| Abstract|| |
Context: Yoga has origins speculated to date back to pre-Vedic Indian period and is practiced as a common exercise, both in India as well as all around the world. Although the yoga practices are ages old, there is not much research literature available. Moreover, with the advancement in technology, the modern analysis tools are not used up to their full potential. Aims: This research focuses on developing a framework for analyzing trikonasana, using the optical motion capture system, and validating the noninvasive method for analyzing muscle activity in prominent muscles while performing trikonasana. Subjects and Methods: We have adopted the noninvasive analysis method using optical motion capture system OptiTrack™ for recording the human motion and musculoskeletal modeling software LifeMod™ to analyze the muscle activity while performing trikonasana. Surface electromyography (sEMG) studies were performed using Trigno™ (Delsys Inc.) wireless sEMG sensors to validate the LifeMod simulation results pertaining muscle activation. Results: It was observed that the characteristics of the sEMG match to that of the estimated muscle tension from the architecture used in this study. The muscle groups such as external right obliques muscles, rectus abdominis of the front leg, and gluteus maximus and gluteus medius of the rear leg were observed to undergo major activation during an isometric contraction while performing trikonasana. The magnitudes of the muscle tension during the left bend depict a close resemblance to the muscle tension magnitudes during the right bend. Conclusions: The optical motion capture system and musculoskeletal modeling software can be used to analyze muscle activity in any yoga exercise noninvasively. Since the yoga exercises majorly require the practitioner to maintain a certain posture for a considerable duration, our approach can be used to find the important muscles involved and their corresponding muscle tension when they undergo isometric contraction.
Keywords: Isometric contraction, LifeMod, musculoskeletal modeling, optical motion capture, trikonasana, yoga
|How to cite this article:|
Kumar A, Kapse RC, Paul N, Vanjare AM, Omkar S N. Musculoskeletal modeling and analysis of trikonasana. Int J Yoga 2018;11:201-7
| Introduction|| |
The word yoga is derived from the Sanskrit word yuj meaning union. Yoga utilizes mental and physical exercises to attain samadhi or the union of the individual self with the infinite. The regular practice of yoga helps establish natural harmony between various organs of the body. This, in turn, helps maintain the well-being of the individual. The origin of yoga though a highly debated topic could be traced back since the evolution of the Indus Valley civilization, as a typical Indian cultural phenomenon, firmly rooted in the society., Since then yoga has undergone several variations in the style of practice pertaining to there are four major periods or generations of various styles of yoga practised; Vedic, Pre-Classical, Classical and Post-Classical.,
Yoga is a broad concept involving various physical, mental, and spiritual practices. The most popular among all yoga asanas is the Hatha Yoga. Some historians claim the emergence of Hatha yoga practice predates even before the 11th century and sheds its persistence over following a proper exercise routine while maintaining specific body postures (asanas). During the late 20th century, the practice of Hatha yoga was highly popularized in the Western world by Iyengar.
Trikonasana, an asana in Hatha Yoga, largely focuses on the lateral bend-stretching, resulting from the wider spacing of the feet. Performing trikonasana, while maintaining the perfect posture largely influences the lateral spinal mobility, elongates the leg muscles, and removes stiffness in the legs and hips., Performing any yogasana (especially trikonasana) is governed by step-wise body posture maintenance and stability which are the significant factors.
The yoga asanas are usually taught by the practioners who learned from their predecessors, who seldomly have tried to step outside of the tradition and look at yoga from a research point of view. The start of 21st century saw more of elaborate scientific research focused toward the positive aspects of performing yoga for various diseases and body pain and results delivered show yoga as a relatively safer form of exercise, capable of increasing strength, flexibility, balance, and functional endurance of people in good health conditions and also those with musculoskeletal affliction.,, Ni et al. analyzed 11 yoga postures from Vinyasa yoga for trunk and hip muscles. Longpré et al. primarily compared the muscle activation of the thigh muscles in various standing poses, one of which being trikonasana. In one study, the authors have used electromyography (EMG) to analyze and compare abdominal muscles activity while doing breathing yoga exercise versus traditional curls. Salem et al. proposed one of the first biomechanical analysis methods while performing Hatha yoga using motion capture system integrated with surface EMG (sEMG) to quantify the physical demands of performing the asanas and evaluating joint angles, muscle activation levels, etc., for various Hatha yoga postures.
Although the sEMG sensors have been used to study the muscle activation patterns, it has a couple of demerits to its account. The sEMG can give the insights of the superficial muscles only. The human body is prone to the phenomenon of perspiration under high temperature or humid conditions, which can lead to sliding of the sEMG sensors and eventually falling-off the skin. Excessive fat deposition in the body can hinder the results of the sEMG sensors. Moreover, they are prone to noise and often mixed up with other muscle signals. The solution for this is to use inverse dynamic algorithms developed in robotics to obtain the joint torques of the skeleton and then run numerical optimization to compute the muscle tensions. Here, the joint torques are computed from the motion data and the geometric and inertial parameters of the skeleton model, and the mathematical optimization with physiologically appropriate criteria, for example, minimizing the signal-dependent noise, distributes them to the muscle tensions. The motion data can be captured by various methods one of which is optical motion capture. The notion of motion capture system was primarily intended for research applications in the field of computer vision, computer-generated imagery, and motion planning in robotics. However, recent developments in the field of biomechanical research have witnessed marker-based optical motion capture system as a noninvasive and highly valuable tool for recording and analyzing of the human motion, particularly pertaining to sports and rehabilitation studies., One of the reasons for using the optical motion capture system in biomechanical studies is the ease of interfacing the acquired human motion data with different musculoskeletal modeling software (such as OpenSim, Adams LifeMod, etc.,) for analyzing human gait patterns, kinematics, dynamics, and determination of various associated parameters such as muscle activation and joint torques and angles. In the field of sports-rehabilitation sciences, optical motion capture has been extensively used as a medium of recording and analyzing the motion patterns of the subjects. Although previously there have been experimental studies conducted on various yogasana based on sEMG which involves integrating inertial measurement unit, accelerometer, or gyroscope-based sensors, very few have taken the approach of the utilization of optical motion capture system as a means of data acquisition. Moreover, since body posture is very critical in yoga, optical motion capture system can be very useful to capture full body to check proper alignment.
Hill type muscle model
The hill-type muscle consists of three elements and is used to depict the mechanical muscle response.,, The major components of this model include the contractile element (CE), in series with the series elastic element, and parallel with the parallel elastic (PE) element depicted in [Figure 1]. The CE is responsible for generating the active whereas the passive element determines the passive elastic properties of the muscle fiber.
The hill equation formulates that Fmuscle generated is the aggregate of passive element force (FPE) and the CE force (FCE) while ignoring the SEE shown in equation (1). The CE (also known as an active element) is dependent on instantaneous muscle length (list), instantaneous muscle shortening velocity, and activation state (A [t]) which is a function of time in equation (2). The passive element is the function of list.
FMuscle = FCE + FPE(1)
The contractile elemental force (FCE) is computed based on equation (2), which models the active muscle behavior with normalized activation state and a maximum Fmuscle at activation. It is essentially dependent on the force-length (fV) and force-velocity relation (fL).,
FCE = A (t). Fmax. fV. (vl). fL. (ll)(2)
Where, A (t) = Activation state which is normalized between the resting state, 0 to the maximum activation state, 1.
Fmax is Fmuscle at maximum activation isometric conditions.
fV is the normalized active force-velocity relation (based on hill-curve).
fL is the normalized active force-length relation.
vl is the dimensionless lengthening velocity.
ll is dimensionless muscle length.
The FPE is modeled based on the product of passive muscle stress (σ) and physiological cross-sectional area (pCSA) of the specific muscle equation (3). The nonlinear stress-strain relationship is used to model the passive muscle stress.
FPE = σ. pCSA(3)
Following literature survey, analysis of muscle activity while performing trikonasana using noninvasive techniques was considered necessary. Trikonasana has been considered specifically for the current study because of its simple geometry through which all the aspects of the study can be illustrated and detectability of all the markers by the motion capture cameras while performing the exercise and the isometric nature of performing the asana.
Hence, the current study was undertaken to create an architecture for biomechanical analysis of the subject performing trikonasana using optical motion capture system OptiTrack™ and musculoskeletal modeling and simulation using LifeMod. In addition, information regarding key muscle groups recruited while performing the asana and hence elucidating the kinds of muscle groups strengthened was explored.
The clinical relevance of this study is to provide an initial basis for better prescription of Trikonasana as a physical therapy to better recruit trunk and thigh muscles and also as a biofeedback tool to assist training and rehabilitation.
| Subjects and Methods|| |
The study was performed at the Biomechanics Laboratory, Indian Institute of Science (IISc), Bangalore. One healthy male subject of age 56 years, with a height of 165 cm and weight 64 kg with prior experience and practice for 30 years in the field of yoga was selected. The subject gave a written consent of his willingness to participate in this experiment, verified by the Human Ethics Committee, IISc.
The exercise was performed in the following protocols in the gait cycle of the trikonasana as depicted in [Figure 2].
- Initially, the subject stood in an upright position (tadasana) maintaining around one-foot distance between the feet
- Then, subject was instructed to raise both hands to shoulder level with palm facing the ground and stretch both legs to an extent of 3–4 feet
- Subsequently, turning the right foot outward about 90°, while keeping the left foot rotated inward about 15° ensuring the direction of the right foot is to the center of the left foot
- The subject bends the entire upper torso, laterally at 90° toward the right side, till the right-hand rests on the right shin bone
- The left hand is held straight up
- The head is turned gently upward and the gaze is set toward the tip of the left palm while ensuring that the legs, arms, and the torso are in the same plane. Both arms are maintained along the same straight line
- The current posture is maintained for about 20 s, with normal breathing
- After a time span of 20 s, the subject returns to the initial position while retaining the stretch, inhaling slowly and come up
- Later, the same procedure is repeated, changing the side (i.e. left side).
Motion capture data acquisition using OptiTrack motion capture system
OptiTrack™ motion capture system was used for biomechanical analysis. A total of eight infrared-optical motion capture cameras (OptiTrack Prime 13W, each sampling at 120 Hz) were used for motion tracking and recording. The software platform used was OptiTrack Motive. Optical motion capture system although effective in tracking human motion comes with a certain range of diminution. First, the optical cameras were calibrated, placing retro-reflective markers on the dark suit worn by the subject. As per marker placement protocols in OptiTrack Motive software, the standard Plug-In-Gait (39 markers) is adopted for the experiment, defining each body segments, head, torso, upper arms (left and right), upper hands, forearms, both upper and lower legs, and feet., The 39-optical reflective markers, 12 on lower body, 23 along the upper body, and 4 on the head, are placed on the subject as shown in [Figure 3]. Based on the information from all the optical motion tracking cameras, the marker locations were determined and the subsequent three-dimensional (3D) motion trajectories are confirmed more were accurate.
Musculoskeletal modeling of the human body in LifeMOD
The optical motion capture data recorded through the OptiTrack Motive software was exported in the comma-separated values format file. The file mostly contains the marker position in mm and also the motion trajectory data. There are total of 117 position values, three (x, y, and z coordinates) for each marker position. These values were copied to the text files in the time-stamped format. A script compiled in python was used to accomplish this. The script also offsetted each value by the distance of the hip's COM location to translate the actual origin to the LifeMOD's origin which is at the center of the pelvis.
The 3D musculoskeletal model of the human body was generated through LifeMOD Biomechanics modeler, based on the subject's anthropometric data. The extent of the application of the LifeMOD/ADAMS in medical and sports research is commendable while analyzing the human body motion through generated models, simulations, and delivering a substantial amount of outputs., Even in the field of yogasana, LifeMOD has been tremendously useful as a simulation tool for analyzing the human body dynamics.
The musculoskeletal model of the subject in LifeMOD environment follows the closed hill type muscle model [Figure 4], with a group of main muscles that govern the rigorous motions of the body segments. After setting up the human model with LifeMOD, motion trajectory was imported and an inverse dynamic simulation was carried out by ADAMS. Before running the inverse dynamic simulation, the muscle tissue properties such as maximum allowable stress to each specific muscle and physiological cross-section area (pCSA) are assigned, from the GeBod anthropometric database. The spring-damper complexes' (muscles) contraction histories were recorded in the inverse dynamic simulation. Then, the motion driver was removed and the muscle contraction histories were employed to perform a forward dynamic simulation. During forward dynamic simulation, the trikonasana was performed by computer-generated human model to recreate the motion history and show the relationship between muscles force and body motion.
An important factor of this research was to determine the muscle tension during the extreme bending positions while performing trikonasana when isometric contraction takes place. From the previous studies, it can be inferred that there is a linear relationship between the muscle tension and sEMG. Therefore, the surface EMG techniques can be used in the current study to validate the results acquired from the LifeMod simulations. For the data acquisition, we used the Trigno™ wireless sEMG sensor, courtesy of Delsys Inc. The sensor was communicating wirelessly with a personal computer running the software, where the data were streamed and recorded.
Before attaching the sensors, the subject's skin was cleaned by swabbing it with 70% isopropyl alcohol. Double-sided adhesive interfaces tailored to match the contours of the sensors were applied on the sEMG sensors to allow it to stick to the skin. One wireless sEMG sensor was placed over the right external oblique muscle, located approximately 15 cm lateral and transverse to the umbilicus. The sEMG sensor was kept on, during the full cycle of performing trikonasana. The sEMG sensor was switched-on before recording any data. sEMG data were simultaneously recorded while the subject was performing the exercise in the motion capture suit to maintain the sync between the two systems.
| Results|| |
The [Figure 5] and [Figure 6] show the muscle tension in the prominent muscles involved during right and left holding positions. It can be observed from the figures that the muscle groups such as obliques, rectus abdominis, and psoas major of the front leg and gluteus maximus and gluteus medius of the back leg undergo major activation during isometric contraction. We can see that the magnitudes of the muscle tension during the left bend, closely match the magnitudes during the right bend.
|Figure 5: LifeMod estimation of muscle tensions during right bend isometric contraction|
Click here to view
|Figure 6: LifeMod estimation of muscle tensions during left bend isometric contraction|
Click here to view
To validate the muscle tension estimated by the musculoskeletal model, sEMG was measured for muscle groups such as obliques. Various studies have shown that there is a linear relationship between Fmuscle and EMG at isometric contraction.,
F = g1E + constant(4)
Where g is a constant gain factor and E is the mean rectified EMG.
[Figure 7] is a plot of surface EMG root mean square (RMS) of right oblique during right bend and it can be observed that the characteristics of the EMG match to that of the estimated muscle tension as shown in [Figure 8], from musculoskeletal modeling software LifeMod. Only the sEMG (RMS) data of the right oblique muscle groups during right bend are formulated because those are superficial muscle groups that delivered nearly accurate output compared to the ones from LifeMod.
|Figure 7: Surface electromyography (root mean square) of right oblique muscle at isometric contraction|
Click here to view
|Figure 8: LifeMod estimation of muscle tension in right oblique muscle at isometric contraction|
Click here to view
| Discussion|| |
The primary purpose of our research was to investigate the application of optical motion capture and musculoskeletal modeling as a useful noninvasive tool to measure muscle activity while performing trikonasana. The linear relationship between Fmuscle and sEMG during isometric contraction was positively confirmed by the results as illustrated in [Figure 5] and [Figure 6]. The architecture used in this study is very useful to understand what are the major muscles involved in performing any asana. For example, referring to the [Figure 5] and [Figure 6], it can be inferred that trunk and thigh muscles are majorly effected by maintaining the posture in trikonasana. One of the important benefits of knowing the behavior of muscles is that this allows better prescription of the asana as a physical therapy to better recruit trunk and thigh muscles. The other benefit of using this noninvasive architecture is to analyze not only the behavior of superficial muscles but also internal muscles such as psoas major.
This setup can very well be used to analyze muscle activity in any yoga exercise. Since the yoga exercises majorly require the practitioner to maintain a certain posture for a considerable duration, our approach can be used to find the important muscles involved and their corresponding muscle tension when they undergo isometric contraction. Although the current study primarily focuses on the muscle activity at the final postures, the illustrated setup can be used to capture the full cycle of the asanas, thereby facilitating the analysis of not only the muscle activity but also the nature of geometry involved. This is important to make sure the posture is correct. The execution of the yoga asana has to be very graceful. The dynamics of motion estimated by the musculoskeletal system can help identify any abrupt behavior in performing the asana.
In the future studies, the architecture of implementing optical motion capture could be extended towards various other yoga postures, which will be the main motivation for our future research efforts. Also providing a statistical analysis with a broader experimental sample size will be documented in the future works.
Financial support and sponsorship
This study was financially supported by Science and Technology of Yoga and Meditation, Department of Science and Technology.
Conflicts of interest
There are no conflicts of interest.
| References|| |
Lidell L. The Sivananda Companion to Yoga. New York: Simon & Schuster Inc.; 1983.
Crangle EF. The Origin and Development of Early Indian Contemplative Practices. Germany: Otto Harrassowitz Verlag; 1994.
Flood GD. An Introduction to Hinduism. United Kingdom: Cambridge University Press; 1996.
Murray RJ, Zimmer H, Campbell J. Philosophies of India. Philos Q 1953;3:274.
Crangle EF. The Origin and Development of Early Indian Contemplative Practices. Vol. 29. Germany: Otto Harrassowitz Verlag; 1994.
Mallinson J. Haṭha yoga. In: Bronkhorst J, editor. Brill's Encyclopedia of Hinduism, Handbook of Oriental Studies. Sec. 2. India: Society, Theology, Biography; 2011. p. 770-81.
Kaminoff L. Yoga Anatomy. Vol. 1. United States of America:Human Kinetics; 2007.
Iyengar BK. Light on the Yoga Sutras of Patanjala: Patanjala Yoga Pradipika. Great Britain: Thorsons Publication; 1993.
Greendale GA, Huang MH, Karlamangla AS, Seeger L, Crawford S. Yoga decreases kyphosis in senior women and men with adult-onset hyperkyphosis: Results of a randomized controlled trial. J Am Geriatr Soc 2009;57:1569-79.
Wang MY, Greendale GA, Kazadi L, Salem GJ. Yoga improves upper-extremity function and scapular posturing in persons with hyperkyphosis. J Yoga Phys Ther 2012;2:117.
Raub JA. Psychophysiologic effects of hatha yoga on musculoskeletal and cardiopulmonary function: A literature review. J Altern Complement Med 2002;8:797-812.
Ni M, Mooney K, Harriell K, Balachandran A, Signorile J. Core muscle function during specific yoga poses. Complement Ther Med 2014;22:235-43.
Longpré HS, Brenneman EC, Johnson AL, Maly MR. Identifying yoga-based knee strengthening exercises using the knee adduction moment. Clin Biomech (Bristol, Avon) 2015;30:820-6.
Petrofsky JS, Cuneo M, Dial R, Morris A. Muscle activity during yoga breathing exercise compared to abdominal crunches. J Appl Res Clin Exp Ther 2005;5:501.
Salem GJ, Yu SS, Wang MY, Samarawickrame S, Hashish R, Azen SP, et al.
Physical demand profiles of hatha yoga postures performed by older adults. Evid Based Complement Alternat Med 2013;2013:165763.
Sharma A, Noel MM. Design of a low-cost five-finger anthropomorphic robotic arm with nine degrees of freedom. Robot Comput Integr Manuf 2012;28:551-8.
Kim J, Mastnik S, André E. EMG-based hand gesture recognition for real time biosignal interfacing. In: Proceedings of the 13th
International Conference on Intelligent user Interfaces. Gran Canaria, Spain: ACM; 2008. p. 30-9.
Rasmussen J, Damsgaard M, Voigt M. Muscle recruitment by the min/max criterion – A comparative numerical study. J Biomech 2001;34:409-15.
Harris CM, Wolpert DM. Signal-dependent noise determines motor planning. Nature 1998;394:780-4.
Moeslund TB, Hilton A, Krüger V. A survey of advances in vision-based human motion capture and analysis. Comput Vis Image Underst 2006;104:90-126.
Veloso A, Esteves G, Silva S, Ferreira C, Brandão F. Biomechanics modeling of human musculoskeletal system using Adams multibody dynamics package. In: Proceedings of the 24th
IASTED International Conference on Biomedical Engineering. Innsbruck; 2006. p. 401-7.
Abdallah IB, Bouteraa Y, Rekik C. Design and development of 3D printed myoelectric robotic exoskeleton for hand rehabilitation. Int J Smart Sens Intell Syst 2017;10:341-66.
Liu YS, Tsay TS, Wang TC. Muscles force and joints load simulation of bicycle riding using multibody models. Procedia Eng 2011;13:81-7.
Roy SH, Cheng MS, Chang SS, Moore J, De Luca G, Nawab SH, et al.
A combined sEMG and accelerometer system for monitoring functional activity in stroke. IEEE Trans Neural Syst Rehabil Eng 2009;17:585-94.
Sahai N, Tewari RP, Singh L. Mechanical behavior of muscles during flexion and extension of lower limb on variable age group by using BRG. Life Mod. In: The third International Conference on the Development of Biomedical Engineering in Vietnam. Vol. 27. Berlin, Heidelberg: Springer; 2010. p. 48-50.
Zajac FE. Muscle and tendon: Properties, models, scaling, and application to biomechanics and motor control. Crit Rev Biomed Eng 1989;17:359-411.
van Soest AJ, Bobbert MF. The contribution of muscle properties in the control of explosive movements. Biol Cybern 1993;69:195-204.
Hill AV. The Heat of Shortening and the Dynamic Constants of Muscle. Proc R Soc London B Biol Sci 1938;126:136-95.
Romero F, Alonso FJ. A comparison among different Hill-type contraction dynamics formulations for muscle force estimation. Mech Sci 2016;7:19.
Deng YC, Goldsmith W. Response of a human head/neck/upper-torso replica to dynamic loading – II. Analytical/numerical model. J Biomech 1987;20:487-97.
Kadaba MP, Ramakrishnan HK, Wootten ME, Gainey J, Gorton G, Cochran GV, et al.
Repeatability of kinematic, kinetic, and electromyographic data in normal adult gait. J Orthop Res 1989;7:849-60.
Kadaba MP, Ramakrishnan HK, Wootten ME. Measurement of lower extremity kinematics during level walking. J Orthop Res 1990;8:383-92.
Huynh KT, Gibson I, Lu WF, Jagdish BN. Simulating dynamics of thoracolumbar spine derived from lifeMOD under haptic forces. World Acad Sci Eng Technol 2010;64:278-85.
Kenny IC, Wallace ES, Brown D, Otto SR. Validation of full-body computer simulation of the golf driver for clubs of differing length. In: Moritz EF, Haake S, editors. The Engineering of Sport 6. New York: Springer; 2006.
De Luca CJ. The use of surface electromyography in biomechanics. J Appl Biomech 1997;13:135-63.
Hof A. EMG and muscle force: An introduction. Hum Mov Sci 1984;3:119-53.
S N Omkar
AE128, Department of Aerospace Engineering, Indian Institute of Science, Bengaluru - 560 012, Karnataka
Source of Support: None, Conflict of Interest: None
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6], [Figure 7], [Figure 8]