. Although Layer Perceptron, Radial Basis Networks, and

    frequency-domain representations such as the power Learning Vector Quantization network, Second are

spectrum of a signal often show useful information, subject’s arm is given in fig. 2, from the input

    these representations don’t show how the frequency feature space, the classifier must be able to classify content of a signal evolves over time. Time-the three output classes exploiting the EMG signals Frequency Analysis can identify not only the measurements.

    frequency content of a signal, but also how that For each channel the signal was acquired using a content evolves over time. single bipolar surface electrode pair. A differential

     amplifier with an isolated input is used. The signal

    was sampled at a rate of 4Khz using A/D board in

    an IBM PC/AT compatible microcomputer; this

    algorithm is developed with MATLAB 6 and is

    performed in a PC-based off-line process. The

    human subject was asked to produce a number of

    continuous movements, 34 single contraction

    periods are separated from the corresponding sets of

    continuous movements. Each single contraction

    period extracted from the raw signal by determined

    threshold is analysed with Short time Fourier

    Transform (STFT), which gives a measure of time and frequency information, fig 3 and fig 4, for small Fig. 1. Measured raw EMG signal from channel 2 segments of a signal. For the two channels we (extensor digitorum) muscle, and his absolute value. prepare some EMG training and test data, each class

    has 17 training and 17 test patterns. The four classes There are a number of different methods available

    labelled 1, 2, 3 and 4 have 68 train-samples and 68 for Time Frequency Analysis. Each type shows a

    test-samples. different time-frequency representation. The Short

     Time Fourier Transform (STFT), which is used in

    this paper, is the simplest TFA method and the

    easiest to compute.

3 Experimentation

     Fig. 3. Thumb: 400 ms EMG Signal analyzed with STFT

    method.

     Fig. 2, EMG training and test patterns recorded using

    two pairs of electrodes in Max Planck Institute

    laboratory in Magdeburg, Germany.

Four types of finger and hand movements to be

    classified are selected: thumb, pointer, middle and

    hand close. The placement of EMG surface

    electrodes on muscle groups is important to have

    more information about each movement. Two EMG

    surface electrodes are placed on two muscle groups,

    palnaris longus (channel_1) and extensor digitorum

     (channel_2), the locations of electrodes on the Fig. 4. Contour presentation for the same above signal

    5.1 Fuzzy system initialised with Subtractive 4 EMG Feature extraction

    Clustering method The problem of classification is the partitioning of

    The subtractive clustering algorithm was proposed the feature space, into regions (classes). Relevant

    by Chiu (1994). It estimates the number of clusters features will lead to high and accurate classification

    and the cluster centres in a set of data by an iterative rates. In the time domain, four features are extracted:

    procedure. The clusters obtained, with iterative Mean absolute value (MAV), Variance (VAR),

    optimisation-based clustering methods fuzzy c-Waveform length (WL), and Median Value (Med).

    Nmeans (fcm), are used to initialise the fuzzy sets, for 1MAVemg, (1) model identification method ANFIS. As initial kN1kmodel the first order Takagi-Sugeno (T. S.) model N12is used. We test some values of cluster radius, VARemg, (2) kN1 For these values we test which are: 0.3, 0.4 and 0.7.k1

    N1our model during 20 epochs using the four-extracted

    WLemgemg (3) k1ktime features, see table 1 for results. k1 Table 1: accuracy and number of correct instances for thWhere emg is the k sample data, which is N keach class with extracted time features.

    samples in length. The following Table shows us Classification accuracy (test data) % the worse classification with these features cluster Thumb Pointer Middle HC average

     radius

    In time-frequency domain using STFT (Short Time 0.3 52.94 23.52 0 5.88 20.58 Fourier Transform), the Hannaford’s moments of 0.4 58.82 17.64 0 17.64 23.52 first and second order are extracted as time-0.7 88.23 11.76 0 5.88 26.47 frequency features with the dominant frequency Cluster # of correct classified instances /17

    radius value FDV, which present the frequency value of

     Thumb Pointer Middle HC Total maximum amplitude obtained from spectral analysis.

    0.3 9 4 0 1 14 The nth moment of the frequency distribution at

    0.4 10 3 0 3 16 time t is defined as:

    0.7 15 2 0 1 18 n;;;;MtSTFTt,k (4) nk k

    In the same way we test, table 2, our features n: order, t: time, : frequency.

    extracted with help of time frequency analysis

    (STFT).

     5 Methods and results Table 2: accuracy and number of correct instances for There is a large set of neural networks and fuzzy each class with extracted time-freq features. logic methods in the literature addressing Classification accuracy (test data) % identification problems. Some of these methods are Cluster Thumb Pointer Middle HC average applied, using different features, to identify our four radius fingers and hand movements and to compare their 0.3 82.35 70.58 76.47 82.35 77.94 performances. These methods in case of supervised 0.4 88.23 64.70 76.47 82.35 77.94 learning employ optimisation techniques to 0.6 94.11 76.47 100 94.11 91.17 processes the inputs and compare their resulting 0.7 94.11 76.47 100 82.35 88.23 outputs against the desired outputs. Errors are then

    calculated, causing the system to adjust the Cluster # of correct classified instances /17 parameters. In case of unsupervised learning, radius

    training algorithms attempt to locate clusters in the Thumb Pointer Middle HC Total

    0.3 14 12 13 14 53/68 input data, which approximate the distribution of

    0.4 15 11 13 14 53/68 the data. More details about clustering data in

    0.6 16 13 17 16 62/68 [Kaufman & Rousseeuw, 1990].

    0.7 16 13 17 14 60/68

    5.2 Neural Networks Methods

    Various ANN based models were applied to identify We test our MLP Network with different number of these different four movement classes. Hykin neurons in hidden layer: 5, 10, 20 and 50 neurons publishes a comprehensive foundation for the study for four extracted time features, (IEMG, WL, VAR of Neural Networks [8]. There are many functions and Med, see table 4. It’s obvious in table 4 that and variables to be determined for each neural there is no effect of the number of neurons in network model. The four neural networks methods hidden layer on the rate of classification. The are trained during 100 epochs. increasing in number of neurons in hidden layer

     doesn’t enhance the accuracy.

    In case of extracted time-frequency features, we 5.2.1 Multi Layer Perceptron (MLP)

    This network is used in many different types of obtained the results, which are resumed in table 5. applications. This architecture has a large class of

    Table 5: network types with many different topologies and

     Classification accuracy (test data) % training methods. The number of neurons in the

    # of Thumb Pointer Middle HC average only hidden layer is determined based on their

    neurons performance in training process. For the one-neuron 5 100 52.94 94.11 94.11 85.29 output-layer we use log sigmoid transfer function 10 100 76.47 100 94.11 92.64 “logsig”, which gives an output in the range of 0 to 20 88.23 41.17 94.11 94.11 79.41 1. Our output range between 0 and 1 will be divided 50 94.11 64.70 70.58 94.11 80.88 in four ranges, since we have four classes to be identified, see Table 3. # of # of correct classified instances /17 neurons Table 3: Thumb Pointer Middle HC Total Classes Target Output Type of 5 17 9 16 16 58/68 output range Movement 10 17 13 17 16 63/68 Classe 1 0.125 0 0.25 Thumb 20 15 7 16 16 54/68 Classe 2 0.375 0.25 0.5 Pointer 50 16 11 12 16 55/68 Classe 3 0.625 0.5 0.75 Middle Classe 4 0.875 0.75 - 1 Hand close 5.2.2 Radial basis Functions (RBF) The classification results are summarized in table4. The RBF Network is a one hidden layer neural Network with several forms of radial basis Table 4: Rate of classification and correct classified activation functions, like Gaussian function. We use instances (extracted time features) for each class and

    the method, which creates neurons one at a time. In average value with four different hidden layer neurons

    each iteration the input vector is used to create a number.

     Classification accuracy (test data) % new neuron. The error of the new network is

    # of Thumb Pointer Middle HC average checked, and if it is not low enough the next neuron

    neurons is added. This procedure is repeated until the error

    5 17.64 35.29 17.64 64.70 33.82 goal is met, or the maximum number of neurons is

    10 35.29 11.76 29.41 29.41 26.47 reached. The output layer is linear and the rate of

    20 52.94 0 5.88 41.17 25 classification is determined by the spread of the 50 41.17 23.52 5.88 52.94 30.88 hidden unit. We give many values between 0.5 and 2.5 with a step of 0.2 to find the optimal value of # of # of correct classified instances /17 spread, which is in our application equal to 0.6 and neurons 0.7, see fig. 5. Thumb Pointer Middle HC Total

    5 3 6 3 11 23/68

    10 6 2 5 5 18/68

    20 9 0 1 7 17/68

    50 7 4 1 9 21/68

    5.2.3 Learning Vector Quantization (LVQ)

    Learning Vector Quantization [9, 10] networks can

    classify, faster than other neural network techniques

    like Back Propagation, any set of input vectors; not

    only linearly separable sets of input vectors. Its

    architecture resembles to that of unsupervised

    competitive learning network, except that each

    output is assigned to a target class and works in two

    steps. First it uses an unsupervised data clustering

    method to locate several clusters. Second it

    optimises the cluster centres. The number of clusters can be specified a priori or determined via Fig. 5. Average accuracy according to spread’s values cluster techniques. It is able to reduce large data sets to a smaller number of codebook vectors (cluster For this value of spread we test our network using centres) suitable for data compressing. the four extracted time features, see table 6 for LVQ network used in this work has 4 neurons in the results. first competitive layer and one neuron for each class in the second linear layer. In comparison with Table 6: accuracy and number of correct instances for previous methods, LVQ needs only 30 to 40 epochs each class with extracted time features.

    to converge, see table 8 and 9. Thumb Pointer Middle HC average

     Acuracy 29.41 11.76 23.52 76.47 35.29

    Table 8: Rate of classification (time extracted features) (Test) %

    for each class and average value.

     Classification accuracy (test data) % # of correct classified instances /17

    competitive Thumb PointerMiddle HC average Thumb Pointer Middle HC Total

     neurons Correct 5 2 4 13 24 / 68

    6 76.47 0 5.88 41.17 30.88 (# / 17)

    12 64.70 0 11.76 58.82 33.82

    28 70.58 0 5.88 64.70 35.29 With the same method we test the classification

     accuracy for the extracted time-frequency features with two values of spread: 0.7 and 1.5, see table 7. Table 9: Rate of classification (time-frequency extracted features) for each class and average value. Table 7: accuracy and number of correct instances Classification accuracy (test data) % (extracted features in time-frequency analysis domain) competitive ThumbPointer MiddleHC average for each class and average value with different values of neurons spread. 6 100 64.70 94.11 88.23 86.76 Classification accuracy (test data) % 12 100 64.70 94.11 88.23 86.76 Spread Thumb Pointer Middle HC average 28 100 64.70 100 88.23 88.23 value

     0.7 88.23 88.23 76.47 82.35 83.82

     1.5 76.47 70.58 64.70 76.47 72.05

    6 Methods comparison

    Spread # of correct classified instances /17 To resume the results obtained with the features value extracted in time-frequency analysis, we present for

     Thumb Pointer Middle HC Total all methods their classification rate according to 0.7 15 15 13 14 57/68 some determined parameters. In fig. 6 four first 1.5 13 12 11 13 49/68 values of classification accuracy, with MLP method,

     are presented according to number of neurons used

    We get the best rate classification for Gaussian in hidden layer, which are: 5, 10, 20 and 50 neurons.

    functions with spread value of 0.7 by extracted The second four values group given by

    features in time-frequency analysis domain. unsupervised fuzzy subtractive clustering method corresponding to the different values of cluster

radius, 0.3, 0.4, 0.6 and 0.7. The two classification recognition as a control command of man

    accuracy values obtained with RBF method machine interface”. Med Eng Phys. 1996 corresponding to 0.7 and 1.5 spread parameter Oct;18(7):529-37. PMID: 8892237 [PubMed

    values. Finally LVQ method shows three rate of indexed for MEDLINE].

    classification, corresponding to three values of [2] Hefftner G, Jaros G, “The electromyogram number of neurons (6, 12 and 28) in competitive (EMG) as a control signal for functional layer. neuromuscular, Stimulation, part I:

    Autoregressive modeling as a mean of EMG

    signature discrimination”. IEEE Trans Biomed

    Eng 1988, 35(6), pp. 22835.

     [3] Zardoshti-Kermani M, Wheeler BC, Badie K,

    Hashemi RM, “EMG feature evaluation for

    movement, control of upper extremity

    prostheses”. IEEE Trans Rehabil Eng

    1995;Vol.3, No.4, pp. 324333.

     [4] C. Bonivento, A. Davalli, C. Fantuzzi, R.

    Sacchetti, S. Terenzi, “Automatic tuning of

    myoelectric prostheses” Journal of Rehabilitation Research and Development Vol. Fig. 6. Classification accuracy comparison between four 35 No. 3, July 1998, pp. 294-304. different methods using extracted features in time 5] K.Englehart, B. Hudgin, and P. A. Parker, “A [frequency domain analysis.

     wevelet based continuous classification scheme for

    multifunction myoelectric control”, IEEE

    Transactions on Biomedical Engineering, vol. 7. Conclusion 48, no. 3, pp. 302-311, Mars 2001. As conclusion, first we can conclude that the [6] Z. K. Mahyar, W. C. Bruce, B. Kambiz, and M. identification methods cannot and don’t help to H. Reza, EMG feature evaluation for perform accuracy classification if the feature movement control of Upper extremity measures selected are not a relevant features. For prostheses”, IEEE Transactions on this reason the determination of a complete set of Rehabilitation Engineering, Dec 1995. discriminatory features is very important. [7] B. Hudgins, P. Parker and R. N. Scott, “New Both methods Neural Networks and Fuzzy Logic strategy for multi-function myoelectric are showed clearly that the selection of the feature control ”IEEE Transactions on Biomedical are of great importance to enhance the recognition Engineering, vol. 40, no. 1, pp. 82-94, Jan. rate of myoelectric patterns for the four movements, 1993. hand close and three finger (Thumb, pointer and [8] S. Haykin, Neural Networks: A Comprehensive middle). In this case the use of time-frequency Foundation, (2nd Edition 1998), ISBN: domain as features extraction domain is necessary 0132733501. Prentice Hall, 1998. to perform the identification. [9] T. Kohonen. Self organization and associate memory. Springer-Verlag, London, 3rd edition, 1989. Acknowledgement [10] T. Kohoenen. Improved versions of learning This work is supported by the DAAD PPP-program vector quantization. In International Joint 2005-2006 between Otto-von-Guericke-Universität conference on Neural Networks. Vol:1, p:545-Magdeburg, Germany, and University of Chemical 550, San Diego, 1990. Technology and Metallurgy, Sofia, Bulgaria.

References:

     [1] Chang G-C, Kang W-J, Luh J-J, Cheng C-K,

    Lai J-S, Chen J-JJ, Kuo T-S, “Real-time

    implementation of electromyogram pattern

Report this document

For any questions or suggestions please email
cust-service@docsford.com