![](bg7.png)
order to find the best configuration. Our procedure is
motivated by the relative limited dataset, which could
be increased for example by the use of data augmen-
tation methods. Indeed, in a classical way, it is im-
perative to have a sufficiently large dataset in order
to have a separate validation set for selecting the best
configuration. This validation set must be different
from the test set, which should only be used for the fi-
nal performance assessment. Evaluating different pa-
rameters on the test set, which is then used for report-
ing the final classification accuracy inevitably causes
leakage, as such test set cannot be considered "new"
or "unseen" by the algorithm since it was used for
making modeling decisions. Alternative to the sin-
gle training / validation / test split would be a proce-
dure called nested cross-validation (Cawley and Tal-
bot, 2010), often applied in tasks involving small data,
which can be investigated in this work.
Future works will also investigate data of
ECOTECH recorded on other muscles involved in
gait movement.
ACKNOWLEDGMENTS
The present paper used collected data from the French
national project ECOTECH supported by the French
National Agency for research under the contract No.
ANR-12-TECS-0020.
REFERENCES
Adam, T., Salam, M., and Gunawan, T. S. (2013). Wavelet
based cepstral coefficients for neural network speech
recognition. In 2013 IEEE International Conference
on Signal and Image Processing Applications, pages
447–451. IEEE.
Andrade, A. O. and Soares, A. (2001). EMG pattern recog-
nition for prosthesis control.
Barbero, M., Merletti, R., and Rainoldi, A. (2012). Atlas
of muscle innervation zones: understanding surface
electromyography and its applications. Springer Sci-
ence & Business Media.
Barkhaus, P. E. (2001). Motor unit action potential quantita-
tion. In An AAEM workshop. Rochester, Minn: Amer-
ican Association of Electrodiagnostic Medicine, vol-
ume 2.
Bengacemi, H., Abed-Meraim, K., Buttelli, O., Ouldali, A.,
and Mesloub, A. (2020). A new detection method for
EMG activity monitoring. Medical & Biological En-
gineering & Computing, 58(2):319–334.
Bhoi, A. K. (2017). Classification and clustering of parkin-
son’s and healthy control gait dynamics using LDA
and K-means. International Journal Bioautomation,
21(1).
Bonato, P., Alessio, T. D., and Knaflitz, M. (1998). A sta-
tistical method for the measurement of muscle activa-
tion intervals from surface myoelectric signal during
gait. Transactions on Biomedical Engineering, IEEE,
45(3):287–299.
Buttelli, O. (2012). Agence nationale de la recherche.
http://www.agence-nationale-recherche.fr/Projet-
ANR-12-TECS-0020.
Carletti, T., Fanelli, D., and Guarino, A. (2006). A new
route to non invasive diagnosis in neurodegenerative
diseases? Neuroscience letters, 394(3):252–255.
Cawley, G. and Talbot, N. (2010). On over-fitting in
model selection and subsequent selection bias in per-
formance evaluation. Journal of Machine Learning
Research, 11:2079–2107.
Elamvazuthi, I., Duy, N., Ali, Z., Su, S., Khan, M. A., and
Parasuraman, S. (2015). Electromyography (EMG)
based classification of neuromuscular disorders using
multi-layer perceptron. Procedia Computer Science,
76:223–228.
Englehart, K., Hudgins, B., Parker, P. A., and Stevenson,
M. (1999). Classification of the myoelectric signal
using time-frequency based representations. Medical
Engineering and Physics, 21(6):431–438.
Furui, S. (1981). Cepstral analysis technique for automatic
speaker verification. IEEE Transactions on Acoustics,
Speech, and Signal Processing, 29(2):254–272.
Furui, S. (1986). Speaker-independent isolated word recog-
nition using dynamic features of speech spectrum.
IEEE Transactions on Acoustics, Speech, and Signal
Processing, 34(1):52–59.
Ghassemi, N. H., Marxreiter, F., Pasluosta, C. F., Kugler,
P., Schlachetzki, J., Schramm, A., Eskofier, B. M.,
and Klucken, J. (2016). Combined accelerometer and
EMG analysis to differentiate essential tremor from
parkinson’s disease. In Engineering in Medicine and
Biology Society (EMBC), 2016 IEEE 38th Annual In-
ternational Conference of the, pages 672–675. IEEE.
Hacine-Gharbi, A. and Ravier, P. (2018). Wavelet cepstral
coefficients for electrical appliances identification us-
ing hidden markov models. In ICPRAM, pages 541–
549.
Hausdorff, J. M., Cudkowicz, M. E., Firtion, R., Wei, J. Y.,
and Goldberger, A. L. (1998). Gait variability and
basal ganglia disorders: Stride-to-stride variations of
gait cycle timing in parkinson’s disease and hunting-
ton’s disease. Movement disorders, 13(3):428–437.
Hausdorff, J. M., Mitchell, S. L., Firtion, R., Peng, C.-
K., Cudkowicz, M. E., Wei, J. Y., and Goldberger,
A. L. (1997). Altered fractal dynamics of gait: re-
duced stride-interval correlations with aging and hunt-
ington’s disease. Journal of applied physiology,
82(1):262–269.
Henmi, O., Shiba, Y., Saito, T., Tsuruta, H., Takeuchi, A.,
Shirataka, M., Obuchi, S., Kojima, M., and Ikeda, N.
(2009). Spectral analysis of gait variability of stride
interval time series: comparison of young, elderly
and parkinson’s disease patients. Journal of Physical
Therapy Science, 21(2):105–111.
Surface EMG Signal Classification for Parkinson’s Disease using WCC Descriptor and ANN Classifier
293