Frequency analysis of biophysiological signs of people with tremor

International Journal of Mechanical Engineering and Technology (IJMET)
Volume 10, Issue 12, December 2019, pp. 303-310, Article ID: IJMET_10_12_033
Available online at />ISSN Print: 0976-6340 and ISSN Online: 0976-6359
© IAEME Publication

FREQUENCY ANALYSIS OF
BIOPHYSIOLOGICAL SIGNS OF PEOPLE WITH
TREMOR
Angie J Valencia C
Faculty of Engineering
Militar Nueva Granada University, Bogotá D.C., Colombia
Mauricio Mauledoux
Faculty of Engineering
Militar Nueva Granada University, Bogotá D.C., Colombia
Edilberto Mejia-Ruda
Faculty of Engineering
Militar Nueva Granada University, Bogotá D.C., Colombia
Ruben D. Hernández
Faculty of Engineering
Militar Nueva Granada University, Bogotá D.C., Colombia
Oscar F. Avilés
Faculty of Engineering
Militar Nueva Granada University, Bogotá D.C., Colombia
ABSTRACT
Tremor is defined as an oscillatory, rhythmic and involuntary movement of one or
more parts of the body. People who suffer from this type of disorder usually have
difficulty performing daily tasks, such as: working, bathing, dressing and eating;
which significantly decreases the quality of life, being shameful and even disabling [1,
2, 3]. The manifestations of essential tremor usually worsen with age. In addition,
there is some evidence that people with essential tremor are more likely than average
to develop other neurodegenerative diseases such as Parkinson's or Alzheimer's

diseases, especially if the tremor first appears after 65 years old [4, 5, 6, 7]. With the
above, this work is carried out in frequency analysis of electromyography signals
acquired through a MYO in patients who emulate tremor to identify the tremor of the
voluntary components of a recorded signal.
Keywords: Quality of Life, Neurodegenerative Diseases, MYO, Oscillating
Movements, Tremor, Fourier Transform

/>
303




Angie J Valencia C, Mauricio Mauledoux, Edilberto Mejia-Ruda, Ruben D. Hernández, Oscar F. Avilés

Cite this Article: Angie J Valencia C, Mauricio Mauledoux, Edilberto Mejia-Ruda,
Ruben D. Hernández, Oscar F. Avilés, Frequency Analysis of Biophysiological Signs
of People with Tremor. International Journal of Mechanical Engineering and
Technology 10(12), 2019, pp. 303-310.
/>
1. INTRODUCTION
Tremor is defined by a series of muscular contractions that produce agitated movements in
different parts of the body, most often affecting the hands, followed by the arms, head, vocal
cords, torso and legs. This can be constant or intermittent, sporadic or the result of another
usually neurological disorder. The frequency of the tremor is the "speed" of the shake and
may decrease as the person ages, while the intensity of the tremor increases [8].
This muscular disorder is usually caused by a problem in the deep parts of the brain that
are responsible for muscle movement. Other causes include the use of certain medications,
alcoholism or alcohol withdrawal after a period of excessive consumption, mercury
poisoning, overactive thyroid, liver or kidney failure, and anxiety or panic [9]. The tremor is

considered not to be life-threatening, but it can be embarrassing and disabling, which makes it
difficult or impossible to work or perform tasks in everyday life [3, 10].
Muscle disorders can be classified according to their appearance, their cause or origin, so
there are about 20 types of tremor. Among the most common include: Essential tremor, which
is one of the most common disorders, being mild and remaining stable for many years. The
tremor usually appears on both sides of the body, but often becomes stronger in the dominant
hand because it is an action tremor [11, 12]. Additionally, there is dystonic tremor, which
occurs in people affected by movement disorders where incorrect brain messages cause
muscle hyperactivity, which result in abnormal postures due to strong muscle spasms or
cramps, reducing their intensity by touching the part of the affected body or muscle [11].
There is also the so-called cerebellar tremor, which is defined as being a slow and wide
tremor of the extremities, which occurs at the end of an intentional movement, such as trying
to press a button. As the name implies, it is caused by an injury to the cerebellum and the
pathways from this to other brain regions [11]. There is also the physiological tremor that
occurs in all healthy people, which is not considered a disease, but a normal human
phenomenon that results from the physical properties of the body [12]. And finally, there is
parkinsonian tremor and orthostatic tremor. The first presents itself as a common symptom of
Parkinson's disease, and manifests itself by shaking one or both hands when they are at rest.
While orthostatic tremor is a rare disorder that is characterized by rapid muscle contractions
in the legs when standing. It is usually accompanied by a feeling of instability or imbalance,
which makes people immediately try to sit or walk [12, 13].
For the proper management of the techniques that must be implemented in the
development of devices that contribute to the rehabilitation of patients with pathological
tremor, the tremor and voluntary components of a registered signal must be distinguished in
the first instance, either for diagnosis or treatment , with which strategies have been used
ranging from linear filtering to stochastic estimators. In [14] an optimal digital filter was
designed offline through follow-up tasks. In [15] they used a second-order low pass filter
applied to an electromyography tremor signal to be transmitted to a neural network, intended
to control an elbow device. In [16] they used a high-pass filter to separate the tremor
component before passing it to a repetitive control loop using a functional electrical

stimulation system. Another recent work used a tremor estimator in the form of a high-pass
filter, which resulted in a significant phase change, which was corrected before being applied
to the suppressor actuator [17].

/>
304




Frequency Analysis of Biophysiological Signs of People with Tremor

Another estimation method is the linear weighted Fourier combiner that adaptively shakes
a tremor signal by tracking its frequency, amplitude and phase [18, 19]. However, to obtain
the best performance, it is recommended to perform a pre-filtering stage with a high pass filter
[20, 21]. A different approach was presented with an adaptive bandpass filter, proposed by
[22] and compared favorably with the weighted frequency Fourier linear combiner.
The Kalman filter is a stochastic estimator, based on a Bayesian model that has been used
for the suppression of tremor by several researchers. In [23] and [24] they implemented a
Kalman filter to track the voluntary movement and by subtracting it from the total movement,
obtain an estimate of the tremor, used to control an upper arm orthosis of three degrees of
freedom. Additionally, GH and Benedict-Bordner filters were used, differentiated by the
method of weight selection [25]. In [26] they have also implemented the Kalman filter,
merging the information from the accelerometer and electromyography data, to obtain a single
estimate of the tremor angle that will be used in the diagnosis, classification and applications
of functional electrical stimulation.
The present work is structured as follows: Section 1, describes the stage of calibration of
the system with basic movements performed by the user. In section 2, the frequency analysis
of results and simulations will be performed. Finally, in section 3 conclude on the data
obtained by applying Fourier transform.

2. ACQUISITION OF ELECTROMIOGRAPHIC SIGNS
2.1. System Calibration
For the analysis of electromyography data in patients with tremor, the development of an
interface that provides the possibility to acquire and save in tables the values obtained during
the execution of the tests is carried out. From there it begins with initial calibration stages
from which the values in orientation, rotation and acceleration are obtained for a forearm
flexion movement, an extension movement, lateral forearm rotation, average forearm rotation,
supination movement and pronation movement.

2.2. Frequency Spectrum of People with Tremor
Biophysiological signals are acquired through MYO, from people with and without tremor for
movement patterns given by drinking a drink. Having as constants an empty glass that allows
to have the same weight in all the samples obtained. After the collection of information, an
analysis is carried out in the frequency spectrum through the Fourier Transform (FT). This in
order to observe the behaviors in module and phase that are characteristic of people with
tremor.

3. RESULTS AND DISCUSSION
Of the data that should be considered are those acquired by the gyroscope and accelerometer
at the coordinates X, Y, Z. In Figure 1 and 2, the phase behavior of a healthy and sick person
is observed for an accelerometer movement in the coordinate Z respectively.

/>
305




Angie J Valencia C, Mauricio Mauledoux, Edilberto Mejia-Ruda, Ruben D. Hernández, Oscar F. Avilés

Figure 1. Z accelerometer phase – Sick.

Figure 2. Z accelerometer phase – Healthy.

From there, differences are observed in the phase distribution of a sick and healthy person.
For example, in Figure 2, the largest number of samples are concentrated in the range of -0.3
to 0.3, which if analyzed in the real (abscissa) and imaginary (ordered) planes, mean
behaviors without oscillations.
In the phase diagram, the dispersion in the histogram samples is directly related to the
oscillations of the signal, so the behavior in Figure 1 corresponds to an increase in the
oscillations generated by involuntary muscle movements with respect to the signs of a healthy
person (Figure 2).
On the other hand, the modulus behavior (real part of the FT) of the Z accelerometer
signal is observed and compared for sick (Figure 3) and healthy person (Figure 4).

/>
306




Frequency Analysis of Biophysiological Signs of People with Tremor

Figure 3. Z accelerometer module – Sick.

Figure 4. Z accelerometer module – Healthy.

From there, symmetric distributions are observed in both signals that are reflected in
similar response times, which means that both movements took approximately 2 seconds to

execute. So, the phase behaviors for the X and Y coordinates will be analyzed, which are
those in which the vibrations have their greatest effect. Starting with the signals in X
coordinates, shown in Figures 5 and 6, and for Y coordinates in Figures 7 and 8.

Figure 5. Accelerometer phase in X – Sick.

/>
307




Angie J Valencia C, Mauricio Mauledoux, Edilberto Mejia-Ruda, Ruben D. Hernández, Oscar F. Avilés

Figure 6. Accelerometer phase in X - Healthy.

Figure 7. Accelerometer phase in Y - Sick.

Figure 8. Accelerometer phase in Y - Healthy.

Similar behaviors are observed to those obtained in Z, which verify the existence of
vibrations (tremor) by the distribution in the phase histogram of the MYO accelerometer
signal.

/>
308

Frequency Analysis of Biophysiological Signs of People with Tremor

4. CONCLUSIONS
With the analysis of the discrete Fourier transform, the frequency spectra in magnitude and
phase of the signals of people with and without tremor are observed. This in order to make a
subsequent filter that allows to generate mechanical solutions that mitigate the involuntary
movements of people with muscular disorders and thus improve their quality of life.
From the acquired signals, it is observed that the accelerometer in the Z coordinate is the
one that is most affected by these oscillatory movements, which is part of this for the design
of filters and solutions in future works.

ACKNOWLEDGEMENT
The research for paper was supported by Military University Nueva Granada by research
project ING 2658.

REFERENCES
[1]

Gironell, A., Kulisevsky, J., Pascual-Sedano, B., Otermín, P., Barbanoj, M., & Gich, I.
(2001). Temblor postural: estudio clínico y neurofisiológico en una serie consecutiva de
300 pacientes. Medicina clínica, 117(16), 601-606.

[2]

Shahed, Joohi, and Joseph Jankovic (2007). "Exploring the relationship between essential
tremor and Parkinson's disease." Parkinsonism & related disorders 13, (2), 67-76.

[3]

Elble, R. J. (2009). Tremor: clinical features, pathophysiology, and treatment. Neurologic

clinics, 27(3), 679-695.

[4]

Högl, Birgit, S. Kiechl, J. Willeit, M. Saletu, B. Frauscher, K. Seppi, J. Müller et al
(2005). "Restless legs syndrome A community-based study of prevalence, severity, and
risk factors." Neurology 64(11), 1920-1924.

[5]

Shahed, Joohi, and Joseph Jankovic (2007). "Exploring the relationship between essential
tremor and Parkinson's disease." Parkinsonism & related disorders 13, (2), 67-76.

[6]

Tse, Winona, Leslie S. Libow, Richard Neufeld, Gerson Lesser, Judith Frank, Susan
Dolan, Chaim Tarshish et al (2008). "Prevalence of movement disorders in an elderly
nursing home population." Archives of gerontology and geriatrics 46, (3), 359-366.

[7]

Lorga Filho, A. M., A. D. Azmus, A. M. Soeiro, A. S. Quadros, A. Avezum Junior, A. C.
Marques, A. Franci et al (2013). "Diretrizes brasileiras de antiagregantes plaquetários e
anticoagulantes em cardiologia." Arquivos Brasileiros de Cardiologia 101, (3), 1-95.

[8]

Elble, Rodger, and Günther Deuschl (2011). "Milestones in tremor research." Movement
Disorders 26(6), 1096-1105.

[9]

Lorga Filho, A. M., A. D. Azmus, A. M. Soeiro, A. S. Quadros, A. Avezum Junior, A. C.
Marques, A. Franci et al (2013). "Diretrizes brasileiras de antiagregantes plaquetários e
anticoagulantes em cardiologia." Arquivos Brasileiros de Cardiologia 101, (3), 1-95.

[10]

Rocon, E., Belda-Lois, J. M., Sanchez-Lacuesta, J. J., & Pons, J. L. (2004). Pathological
tremor management: Modelling, compensatory technology and evaluation. Technology
and Disability, 16(1), 3-18.

[11]

Rodríguez, M. I., Gajardo Cynthia,; Dystonia and Rehabilitation in Children, World´s
Largest Science, Technology & Medicine Open Access book publisher, Instituto de
Rehabilitación Infantil Teletón Santiago de Chile.

[12]

Gironell, A., Kulisevsky, J., Pascual-Sedano, B., Otermín, P., Barbanoj, M., & Gich, I.
(2001). Temblor postural: estudio clínico y neurofisiológico en una serie consecutiva de
300 pacientes. Medicina clínica, 117(16), 601-606.

/>
309

Angie J Valencia C, Mauricio Mauledoux, Edilberto Mejia-Ruda, Ruben D. Hernández, Oscar F. Avilés

[13]

Yagüe, S., Veciana, M., Pedro, J., & Campdelacreu, J. (2011). Importancia del estudio
electromiográfico en el diagnóstico del temblor ortostático. Neurología, 26(1), 53-55.

[14]

Gonzalez, J. G., Heredia, E. A., Rahman, T., Barner, K. E., & Arce, G. R. (2000). Optimal
digital filtering for tremor suppression. IEEE Transactions on Biomedical Engineering,
47(5), 664-673.

[15]

Ando, T., Watanabe, M., Nishimoto, K., Matsumoto, Y., Seki, M., & Fujie, M. G. (2012).
Myoelectric-controlled exoskeletal elbow robot to suppress essential tremor: extraction of
elbow flexion movement using STFTs and TDNN. Journal of Robotics and Mechatronics,
24(1), 141-149.

[16]

Verstappen, R. J., Freeman, C. T., Rogers, E., Sampson, T., & Burridge, J. H. (2012,
October). Robust higher order repetitive control applied to human tremor suppression. In
Intelligent Control (ISIC), 2012 IEEE International Symposium on (pp. 1214-1219).
IEEE.

[17]

Taheri, B., Case, D., & Richer, E. (2014). Robust controller for tremor suppression at

musculoskeletal level in human wrist. IEEE Transactions on neural systems and
rehabilitation engineering, 22(2), 379-388.

[18]

Riviere, C. N., & Thakor, N. V. (1996). Modeling and canceling tremor in humanmachine interfaces. IEEE engineering in medicine and biology magazine, 15(3), 29-36.

[19]

Riviere, C. N., Rader, R. S., & Thakor, N. V. (1998). Adaptive cancelling of physiological
tremor for improved precision in microsurgery. IEEE Transactions on Biomedical
Engineering, 45(7), 839-846.

[20]

Veluvolu, K. C., Latt, W. T., & Ang, W. T. (2010). Double adaptive bandlimited multiple
Fourier linear combiner for real-time estimation/filtering of physiological tremor.
Biomedical Signal Processing and Control, 5(1), 37-44.

[21]

Wang, S., Gao, Y., Zhao, J., & Cai, H. (2014). Adaptive sliding bandlimited multiple
fourier linear combiner for estimation of pathological tremor. Biomedical Signal
Processing and Control, 10, 260-274.

[22]

Popović, L. Z., Šekara, T. B., & Popović, M. B. (2010). Adaptive band-pass filter (ABPF)
for tremor extraction from inertial sensor data. Computer methods and programs in
biomedicine, 99(3), 298-305.

[23]

Rocon, E., Belda-Lois, J. M., Ruiz, A. F., Manto, M., Moreno, J. C., & Pons, J. L. (2007).
Design and validation of a rehabilitation robotic exoskeleton for tremor assessment and
suppression. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 15(3),
367-378.

[24]

Rocon, E., & Pons, J. L. (2011). Exoskeletons in rehabilitation robotics: Tremor
suppression (Vol. 69). Springer.

[25]

Brookner, E. (1998). Frontmatter and Index (pp. i-xxiii). John Wiley & Sons, Inc.

[26]

Widjaja, F., Shee, C. Y., Latt, W. T., Au, W. L., Poignet, P., & Ang, W. T. (2008, May).
Kalman filtering of accelerometer and electromyography (EMG) data in pathological
tremor sensing system. In Robotics and Automation, 2008. ICRA 2008. IEEE
International Conference on (pp. 3250-3255). IEEE.

/>
310