The Future of Fall Prevention: Integrating OpenPose with Cutting-Edge ML Models

Shina Samuel Kolawole; Gautam Siddharth Kashyap; Olamide Emmanuel Kolawole; Miao Yu

doi:10.4108/eetpht.11.9013

Authors

Shina Samuel Kolawole University of Lincoln
Gautam Siddharth Kashyap Indraprastha Institute of Information Technology Delhi
Olamide Emmanuel Kolawole Aston University
Miao Yu University of Lincoln

DOI:

https://doi.org/10.4108/eetpht.11.9013

Keywords:

Deep Learning, Fall, Healthcare, Machine Learning, OpenPose

Abstract

The research paper aims to assess ML models for video-recorded gaits with an aim of classifying people into high or low risks to fall groups. Several ML algorithms were tried employing OpenPose for CV, with RF showing the best outcomes: 93% accuracy along with F1-score as well as balanced sensitivity (93.50%) as well as specificity (92.50%). Some important determining factors were speed per unit distance, angle among other statistical measures. In comparison to wearables-based DL approaches plus traditional fall detection methods, this study’s approach showed higher accuracy and adaptability within health care settings.

Downloads

No metrics available.

-

see details

References

[1] Song W, Latham NK, Liu L, Rice HE, Sainlaire M, Min L, et al. Improved accuracy and efficiency of primary care fall risk screening of older adults using a machine learning approach. J Am Geriatr Soc [Internet]. 2024 Apr 1 [cited 2024 Jun 5];72(4):1145–54. Available from: https://onlinelibrary.wiley.com/doi/full/10.1111/jgs.18776

[2] Steckhan GM, Warner LM, Fleig L. Falls prevention is more than just promoting physical health: evaluation of the group-based, out-patient prevention program ‘Staying safe and active in old age – falls prevention.’ Heal Psychol Behav Med [Internet]. 2024 Dec 31 [cited 2024 Jun 5];12(1). Available from: https://www.tandfonline.com/doi/abs/10.1080/21642850.2024.2358915

[3] Bevilacqua G, Westbury LD, Bloom I, Zhang J, Lawrence WT, Barker ME, et al. General self-efficacy, not musculoskeletal health, was associated with social isolation and loneliness in older adults during the COVID-19 pandemic: findings from the Hertfordshire Cohort Study. Aging Clin Exp Res [Internet]. 2024 Dec 1 [cited 2024 Jun 5];36(1):1–9. Available from: https://link.springer.com/article/10.1007/s40520-023-02676-5

[4] Banerjee S, Mihailovic A, Miller R, Jian-Yu E, Gitlin LN, Xiong Y, et al. Visual Impairment and Real-World Home Physical Activity with Home Environment in an Older Population. JAMA Ophthalmol [Internet]. 2024 Mar 1 [cited 2024 Jun 5];142(3):208–14. Available from: https://jamanetwork.com/journals/jamaophthalmology/fullarticle/2814751

[5] Ayvat E, Doğan M, Ayvat F, Kılınç ÖO, Sütçü G, Kılınç M, et al. Usefulness of the Berg Balance Scale for prediction of fall risk in multiple sclerosis. Neurol Sci [Internet]. 2024 Jun 1 [cited 2024 Jun 5];45(6):2801–5. Available from: https://link.springer.com/article/10.1007/s10072-024-07318-w

[6] Takase M, Kisanuki N, Nakayoshi Y, Uemura C, Sato Y, Yamamoto M. Exploring nurses’ clinical judgment concerning the relative importance of fall risk factors: A mixed method approach using the Q Methodology. Int J Nurs Stud. 2024 May 1;153:104720.

[7] Jahangiri S, Abdollahi M, Patil R, Rashedi E, Azadeh-Fard N. An inpatient fall risk assessment tool: Application of machine learning models on intrinsic and extrinsic risk factors. Mach Learn with Appl. 2024 Mar 1;15:100519.

[8] Liang HW, Ameri R, Band S, Chen HS, Ho SY, Zaidan B, et al. Fall risk classification with posturographic parameters in community-dwelling older adults: a machine learning and explainable artificial intelligence approach. J Neuroeng Rehabil [Internet]. 2024 Dec 1 [cited 2024 Jun 5];21(1):1–17. Available from: https://link.springer.com/articles/10.1186/s12984-024-01310-3

[9] Takada H, Yamashita K, Osawa L, Komiyama Y, Muraoka M, Suzuki Y, et al. Assessment of lower limb muscle strength can predict fall risk in patients with chronic liver disease. Sci Rep [Internet]. 2024 Jan 2 [cited 2024 Jun 5];14(1):1–10. Available from: https://www.nature.com/articles/s41598-023-50574-7

[10] Murayama A, Higuchi D, Saida K, Tanaka S, Shinohara T. Fall Risk Prediction for Community-Dwelling Older Adults: Analysis of Assessment Scale and Evaluation Items without Actual Measurement. Int J Environ Res Public Health [Internet]. 2024 Feb 14 [cited 2024 Jun 5];21(2):224. Available from: https://www.mdpi.com/1660-4601/21/2/224/htm

[11] Olesen AE, Væver TJ, Simonsen M, Simonsen PG, Høj K. Deprescribing in primary care without deterioration of health-related outcomes: A real-life, quality improvement project. Basic Clin Pharmacol Toxicol [Internet]. 2024 Jan 1 [cited 2024 Jun 5];134(1):72–82. Available from: https://onlinelibrary.wiley.com/doi/full/10.1111/bcpt.13925

[12] Aarts J, Saddal SRD, Bosmans JE, de Groot V, de Jong BA, Klein M, et al. Combining Exercise and Cognitive Training to Postpone Cognitive Decline in People with Multiple Sclerosis: Part of the Don’t be late! Study Protocol. Mult Scler Relat Disord. 2024 Jun 1;86:105590.

[13] Yi MK, Han KH, Hwang SO. Fall Detection of the Elderly Using Denoising LSTM-based Convolutional Variant Autoencoder. IEEE Sens J. 2024;

[14] Yu J, Yang X, Deng Y, Krefman AE, Pool LR, Zhao L, et al. Incorporating longitudinal history of risk factors into atherosclerotic cardiovascular disease risk prediction using deep learning. Sci Rep [Internet]. 2024 Jan 31 [cited 2024 Jun 5];14(1):1–10. Available from: https://www.nature.com/articles/s41598-024-51685-5

[15] Habib Z, Mughal MA, Khan MA, Shabaz M. WiFOG: Integrating deep learning and hybrid feature selection for accurate freezing of gait detection. Alexandria Eng J. 2024 Jan 1;86:481–93.

[16] Huang D, Wu C, Wang Y, Zhang Z, Chen C, Li L, et al. Episode-level prediction of freezing of gait based on wearable inertial signals using a deep neural network model. Biomed Signal Process Control. 2024 Feb 1;88:105613.

[17] Sruthi P, Udgata SK. Wi-Fi sensing based person identification and activity recognition using two-phase deep learning model. Eng Appl Artif Intell. 2024 Jun 1;132:107904.

[18] Liu PL, Chang CC. Simple method integrating OpenPose and RGB-D camera for identifying 3D body landmark locations in various postures. Int J Ind Ergon. 2022 Sep 1;91:103354.

[19] Hu R, Diao Y, Wang Y, Li G, He R, Ning Y, et al. Effective evaluation of HGcnMLP method for markerless 3D pose estimation of musculoskeletal diseases patients based on smartphone monocular video. Front Bioeng Biotechnol. 2023 Jan 9;11:1335251.

[20] Mora A, Ozeki T, Mouri T, Barber R. A ROS-Integrated Skeleton Detection Framework Enabling Human-Centered Robotic Applications. 2024 IEEE Int Conf Auton Robot Syst Compet [Internet]. 2024 May 2 [cited 2024 Jun 5];3–8. Available from: https://ieeexplore.ieee.org/document/10535944/

[21] Sheng B, Chen L, Cheng J, Zhang Y, Hua Z, Tao J. A markless 3D human motion data acquisition method based on the binocular stereo vision and lightweight open pose algorithm. Meas J Int Meas Confed. 2024 Feb 15;225:113908.

[22] Murugan AS, Noh G, Jung H, Kim E, Kim K, You H, et al. Optimising computer vision-based ergonomic assessments: sensitivity to camera position and monocular 3D pose model. Ergonomics [Internet]. 2024 [cited 2024 Jun 5]; Available from: https://www.tandfonline.com/doi/abs/10.1080/00140139.2024.2304578

[23] Dutt M, Gupta A, Goodwin M, Omlin CW. An Interpretable Modular Deep Learning Framework for Video-Based Fall Detection. Appl Sci 2024, Vol 14, Page 4722 [Internet]. 2024 May 30 [cited 2024 Jun 5];14(11):4722. Available from: https://www.mdpi.com/2076-3417/14/11/4722/htm

[24] Nishiyama D, Arita S, Fukui D, Yamanaka M, Yamada H. Accurate fall risk classification in elderly using one gait cycle data and machine learning. Clin Biomech. 2024 May 1;115:106262.

[25] Jahandideh S, Hutchinson AF, Bucknall TK, Considine J, Driscoll A, Manias E, et al. Using machine learning models to predict falls in hospitalised adults. Int J Med Inform. 2024 Jul 1;187:105436.

[26] Juneau P, Lemaire ED, Bavec A, Burger H, Baddour N. Automated step detection with 6-minute walk test smartphone sensors signals for fall risk classification in lower limb amputees. PLOS Digit Heal [Internet]. 2022 Aug 18 [cited 2024 Jun 5];1(8):e0000088. Available from: /pmc/articles/PMC9931302/

[27] Sellers WI, Cross CF, Fukuhara A, Ishiguro A, Hirasaki E. Producing non-steady-state gaits (starting, stopping, and turning) in a biologically realistic quadrupedal simulation. Front Ecol Evol [Internet]. 2022 Sep 8 [cited 2024 Jun 5];10:954838. Available from: https://tohoku.elsevierpure.com/en/publications/producing-non-steady-state-gaits-starting-stopping-and-turning-in

[28] Kour N, Sunanda, Arora S. A Vision-Based Gait Dataset for Knee Osteoarthritis and Parkinson’s Disease Analysis with Severity Levels. In Mendeley Data; 2022. p. 303–17.