نوع مقاله : مقاله مروری

نویسندگان

1 دانشجوی دکتری بیومکانیک ورزشی، دانشکده علوم ورزشی، دانشگاه مازندران

2 دانشیار گروه بیومکانیک ورزشی، دانشکده علوم ورزشی، دانشگاه مازندران

3 دانشیار گروه آموزشی ریاضی، دانشکده ریاضی، دانشگاه مازندران

چکیده

تاکنون مطالعه سیستماتیکی به منظور دسته‌بندی روش‌های برآورد غیرمستقیم نیروی عکس‌العمل زمین با استفاده از دیتای کینماتیک حین تکلیف دویدن و بویژه اطلاع‌رسانی درباره‌ی مطالعات انجام شده در این زمینه با استفاده از روش جدید هوش مصنوعی انجام نشده است. بنابراین مطالعه‌ی مروری حاضر، مطالعات انجام شده در زمینه‌ی برآورد نیروی عکس‌العمل زمین با استفاده از دیتای کینماتیک حین دویدن از سال2010 تا 2023 جمع‌‌آوری و بررسی کرده است. این مقالات از طریق جستجو در وب‌سایت‌های Scopus، IEEE Xplore، Medline، ScienceDirect و PubMed با کلید واژه‌های Ground Reaction Force، Kinematics، Kinetics، Camera، Video Analysis، Accelerometer، IMU، in field، Running و ترکیبی از آن‌ها، پیدا شدند. 213 مقاله در منابع مختلف یافت شد که پس از تطبیق با معیارهای ورود به تحقیق، در نهایت 17 مطالعه بررسی گردید. مرور مقالات نشان داد به علت چالش‌هایی که پژوهشگران هنگام به کار بردن ابزارهای اندازه‌گیری مستقیم نیرو با آن مواجه هستند، برآورد غیرمستقیم نیرو از داده‌های کینماتیکی مورد توجه قرار گرفته است. از میان روش‌های برآورد غیرمستقیم، مدل بیومکانیکی و آماری عدم دقت و عدم اطمینان را به همراه دارند. روش‌های یادگیری ماشین به عنوان یک ابزار انعطاف‌پذیر برای مدل‌سازی غیرخطی و بسیار کارآمد، نیازی به دانش قبلی مدل ندارند و مدل خود را براساس الگوریتم‌های شبکه عصبی مصنوعی که یک رابطه قوی بین متغیرهای ورودی و هدف می‌یابد، ایجاد می‌کنند. به همین دلیل امروزه استفاده از روش‌های یادگیری ماشین به عنوان یک رویکرد مدرن که مدل‌سازی و استراتژی‌های جمع‌آوری داده‌ها را ساده می‌کند برای برآورد نیرو در کانون توجه است.

کلیدواژه‌ها

موضوعات

عنوان مقاله [English]

The Estimation of Ground Reaction Force Using Kinematic Data During Running: Can Machine Learning Methods Solve the Limitations?

نویسندگان [English]

  • Fatimah Ahmadi Godini 1
  • Mansour Eslami 2
  • Rohollah Yousef pour 3

1 Ph.D. Candidate, Department of Sports Biomechanics, Faculty of Sports Sciences, University of Mazandaran, Babolsar, Iran

2 Associate Professor, Department of Sports Biomechanics, Faculty of Sports Sciences, University of Mazandaran, Babolsar, Iran

3 Associate Professor, Department of Mathematics, Faculty of Sciences, University of Mazandaran, Babolsar, Iran

چکیده [English]

Background and Purpose
In running analysis and evaluating athletes' performance, ground reaction force (GRF) components are essential parameters in estimating joint reaction forces and joint moments. In a biomechanics lab, Force platforms, instrumented treadmills and wearable sensors are the most common and reliable tools to measure GRF. However, using these tools require specific spaces such as motion analysis laboratories, skilled operators and they may alter foot-ground interaction and shoe stiffness. Furthermore, the mechanical and repeated stress applied to these sensors is high, so that the sensors can be easily worn or damaged. Because of these limitations, researchers have recently focused on indirect model in the estimation of GRF from kinematics data.
In biomechanical models, most of the proposed methods require modeling of the biomechanical system. This modeling, in turn, requires extensive knowledge of the subject's special parameters, such as mass of segments, dimensions, moment of inertia and etc. This inevitably brings inaccuracy and uncertainty. To calculate the mechanical parameters of each segment, extensive use of standard tables is required. Such statistical values may cause inaccuracies in the estimation of the desired values and they may not be generalizable to all societies.
In statistical models, many of the anatomical features of people are not considered. Therefore, this method may not be applicable for activities that involve repeated loads, such as running or during a training session. Furthermore, the accuracy of the estimated GRF by the statistical model is lower than that of other methods.
Recently, the use of machine learning methods, as a modern approach to the GRF estimation, are in the spotlight. These methods are based on the hypothesis that there is a relationship between the acceleration of each segment and the ground reaction forces. Machine learning methods do not require prior knowledge of model and build their own model using the training data obtained in previous experiments. Artificial neural network (ANN) is reported to be a good flexible tool for nonlinear modeling and very efficient for GRF estimation. Indeed, the use of neural networks simplifies modeling and data collection strategies. However, the disadvantages of ANNs include the sensitivity to the chosen input parameters, computationally expensive, as well as requiring a large amount of data for training the system to achieve acceptable accuracy.
According to the review conducted in the current study, most of the reviewed articles reported an acceptable estimation of the vertical component of GRF, while a few focused on its medial-lateral components. However, finding was attributed to the lower absolute values of the medial- lateral force components. Furthermore, while the correlation between body acceleration and vertical GRF was good in most cases, the absolute values of GRF were not correctly estimated. A combination of methods based on biomechanical modeling and machine learning, as well as the use of more sophisticated algorithms, seems to be a promising way to increase the overall accuracy, even in estimating the medial-lateral component of GRF.
Ground reaction force (GRF) components are critical parameters in running analysis, as they are essential for estimating joint reaction forces and joint moments, which are key to evaluating athletes' performance. Traditionally, GRF is measured using force platforms, instrumented treadmills, and wearable sensors, which are considered reliable tools in biomechanics laboratories. However, these methods require specialized spaces, skilled operators, and may alter natural foot-ground interactions and shoe stiffness. Additionally, the mechanical and repeated stress applied to these sensors can lead to wear and damage, further limiting their practicality. Due to these limitations, researchers have increasingly focused on indirect methods for estimating GRF from kinematic data.
Traditional biomechanical models for GRF estimation require detailed modeling of the biomechanical system, which involves extensive knowledge of subject-specific parameters such as segment mass, dimensions, and moments of inertia. This reliance on statistical data from standard tables can introduce inaccuracies and uncertainties, making the results less generalizable across different populations. Statistical models, on the other hand, often overlook many anatomical features, reducing their applicability for activities involving repeated loads, such as running or training sessions. Moreover, the accuracy of GRF estimation using statistical models is generally lower compared to other methods.
Recent advancements in machine learning (ML) methods have provided a promising alternative for GRF estimation. These methods are based on the hypothesis that a relationship exists between segmental accelerations and ground reaction forces. Unlike traditional approaches, machine learning methods do not require prior knowledge of the biomechanical model and can build their own models using training data from previous experiments. Artificial neural networks (ANNs), in particular, have been reported as flexible and efficient tools for nonlinear modeling, making them highly suitable for GRF estimation. However, ANNs are sensitive to the choice of input parameters, computationally expensive, and require large datasets for training to achieve acceptable accuracy.
A review of the current literature indicates that most studies report acceptable estimation accuracy for the vertical component of GRF, while fewer studies focus on the medial-lateral components. This discrepancy is attributed to the lower absolute values of medial-lateral forces, which are more challenging to estimate accurately. Although a strong correlation between body acceleration and vertical GRF has been observed in many cases, the absolute values of GRF are often not estimated correctly. To address these limitations, a combination of biomechanical modeling and machine learning methods, along with the use of more sophisticated algorithms, appears to be a promising approach for improving the overall accuracy of GRF estimation, including the medial-lateral components.
Machine learning methods, particularly ANNs, offer a promising solution to the limitations of traditional GRF estimation techniques. By leveraging the relationship between kinematic data and GRF, these methods can simplify modeling and data collection strategies. However, challenges such as sensitivity to input parameters, computational costs, and the need for large training datasets must be addressed to improve their accuracy and applicability. Future research should focus on integrating biomechanical modeling with machine learning and developing more advanced algorithms to enhance the estimation of both vertical and medial-lateral GRF components.
Article message
Machine learning methods, especially neural networks, are a promising alternative to traditional biomechanical models in estimating ground reaction force during running using kinematic data.



کلیدواژه‌ها [English]

  • Machine Learning
  • Ground Reaction Force
  • Kinematics
  • Estimation
  • Running
  1. Slater A, Campbell A, Smith A, Straker L. Greater lower limb flexion in gymnastic landings is associated with reduced landing force: a repeated measures study. Sport Biomech. 2015;14(1):45–56.
  2. Cavanagh PR, Lafortune MA. Ground reaction forces in distance running. J Biomech. 1980;13(5):397–406.
  3. Donoghue OA, Shimojo H, Takagi H. Impact Forces of Plyometric Exercises Performed on Land and in Water. 2011;3(3):303–9.
  4. Eslami M, Gandomkar A, Hosseini nejad SE, Jahedi V, Gandomkar E. Comparison of the Effect of unstable and control Shoes on the Variables Related to Tibia Stress Fracture during Running in Recreational Runners. J Res Rehabil Sci. 2014;9(6). (Persian)
  5. Pappas E, Sheikhzadeh A, Hagins M, Nordin M. The effect of gender and fatigue on the biomechanics of bilateral landings from a jump: Peak values. J Sport Sci Med. 2007;6(1):77–84.
  6. Hedayat Pour N, Shabani M, Eslami M. The effect of muscle fatigue on ankle joint moment and center of pressure during perturbation of single-leg stance. Razi J Med Sci. 2013; 20 (106) :57-64. (Persian)
  7. Bassiri Z, Eslami M, Ghaemy M, Hosseninejad SE, Rabiei M. The Effect of Shoe Outsole Containing Nanoclay Particles on Knee Joint Power during the Stance Phase of Running. Ann Appl Sport Sci. 2014;2(3):33–40.
  8. Hoseini SZ, Eslami M. Effect of Five-Finger Shoes on Vertical Ground Reaction Force Loading Rates and Perceived Comfort during the Stance Phase of the Running. J Sport Biomech. 2015;37–47. (Persian)
  9. Tazike Lamski Z, Eslami M, Habibi Tirtashi F. The Effect of Shoe Insole Stiffness on Leg Stiffness during Stance Phase of Running in Two Different Speeds among Active Men. J Res Rehabil Sci. 2016;12(1):34–41. (Persian)
  10. Hoseinzadeh E, Eslami M, Taghipur M, Fayyaz-Movaghar A. The Role of Leg Stiffness in Prediction of Medial Tibial Stress Syndrome in Active People: A Prospective Cohort Study. J Res Rehabil Sci. 2017;13(5):286-95. (Persian)
  11. Hatami Joushghan Z, Eslami M. The Effect of Fatigue on the Stiffness Changes in Legs, Ankle and Knee Joints in Lower Limb During a Thirty-Second Continuous Vertical Jump Test in Female Volleyball Players. J Sport Biomech. 2015;1(2):15–23. (Persian)
  12. Eslami M, Begon M, Hinse S, Sadeghi H, Popov P, Allard P. Effect of foot orthoses on magnitude and timing of rearfoot and tibial motions, ground reaction force and knee moment during running. J Sci Med Sport. 2009;12(6):679–84.
  13. Gandomkar A, Eslami M, Hosseini nejad SE, Jahedi V. Effect of unstable shoes on lower extremity join power during stance phase of running. Razi J Med Sci. 2014;21(124):44-43. (Persian)
  14. Boroushak N, Eslami M, Khoshnoodi H. A Review on the Impact Skills Analysis Using two Physical Principles; Impulse- Momentum and Work- Energy in Martial Arts. J Sport Biomech. 2017;3(3):57-67.
  15. Eslami M, Hoseininejad E, Fayyaz A, Sadeghi H. Do minimal shoes imitate barefoot running? Lower limbs mechanical energy using component analysis. J Appl Exerc Physiol. 2017;87–96. (Persian)
  16. Eslami M, Hoseninejad SE, Salari Esker F, Yousefpour R, Fayyaz Movaghar A, Sadeghi H. Determination of functional groups in different levels in running gait; lower limb mechanical energy analysis. In: 35th Conference of the International Society of Biomechanics in Sports, Cologne, Germany. 2017. P. 378–9.
  17. Jiang X, Napier C, Hannigan B, Eng JJ, Menon C. Estimating Vertical Ground Reaction Force during Walking Using a Single Inertial Sensor. Sensors. 2020;20(15):4345.
  18. Ancillao A. Stereophotogrammetry in Functional Evaluation: History and Modern Protocols. In 2018. p. 1–29.
  19. Veltink PH, Liedtke C, Droog E, van der Kooij H. Ambulatory measurement of ground reaction forces. IEEE Trans Neural Syst Rehabil Eng. 2005;13(3):423–7.
  20. Owings TM, Grabiner MD. Measuring step kinematic variability on an instrumented treadmill: how many steps are enough? J Biomech. 2003;36(8):1215–8.
  21. van der Krogt MM, Sloot LH, Harlaar J. Overground versus self-paced treadmill walking in a virtual environment in children with cerebral palsy. Gait Posture. 2014;40(4):587–93.
  22. Davis BL, Perry JE, Neth DC, Waters KC. A Device for Simultaneous Measurement of Pressure and Shear Force Distribution on the Plantar Surface of the Foot. J Appl Biomech. 1998 ;14(1):93–104.
  23. Razian MA, Pepper MG. Design, development, and characteristics of an in-shoe triaxial pressure measurement transducer utilizing a single element of piezoelectric copolymer film. IEEE Trans Neural Syst Rehabil Eng. 2003;11(3):288–93.
  24. Abdul Razak AH, Zayegh A, Begg RK, Wahab Y. Foot Plantar Pressure Measurement System: A Review. Sensors. 2012;12:9884–9912.
  25. Fong DT-P, Chan Y-Y, Hong Y, Yung PS-H, Fung K-Y, Chan K-M. Estimating the complete ground reaction forces with pressure insoles in walking. J Biomech. 2008;41(11):2597–601.
  26. Wundersitz DWT, Netto KJ, Aisbett B, Gastin PB. Validity of an upper-body-mounted accelerometer to measure peak vertical and resultant force during running and change-of-direction tasks. Sport Biomech. 2013;12(4):403–12.
  27. Nigg BM, Liu W. The effect of muscle stiffness and damping on simulated impact force peaks during running. J Biomech. 1999;32(8):849–56.
  28. Young WB, Hepner J, Robbins DW. Movement Demands in Australian Rules Football as Indicators of Muscle Damage. J Strength Cond Res. 2012;26(2):492–6.
  29. Bobbert MF, Schamhardt HC, Nigg BM. Calculation of vertical ground reaction force estimates during running from positional data. J Biomech. 1991 Jan;24(12):1095–105.
  30. Clark KP, Weyand PG. Are running speeds maximized with simple-spring stance mechanics? J Appl Physiol. 2014;117(6):604–15.
  31. Clark KP, Ryan LJ, Weyand PG. A general relationship links gait mechanics and running ground reaction forces. J Exp Biol. 2017; 220(2):247–58.
  32. Raper DP, Witchalls J, Philips EJ, Knight E, Drew MK, Waddington G. Use of a tibial accelerometer to measure ground reaction force in running: A reliability and validity comparison with force plates. J Sci Med Sport. 2018;21(1):84–8.
  33. Oosterwaal M, Telfer S, Tørholm S, Carbes S, van Rhijn LW, Macduff R, et al. Generation of subject-specific, dynamic, multisegment ankle and foot models to improve orthotic design: a feasibility study. BMC Musculoskelet Disord. 2011;12(1):256.
  34. Vanheule V, Andersen MS, Wirix-Speetjens R, Jonkers I, Victor J, Sloten J Vanden. Aalborg Universitet Modeling of patient-specific knee kinematics and ligament behavior using force- dependent kinematics. In: XXIV Congress of the International Society of Biomechanics. 2013.
  35. de Vries WHK, Veeger HEJ, Baten CTM, van der Helm FCT. Magnetic distortion in motion labs, implications for validating inertial magnetic sensors. Gait Posture. 2009;29(4):535–41.
  36. Leardini A, Chiari L, Croce U Della, Cappozzo A. Human movement analysis using stereophotogrammetry. Gait Posture. 2005;21(2):212–25.
  37. Alexander EJ, Andriacchi TP. Correcting for deformation in skin-based marker systems. J Biomech . 2001 Mar;34(3):355–61.
  38. Neugebauer JM, Hawkins DA, Beckett L. Estimating Youth Locomotion Ground Reaction Forces Using an Accelerometer-Based Activity Monitor. Garatachea N, editor. PLoS One. 2012;7(10):e48182.
  39. Fox J, Weisberg S. Mixed-effects models in R. An R Companion to Applied Regression; SAGE: Thousand Oaks, CA, USA. 2002.

 

  1. Neugebauer JM, Collins KH, Hawkins DA. Ground Reaction Force Estimates from ActiGraph GT3X+ Hip Accelerations. Hug F, editor. PLoS One. 2014;9(6):e99023.
  2. Thiel D V., Shepherd J, Espinosa HG, Kenny M, Fischer K, Worsey M, et al. Predicting Ground Reaction Forces in Sprint Running Using a Shank Mounted Inertial Measurement Unit. In: The 12th Conference of the International Sports Engineering Association. Basel Switzerland: MDPI; 2018. p. 199.
  3. White JD, Carson N, Baum BS, Reinking MF, McPoil TG. Use of 2-dimensional sagittal kinematic variables to estimate ground reaction force during running. Int J Sports Phys Ther. 2019;14(2):174–9.
  4. Wouda FJ, Giuberti M, Bellusci G, Maartens E, Reenalda J, van Beijnum B-JF, et al. Estimation of Vertical Ground Reaction Forces and Sagittal Knee Kinematics During Running Using Three Inertial Sensors. Front Physiol. 2018;9:218.
  5. Ren L, Jones RK, Howard D. Whole body inverse dynamics over a complete gait cycle based only on measured kinematics. J Biomech. 2008;41(12):2750–9.
  6. Johnson WR, Mian A, Donnelly CJ, Lloyd D, Alderson J. Predicting athlete ground reaction forces and moments from motion capture. Med Biol Eng Comput. 2018;56(10):1781–92.
  7. De Bie T, Cristianini N, Rosipal R. Eigenproblems in Pattern Recognition. Handb Geom Comput. 2005;129–67.
  8. Komaris D-S, Perez-Valero E, Jordan L, Barton J, Hennessy L, O’Flynn B, et al. Predicting Three-Dimensional Ground Reaction Forces in Running by Using Artificial Neural Networks and Lower Body Kinematics. IEEE Access. 2019;7:156779–86.
  9. O’Reilly M, Caulfield B, Ward T, Johnston W, Doherty C. Wearable Inertial Sensor Systems for Lower Limb Exercise Detection and Evaluation: A Systematic Review. Sport Med. 2018;48(5):1221–46.
  10. Davidson P, Virekunnas H, Sharma D, Piché R, Cronin N. Continuous Analysis of Running Mechanics by Means of an Integrated INS/GPS Device. Sensors. 2019;19(6):1480.
  11. Pogson M, Verheul J, Robinson MA, Vanrenterghem J, Lisboa P. A neural network method to predict task- and step-specific ground reaction force magnitudes from trunk accelerations during running activities. Med Eng Phys. 2020;78:82–9.
  12. Akenhead R, Nassis GP. Training Load and Player Monitoring in High-Level Football: Current Practice and Perceptions. Int J Sports Physiol Perform. 2016;11(5):587–93.
  13. Girka A, Kulmala J, Äyrämö S. Deep learning approach for prediction of impact peak appearance at ground reaction force signal of running activity. Comput Methods Biomech Biomed Engin. 2020;23(14):1052–9.
  14. Tedesco S, Perez-Valero E, Komaris D, Jordan L, Barton J, Hennessy L, et al. Wearable motion sensors and artificial neural network for the estimation of vertical ground reaction forces in running. In: 2020 IEEE SENSORS. IEEE; 2020. p. 1–4.
  15. Johnson WR, Mian A, Robinson MA, Verheul J, Lloyd DG, Alderson JA. Multidimensional Ground Reaction Forces and Moments From Wearable Sensor Accelerations via Deep Learning. IEEE Trans Biomed Eng. 2021;68(1):289–97.
  16. Sharma D, Davidson P, Müller P, Piché R. Indirect Estimation of Vertical Ground Reaction Force from a Body-Mounted INS/GPS Using Machine Learning. Sensors. 2021;21(4):1553.