Using Machine Learning to Identify Feelings of Energy and Fatigue in Single-Task Walking Gait: An Exploratory Study
Abstract
:1. Introduction
2. Methods
2.1. Study Design
2.2. Participants
2.3. Instruments
2.3.1. Self-Reported Feelings of Energy and Fatigue
2.3.2. Gait
2.4. Procedure
2.5. Statistical Analysis
2.5.1. Pre-Processing of Data
2.5.2. Primary Analyses
- Classification of participants
3. Results
3.1. Feelings of Energy (Vigor)
3.1.1. Feature Importance
3.1.2. Model Training
3.1.3. Differences in Gait Characteristics between Groups
3.2. Feelings of Fatigue
3.2.1. Feature Importance
3.2.2. Model Training
3.2.3. Differences in Gait Characteristics between Groups
3.3. Energy and Fatigue Combined
3.3.1. Feature Importance
3.3.2. Model Training
3.3.3. Differences in Gait Characteristics between Groups
4. Discussion
4.1. Comparing Most Important Features
4.2. Model Accuracy
4.3. Comparing Gait Characteristics between Groups
4.3.1. High Energy vs. Low Energy
4.3.2. High Fatigue vs. Low Fatigue
4.3.3. Comparison between Classes
4.4. Limitations
4.5. Implications
5. Conclusions
Supplementary Materials
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Acknowledgments
Conflicts of Interest
References
- Bergen, G. Falls and fall injuries among adults aged ≥65 years—United States, 2014. MMWR Morb. Mortal. Wkly. Rep. 2016, 65, 993–998. [Google Scholar] [CrossRef]
- Peel, N.M. Epidemiology of falls in older age. Can. J. Aging Rev. Can. Vieil. 2011, 30, 7–19. [Google Scholar] [CrossRef] [PubMed]
- Florence, C.S.; Bergen, G.; Atherly, A.; Burns, E.; Stevens, J.; Drake, C. Medical costs of fatal and nonfatal falls in older adults. J. Am. Geriatr. Soc. 2018, 66, 693–698. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Rubenstein, L.Z. Falls in older people: Epidemiology, risk factors and strategies for prevention. Age Ageing 2006, 35 (Suppl. 2), ii37–ii41. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Grobe, S.; Kakar, R.S.; Smith, M.L.; Mehta, R.; Baghurst, T.; Boolani, A. Impact of cognitive fatigue on gait and sway among older adults: A literature review. Prev. Med. Rep. 2017, 6, 88–93. [Google Scholar] [CrossRef] [PubMed]
- Behrens, M.; Mau-Moeller, A.; Lischke, A.; Katlun, F.; Gube, M.; Zschorlich, V.; Skripitz, R.; Weippert, M. Mental fatigue increases gait variability during dual-task walking in old adults. J. Gerontol. Ser. A 2017, 73, 792–797. [Google Scholar] [CrossRef] [PubMed]
- Morris, A.J.; Christie, A.D. The Effect of Mental Fatigue on Neuromuscular Function is Similar in Young and Older Women. Brain Sci. 2020, 10, 191. [Google Scholar] [CrossRef] [Green Version]
- Morris, A.J.; Christie, A.D. The effect of a mentally fatiguing task on postural balance control in young and older women. Exp. Gerontol. 2020, 132, 110840. [Google Scholar] [CrossRef]
- Vestergaard, S.; Nayfield, S.G.; Patel, K.V.; Eldadah, B.; Cesari, M.; Ferrucci, L.; Ceresini, G.; Guralnik, J.M. Fatigue in a representative population of older persons and its association with functional impairment, functional limitation, and disability. J. Gerontol. Ser. A Biomed. Sci. Med. Sci. 2009, 64, 76–82. [Google Scholar] [CrossRef] [Green Version]
- Verschueren, J.O.; Tassignon, B.; Proost, M.; Teugels, A.; Roelands, B.; Verhagen, E.; Meeusen, R. Does mental fatigue negatively affect outcomes of functional performance tests? Med. Sci. Sports Exerc. 2020, 52, 2002–2010. [Google Scholar] [CrossRef]
- Adão Martins, N.R.; Annaheim, S.; Spengler, C.M.; Rossi, R.M. Fatigue monitoring through wearables: A state-of-the-art review. Front. Physiol. 2021, 2285. [Google Scholar] [CrossRef]
- Maman, Z.S.; Chen, Y.-J.; Baghdadi, A.; Lombardo, S.; Cavuoto, L.A.; Megahed, F.M. A data analytic framework for physical fatigue management using wearable sensors. Expert Syst. Appl. 2020, 155, 113405. [Google Scholar] [CrossRef]
- Ameli, S.; Naghdy, F.; Stirling, D.; Naghdy, G.; Aghmesheh, M. Quantitative and non-invasive measurement of exercise-induced fatigue. Proc. Inst. Mech. Eng. Part P J. Sports Eng. Technol. 2019, 233, 34–45. [Google Scholar] [CrossRef]
- Maman, Z.S.; Yazdi, M.A.A.; Cavuoto, L.A.; Megahed, F.M. A data-driven approach to modeling physical fatigue in the workplace using wearable sensors. Appl. Ergon. 2017, 65, 515–529. [Google Scholar] [CrossRef]
- Aryal, A.; Ghahramani, A.; Becerik-Gerber, B. Monitoring fatigue in construction workers using physiological measurements. Autom. Constr. 2017, 82, 154–165. [Google Scholar] [CrossRef]
- Buckley, C.; O’Reilly, M.A.; Whelan, D.; Farrell, A.V.; Clark, L.; Longo, V.; Gilchrist, M.D.; Caulfield, B. Binary classification of running fatigue using a single inertial measurement unit. In Proceedings of the 2017 IEEE 14th International Conference on Wearable and Implantable Body Sensor Networks (BSN), Eindhoven, The Netherlands, 5–12 May 2017; pp. 197–201. [Google Scholar]
- McDowell, C.P.; Campbell, M.J.; Herring, M.P. Sex-Related Differences in Mood Responses to Acute Aerobic Exercise. Med. Sci. Sports Exerc. 2016, 48, 1798–1802. [Google Scholar] [CrossRef] [Green Version]
- Herring, M.P.; Lindheimer, J.B.; O’Connor, P.J. The effects of exercise training on anxiety. Am. J. Lifestyle Med. 2014, 8, 388–403. [Google Scholar] [CrossRef]
- Herring, M.P.; O’Connor, P.J. The effect of acute resistance exercise on feelings of energy and fatigue. J. Sports Sci. 2009, 27, 701–709. [Google Scholar] [CrossRef]
- Loy, B.D.; O’Connor, P.J.; Dishman, R.K. The effect of a single bout of exercise on energy and fatigue states: A systematic review and meta-analysis. Fatigue Biomed. Health Behav. 2013, 1, 223–242. [Google Scholar] [CrossRef]
- Ward-Ritacco, C.; Poudevigne, M.S.; O’Connor, P.J. Muscle strengthening exercises during pregnancy are associated with increased energy and reduced fatigue. J. Psychosom. Obstet. Gynecol. 2016, 37, 68–72. [Google Scholar] [CrossRef] [Green Version]
- Loy, B.D.; Cameron, M.H.; O’Connor, P.J. Perceived fatigue and energy are independent unipolar states: Supporting evidence. Med. Hypotheses 2018, 113, 46–51. [Google Scholar] [CrossRef] [PubMed]
- Eshragh, J.; Dhruva, A.; Paul, S.M.; Cooper, B.A.; Mastick, J.; Hamolsky, D.; Levine, J.D.; Miaskowski, C.; Kober, K.M. Associations between neurotransmitter genes and fatigue and energy levels in women after breast cancer surgery. J. Pain Symptom Manag. 2017, 53, 67–84.e7. [Google Scholar] [CrossRef]
- Boolani, A.; O’Connor, P.J.; Reid, J.; Ma, S.; Mondal, S. Predictors of feelings of energy differ from predictors of fatigue. Fatigue Biomed. Health Behav. 2019, 7, 12–28. [Google Scholar] [CrossRef]
- Boolani, A.; Gallivan, K.M.; Ondrak, K.S.; Christopher, C.J.; Castro, H.F.; Campagna, S.R.; Taylor, C.M.; Luo, M.; Dowd, S.E.; Smith, M.L.; et al. Trait Energy and Fatigue May Be Connected to Gut Bacteria among Young Physically Active Adults: An Exploratory Study. Nutrients 2022, 14, 466. [Google Scholar] [CrossRef] [PubMed]
- Dupree, E.J.; Goodwin, A.; Darie, C.C.; Boolani, A. A Pilot exploratory Proteomics Investigation of mental fatigue and mental energy. In Advancements of Mass Spectrometry in Biomedical Research; Springer: Heidelberg, Germany, 2019; pp. 601–611. [Google Scholar]
- Loy, B.D.; O’Connor, P.J. The effect of histamine on changes in mental energy and fatigue after a single bout of exercise. Physiol. Behav. 2016, 153, 7–18. [Google Scholar] [CrossRef] [PubMed]
- Manierre, M.; Jansen, E.; Boolani, A. Sleep quality and sex modify the relationships between trait energy and fatigue on state energy and fatigue. PLoS ONE 2020, 15, e0227511. [Google Scholar] [CrossRef] [PubMed]
- Boolani, A.; Manierre, M. An exploratory multivariate study examining correlates of trait mental and physical fatigue and energy. Fatigue Biomed. Health Behav. 2019, 7, 29–40. [Google Scholar] [CrossRef]
- Boolani, A.; Ryan, J.; Vo, T.; Wong, B.; Banerjee, N.K.; Banerjee, S.; Fulk, G.; Smith, M.L.; Martin, R. Do changes in mental energy and fatigue impact functional assessments associated with fall risks? an exploratory study using machine learning. Phys. Occup. Ther. Geriatr. 2020, 38, 283–301. [Google Scholar] [CrossRef]
- Kowalski, K.L.; Boolani, A.; Christie, A.D. Sex differences in the impact of state and trait fatigue on gait variability. Hum. Mov. Sci. 2021, 80, 102884. [Google Scholar] [CrossRef]
- Kowalski, K.L.; Boolani, A.; Christie, A.D. State and Trait Fatigue and Energy Predictors of Postural Control and Gait. Mot. Control 2021, 1, 519–536. [Google Scholar] [CrossRef]
- Boolani, A.; Allen, A.; Barrios, N.; Sames, C. Associaiton between trait energy and fatigue and aquatic functional tests: An exploratory study. J. Aquat. Ther. 2022, in press. [Google Scholar]
- Sprague, B.N.; Zhu, X.; Ehrenkranz, R.C.; Tian, Q.; Gmelin, T.A.; Glynn, N.W.; Rosso, A.L.; Rosano, C. Declining energy predicts incident mobility disability and mortality risk in healthy older adults. J. Am. Geriatr. Soc. 2021, 69, 3134–3141. [Google Scholar] [CrossRef]
- Mahoney, G.; Martin, J.; Martin, R.; Yager, C.; Smith, M.L.; Grin, Z.; Vogel-Rosbrook, C.; Bradley, D.; Appiah-Kubi, K.O.; Boolani, A. Evidence that feelings of energy and fatigue are associated differently with gait characteristics and balance: An exploratory study. Fatigue Biomed. Health Behav. 2021, 9, 125–138. [Google Scholar] [CrossRef]
- Mahlknecht, P.; Kiechl, S.; Bloem, B.R.; Willeit, J.; Scherfler, C.; Gasperi, A.; Rungger, G.; Poewe, W.; Seppi, K. Prevalence and burden of gait disorders in elderly men and women aged 60–97 years: A population-based study. PLoS ONE 2013, 8, e69627. [Google Scholar] [CrossRef]
- McNair, D.M.; Lorr, M.; Heuchert, J.W.P.; Droppleman, L.F. Profile of Mood States—Brief form North Tonawanda: Multi-Health Systems 2003; Multi-Health Systems Inc.: North Tonawanda, NY, USA, 2003. [Google Scholar]
- Curran, S.L.; Andrykowski, M.A.; Studts, J.L. Short form of the Profile of Mood States (POMS-SF): Psychometric information. Psychol. Assess. 1995, 7, 80. [Google Scholar] [CrossRef]
- Zhao, N.; Zhang, Z.; Wang, Y.; Wang, J.; Li, B.; Zhu, T.; Xiang, Y. See your mental state from your walk: Recognizing anxiety and depression through Kinect-recorded gait data. PLoS ONE 2019, 14, e0216591. [Google Scholar] [CrossRef] [Green Version]
- Miao, B.; Liu, X.; Zhu, T. Automatic mental health identification method based on natural gait pattern. PsyCh J. 2021, 10, 453–464. [Google Scholar] [CrossRef]
- Washabaugh, E.P.; Kalyanaraman, T.; Adamczyk, P.G.; Claflin, E.S.; Krishnan, C. Validity and repeatability of inertial measurement units for measuring gait parameters. Gait Posture 2017, 55, 87–93. [Google Scholar] [CrossRef] [Green Version]
- Mancini, M.; Chiari, L.; Holmstrom, L.; Salarian, A.; Horak, F.B. Validity and reliability of an IMU-based method to detect APAs prior to gait initiation. Gait Posture 2016, 43, 125–131. [Google Scholar] [CrossRef] [Green Version]
- Mancini, M.; King, L.; Salarian, A.; Holmstrom, L.; McNames, J.; Horak, F.B. Mobility lab to assess balance and gait with synchronized body-worn sensors. J. Bioeng. Biomed. Sci. 2011, 007. [Google Scholar] [CrossRef]
- Liu, X.; Sun, B.; Zhang, Z.; Wang, Y.; Tang, H.; Zhu, T. Gait can reveal sleep quality with machine learning models. PLoS ONE 2019, 14, e0223012. [Google Scholar] [CrossRef] [PubMed]
- Hatchett, A.; Armstrong, K.; Parr, B.; Crews, M.; Tant, C. The effect of a curved non-motorized treadmill on running gait length, imbalance and stride angle. Sports 2018, 6, 58. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Simoni, L.; Scarton, A.; Gerli, F.; Macchi, C.; Gori, F.; Pasquini, G.; Pogliaghi, S. Testing the Performance of an Innovative Markerless Technique for Quantitative and Qualitative Gait Analysis. Sensors 2020, 20, 6654. [Google Scholar] [CrossRef] [PubMed]
- Troyanskaya, O.G.; Botstein, D.; Altman, R.B. Missing Value Estimation. In A Practical Approach to Microarray Data Analysis; Berrar, D.P., Dubitzky, W., Granzow, M., Eds.; Springer: Boston, MA, USA, 2003; pp. 65–75. [Google Scholar]
- Tukey, J.W. The future of data analysis. Ann. Math. Stat. 1962, 33, 1–67. [Google Scholar] [CrossRef]
- Zhao, X.; Yu, B.; Liu, Y.; Chen, Z.; Li, Q.; Wang, C.; Wu, J. Estimation of poverty using random forest regression with multi-source data: A case study in Bangladesh. Remote Sens. 2019, 11, 375. [Google Scholar] [CrossRef] [Green Version]
- Berrar, D. Cross-validation. Encycl. Bioinform. Comput. Biol. 2019, 1, 542–545. [Google Scholar]
- Efron, B.; Tibshirani, R.J. An Introduction to the Bootstrap (Chapman & Hall/CRC Monographs on Statistics & Applied Probability). Chapman and Hall: Boca Raton, FL, USA; CRC: Boca Raton, FL, USA, 1994. [Google Scholar]
- Chernoff, H. Large-sample theory: Parametric case. Ann. Math. Stat. 1956, 27, 1–22. [Google Scholar] [CrossRef]
- Lehmann, E.L. Elements of Large-Sample Theory; Springer Science & Business Media: New York, NY, USA, 2004. [Google Scholar]
- Willmott, C.J.; Matsuura, K. Advantages of the mean absolute error (MAE) over the root mean square error (RMSE) in assessing average model performance. Clim. Res. 2005, 30, 79–82. [Google Scholar] [CrossRef]
- Boolani, A.; Martin, R.; Goodwin, A.; Avolio, A.; Sur, S.; Smith, M.L.; Fulk, G. Associations for tasks requiring single stimulus and working memory with different aspects of gait and posture: An exploratory study. Int. J. Rehabil. Res. 2019, 42, 160–167. [Google Scholar] [CrossRef]
- Feldman, R.; Schreiber, S.; Pick, C.G.; Been, E. Gait, balance, mobility and muscle strength in people with anxiety compared to healthy individuals. Hum. Mov. Sci. 2019, 67, 102513. [Google Scholar] [CrossRef]
- Feldman, R.; Schreiber, S.; Pick, C.G.; Been, E. Gait, Balance and Posture in Major Mental Illnesses: Depression, Anxiety and Schizophrenia. Austin Med. Sci. 2020, 5, 1039. [Google Scholar]
- Devereaux, C.; Yager, C.; Pickett, A.C.; Smith, M.L.; Martin, J.; Boolani, A. Predictors of Feelings of Anxiety in Graduate Allied Health Students: An Exploratory Study. J. Allied Health 2021, 50, 73E–77E. [Google Scholar]
- Nagano, H.; James, L.; Sparrow, W.A.; Begg, R.K. Effects of walking-induced fatigue on gait function and tripping risks in older adults. J. Neuroeng. Rehabil. 2014, 11, 155. [Google Scholar] [CrossRef] [Green Version]
- Hausdorff, J.M. Gait variability: Methods, modeling and meaning. J. Neuroeng. Rehabil. 2005, 2, 19. [Google Scholar] [CrossRef] [Green Version]
- Verghese, J.; Holtzer, R.; Lipton, R.B.; Wang, C. Quantitative gait markers and incident fall risk in older adults. J. Gerontol. Ser. A 2009, 64, 896–901. [Google Scholar] [CrossRef] [Green Version]
- Michalak, J.; Troje, N.F.; Fischer, J.; Vollmar, P.; Heidenreich, T.; Schulte, D. Embodiment of sadness and depression—Gait patterns associated with dysphoric mood. Psychosom. Med. 2009, 71, 580–587. [Google Scholar] [CrossRef] [Green Version]
- Lemke, M.R.; Wendorff, T.; Mieth, B.; Buhl, K.; Linnemann, M. Spatiotemporal gait patterns during over ground locomotion in major depression compared with healthy controls. J. Psychiatr. Res. 2000, 34, 277–283. [Google Scholar] [CrossRef]
- Boolani, A.; Yager, C.; Reid, J.; Lackman, J.; Smith, M.L. Correlates of depressive mood among graduate-level allied health students: An exploratory study examining trait energy and fatigue. J. Am. Coll. Health 2021, 1–12. [Google Scholar] [CrossRef]
- Timmis, M.A.; Pardhan, S. Patients with central visual field loss adopt a cautious gait strategy during tasks that present a high risk of falling. Investig. Ophthalmol. Vis. Sci. 2012, 53, 4120–4129. [Google Scholar] [CrossRef] [Green Version]
- Boudarham, J.; Roche, N.; Pradon, D.; Bonnyaud, C.; Bensmail, D.; Zory, R. Variations in kinematics during clinical gait analysis in stroke patients. PLoS ONE 2013, 8, e66421. [Google Scholar] [CrossRef]
- Barrios, N.; Sames, C.; Allen, A.; Viera, E.C.; Boolani, A. Intensity of aquatic physical activity is best when tailored to the participants mental and physical capabilities: An exploratory study. J. Aquat. Phys. Ther. 2022, in press. [Google Scholar]
- Miller, M.; Lee-Chambers, J.; Cooper, B.; Boolani, A.; Jansen, E. Associations between physical activity and energy and fatigue depend on sleep quality. Fatigue Biomed. Health Behav. 2020, 8, 193–204. [Google Scholar] [CrossRef]
Variable | Energy | Fatigue | Energy + Fatigue | ||||||||
---|---|---|---|---|---|---|---|---|---|---|---|
Low Energy | High Energy | Sig. | Low Fatigue | High Fatigue | Sig. | High Energy/High Fatigue | Low Energy/High Fatigue | High Energy/Low Fatigue | Low Energy/Low Fatigue | Sig. | |
Male:Female | 20:43 | 27:36 | 0.197 | 24:29 | 23:50 | 0.114 | 13:20 | 14:16 | 10:30 | 10:13 | 0.246 |
Age | 24.25 ± 3.94 | 23.6 ± 3.52 | 0.330 | 23.87 ± 3.93 | 23.97 ± 3.62 | 0.877 | 23.03 ± 2.99 | 24.23 ± 3.97 | 24.75 ± 3.93 | 23.39 ± 3.9 | 0.817 |
Height | 172.56 ± 8.42 | 173.61 ± 8.61 | 0.491 | 173.77 ± 6.84 | 172.58 ± 9.54 | 0.439 | 173.49 ± 9.3 | 173.74 ± 7.94 | 171.83 ± 9.79 | 173.82 ± 5.24 | 0.805 |
Weight | 71.65 ± 12.86 | 74.46 ± 13.66 | 0.236 | 74.08 ± 12.66 | 72.31 ± 13.77 | 0.464 | 73.45 ± 14.12 | 75.58 ± 13.28 | 71.38 ± 13.57 | 72.12 ± 11.81 | 0.859 |
BMI | 23.88 ± 3.68 | 24.63 ± 3.81 | 0.263 | 24.36 ± 3.53 | 24.18 ± 3.92 | 0.791 | 24.39 ± 4.22 | 24.89 ± 3.36 | 24 ± 3.7 | 23.66 ± 3.7 | 0.643 |
Level of Importance | Energy * | Fatigue * | Energy + Fatigue ** | |||
---|---|---|---|---|---|---|
Ranking | Relative Importance | Gait Characteristic | Relative Importance | Gait Characteristic | Relative Importance | Gait Characteristic |
1 | 0.159 | Variation in Turns Velocity (°/s) | 0.146 | Mean Max. Lumbar R Lat. bend (°) | 0.138 | Variation Max. Lumbar Rot ROM to the right (°) |
2 | 0.154 | Imbalance in DLS-GCT between legs (%) | 0.142 | Variation L Heel Strike Angle (°) | 0.108 | Variation in Turns Angle (°) |
3 | 0.124 | Variation Lumbar Flexion/Extension ROM (°) | 0.139 | Variation in Turns Angle (°) | 0.046 | APA 1st step ROM (°) |
4 | 0.094 | Variation R leg Stride Length (m) | 0.131 | Variation in R leg elevation at mid-stance (cm) | 0.031 | Variation L DLS-GCT (%) |
5 | 0.091 | Variation L leg Circumduction (cm) | 0.087 | Mean L Toe Out Angle (°) | −0.031 | Variation L Stride Length (m) |
6 | 0.086 | Mean R leg DLS-GCT (%) | 0.08 | Mean Min. Lumbar Flexion/Extension ROM (°) | −0.092 | Mean Max. Neck Flexion/Extension ROM (°) |
7 | 0.067 | Mean Max. Lumbar Rot ROM (°) | 0.067 | Variation Max. Lumbar L Rot ROM (°) | −0.092 | Mean Neck Flexion/Extension ROM (°) |
8 | 0.049 | Imbalance in Cadence between legs (%) | 0.060 | Imbalance in step variability between legs (%) | −0.092 | Variation L leg Circumduction (cm) |
9 | 0.047 | Weight (kg) | 0.057 | Variation R Toe Off Angle (°) | −0.108 | Lateral APA Peak (m/s2) |
10 | 0.046 | Variation R leg cadence (s) | 0.043 | Mean Trunk Flexion/Extension ROM (°) | −0.108 | Variation R leg elevation at mid-stance (cm) |
11 | 0.042 | APA 1st step time (s) | 0.033 | Variation Max. Lumbar R Rot ROM (°) | −0.108 | Variation Step Time (s) |
12 | 0.042 | Mean Neck R Rot ROM (°) | 0.016 | APA 1st step Time (s) | −0.123 | APA 1st step Time (s) |
Level of Importance | Energy * | Fatigue * | ||
---|---|---|---|---|
Ranking | Relative Importance | Gait Characteristic | Relative Importance | Gait Characteristic |
1 | 0.162 | Number of turns (#) | 0.230 | APA 1st step ROM (°) |
2 | 0.140 | Mean Max. Neck R Lat. bend ROM (°) | 0.228 | Mean Max. Neck Flex/Ext ROM (°) |
3 | 0.103 | Mean Min. Neck Flex/Ext ROM (°) | 0.085 | Imbalance Mean Toe Out Angle between legs (°) |
4 | 0.089 | Mean arm swing velocity (°/s) | 0.084 | Mean Max. Neck L Lat. bend ROM (°) |
5 | 0.086 | Imbalance in cadence between legs (%) | 0.067 | Variation R DLS-GCT (%) |
6 | 0.080 | Mean R Cadence (steps/s) | 0.066 | Variation R Arm Swing Velocity (°) |
7 | 0.073 | Mean Max. Back Flex/Ext ROM (°) | 0.061 | Variation L Cadence (steps/s) |
8 | 0.073 | Mean Gait Speed of both legs (m/s) | 0.053 | Mean Neck R Lat. bend ROM (°) |
9 | 0.063 | Mean Cadence of both legs (steps/s) | 0.050 | Variation Max. Lumbar R Rot ROM (°) |
10 | 0.056 | Mean L Cadence (steps/s) | 0.045 | Variation R SLS-GCT (%) |
11 | 0.044 | Imbalance in Gait Cycle Time between legs (%) | 0.021 | Imbalance in Gait Cycle Time between legs (%) |
12 | 0.031 | Sex | 0.009 | Sex |
Construct | Classifier | Mean | 95% CI | Minimum | Q1 | Q2 | Q3 | Maximum |
---|---|---|---|---|---|---|---|---|
Energy | Gradient Boosting Classifier | 0.743 | 0.708–0.779 | 0.539 | 0.673 | 0.750 | 0.769 | 1 |
Fatigue | Gradient Boosting Classifier | 0.742 | 0.696–0.788 | 0.461 | 0.692 | 0.760 | 0.846 | 0.923 |
Energy/Fatigue | Gaussian Naïve Bayes | 0.455 | 0.398–0.512 | 0.166 | 0.314 | 0.461 | 0.538 | 0.833 |
Regressor R2 | ||||||||
Energy | Random Forest Regressor—Bootstrapped | 0.884 | 0.86–00.90 | 0.863 | 0.875 | 0.881 | 0.888 | 0.900 |
Energy | Random Forest Regressor—with K-fold | 0.310 | 0.30–0.32 | 0.30 | 0.281 | 0.283 | 0.321 | 0.321 |
Fatigue | Random Forest Regressor—Bootstrapped | 0.886 | 0.859–0.91 | 0.862 | 0.876 | 0.885 | 0.895 | 0.895 |
Fatigue | Random Forest Regressor—with K-fold | 0.349 | 0.2–0.498 | 0.239 | 0.308 | 0.340 | 0.397 | 0.397 |
Regressor Mean Absolute Error | ||||||||
Energy | Random Forest Regressor—Bootstrapped | 0.006 | 0.004–0.008 | 0.004 | 0.005 | 0.006 | 0.007 | 0.008 |
Energy | Random Forest Regressor—with K-fold | 0.005 | 0.003–0.006 | 0.005 | 0.005 | 0.005 | 0.004 | 0.004 |
Fatigue | Random Forest Regressor—Bootstrapped | 0.005 | 0.003–006 | 0.004 | 0.004 | 0.005 | 0.005 | 0.006 |
Fatigue | Random Forest Regressor—with K-fold | 0.007 | 0.005–009 | 0.004 | 0.006 | 0.007 | 0.008 | 0.001 |
Variable | Low Energy | High Energy | Sig. | d | |
---|---|---|---|---|---|
Leg | Variation R cadence (steps/min) | 2.36 ± 0.76 | 2.69 ± 0.87 | 0.024 | −0.40 |
Variation R DLS-GCT (%) | 1.08 ± 0.23 | 1.18 ± 0.29 | 0.040 | −0.38 | |
Imbalance in DLS-GCT between legs (%) | 0.71 ± 0.42 | 0.52 ± 0.43 | 0.047 | 0.37 | |
Variation L gait speed (m/s) | 0.046 ± 0.015 | 0.052 ± 0.019 | 0.041 | −0.41 | |
Variation R gait speed (m/s) | 0.047 ± 0.013 | 0.054 ± 0.017 | 0.009 | −0.47 | |
Variation L toe off angle (°) | 1.34 ± 0.44 | 1.54 ± 0.5 | 0.019 | −0.45 | |
Variation R stride length (m) | 0.038 ± 0.01 | 0.042 ± 0.012 | 0.044 | −0.36 | |
Arm | Mean R arm swing velocity (°/s) | 162.54 ± 53.08 | 183.91 ± 60.57 | 0.018 | −0.40 |
Mean swing velocity for both arms (°/s) | 177.57 ± 51.69 | 198.41 ± 61.14 | 0.017 | −0.40 | |
Mean L ROM (°) | 42.85 ± 16.31 | 48.43 ± 17.78 | 0.033 | −0.38 | |
Mean R ROM (°) | 36.05 ± 13.85 | 42.07 ± 15.81 | 0.017 | −0.43 | |
Mean ROM for both arms (°) | 39.66 ± 14.04 | 45.08 ± 15.26 | 0.020 | −0.41 | |
Imbalance in arm ROM between arms (%) | 14.6 ± 10.04 | 11.6 ± 8.49 | 0.038 | 0.38 |
Variable | High Fatigue | Low Fatigue | Sig. | d | |
---|---|---|---|---|---|
Gait Initiation | APA first step time (s) | 0.59 ± 0.06 | 0.56 ± 0.07 | 0.032 | 0.39 |
Lumbar | Mean Min. Lumbar Flex/Ext ROM (°) | −1.68 ± 4.81 | −3.16 ± 3.9 | 0.043 | 0.37 |
Variation Max. Lumbar R Rot ROM (°) | 4.31 ± 2.97 | 5.36 ± 3.38 | 0.024 | −0.40 | |
Variation Max. Lumbar L Rot ROM (°) | 4.44 ± 2.92 | 5.48 ± 3.4 | 0.024 | −0.40 | |
Trunk | Variation Trunk Flex/Ext ROM (°) | 1.1 ± 0.35 | 1.2 ± 0.39 | 0.035 | −0.38 |
Leg | Imbalance in Gait Speed between legs (%) | 0.89 ± 0.67 | 1.15 ± 0.73 | 0.019 | −0.43 |
Mean L Toe Off Angle (°) | 37.21 ± 2.91 | 35.93 ± 3.41 | 0.041 | 0.37 | |
Mean R Toe Off Angle (°) | 37.32 ± 2.78 | 36 ± 3.38 | 0.029 | 0.39 | |
Imbalance in Stride Length between legs | 0.79 ± 0.57 | 1.01 ± 0.63 | 0.019 | −0.43 | |
Arm | Variation L Swing Velocity (°) | 41.28 ± 23.52 | 47.79 ± 28.94 | 0.032 | −0.36 |
Variation R Swing Velocity (°) | 36.86 ± 17.02 | 44 ± 25.63 | 0.017 | −0.38 | |
Turning | Variation Turns Angle (°) | 5.7 ± 1.44 | 5.02 ± 1.2 | 0.010 | 0.48 |
Variable | High Energy/Low Fatigue | High Energy/High Fatigue | Sig. | d |
---|---|---|---|---|
Means ± SD | Means ± SD | |||
Variation R Arm Swing Velocity (°/s) | 50.5 ± 26.9 | 35.24 ± 14.73 | 0.004 | 0.70 |
Variation L Heel Strike Angle (°) | 1.90 ± 0.48 | 1.61 ± 0.35 | 0.033 | 0.69 |
Variation R Arm ROM (°) | 8.66 ± 4.25 | 6.77 ± 2.29 | 0.045 | 0.55 |
Low Energy/Low Fatigue | Low Energy/High Fatigue | |||
Variation Max. Lumbar R Rot ROM (°) | 6.57 ± 3.74 | 4.15 ± 2.7 | 0.018 | 0.74 |
Variation Max. Back L Rot ROM (°) | 6.44 ± 3.82 | 4.11 ± 2.72 | 0.032 | 0.70 |
High Energy/High Fatigue | Low Energy/High Fatigue | |||
Imbalance DLS-GCT between legs (%) | 0.42 ± 0.41 | 0.73 ± 0.42 | 0.045 | −0.75 |
High Energy/Low Fatigue | Low Energy/High Fatigue | |||
Variation R Arm Swing Velocity (°/s) | 50.5 ± 26.9 | 38.19 ± 18.78 | 0.022 | 0.53 |
Mean R Arm ROM (°) | 44.72 ± 16.81 | 36.56 ± 14.23 | 0.048 | 0.52 |
High Energy/Low Fatigue | Low Energy/Low Fatigue | |||
Variation R Toe Off Angle (°) | 1.61 ± 0.5 | 1.3 ± 0.29 | 0.049 | 0.76 |
Imbalance Arm Swing velocity between arms (°/s) | 209.95 ± 65.67 | 164.94 ± 40.06 | 0.042 | 0.83 |
High Energy/High Fatigue | Low Energy/Low Fatigue | |||
Variation Turns Angle (°) | 6.07 ± 1.43 | 4.88 ± 1.13 | 0.018 | 0.92 |
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |
© 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Kadry, A.M.; Torad, A.; Elwan, M.A.; Kakar, R.S.; Bradley, D.; Chaudhry, S.; Boolani, A. Using Machine Learning to Identify Feelings of Energy and Fatigue in Single-Task Walking Gait: An Exploratory Study. Appl. Sci. 2022, 12, 3083. https://doi.org/10.3390/app12063083
Kadry AM, Torad A, Elwan MA, Kakar RS, Bradley D, Chaudhry S, Boolani A. Using Machine Learning to Identify Feelings of Energy and Fatigue in Single-Task Walking Gait: An Exploratory Study. Applied Sciences. 2022; 12(6):3083. https://doi.org/10.3390/app12063083
Chicago/Turabian StyleKadry, Ahmed M., Ahmed Torad, Moustafa Ali Elwan, Rumit Singh Kakar, Dylan Bradley, Shafique Chaudhry, and Ali Boolani. 2022. "Using Machine Learning to Identify Feelings of Energy and Fatigue in Single-Task Walking Gait: An Exploratory Study" Applied Sciences 12, no. 6: 3083. https://doi.org/10.3390/app12063083
APA StyleKadry, A. M., Torad, A., Elwan, M. A., Kakar, R. S., Bradley, D., Chaudhry, S., & Boolani, A. (2022). Using Machine Learning to Identify Feelings of Energy and Fatigue in Single-Task Walking Gait: An Exploratory Study. Applied Sciences, 12(6), 3083. https://doi.org/10.3390/app12063083