*Result*: An automatic approach to assess biomechanical risk using machine learning algorithms and inertial sensors.
Title:
An automatic approach to assess biomechanical risk using machine learning algorithms and inertial sensors.
Authors:
Prisco G; Department of Medicine and Health Sciences, University of Molise, Campobasso, Italy., Cesarelli M; Department of Engineering, University of Sannio, Benevento, Italy., Esposito F; Department of Advanced Medical and Surgical Sciences, University of Campania Luigi Vanvitelli, Naples, Italy., Santone A; Department of Medicine and Health Sciences, University of Molise, Campobasso, Italy., Gargiulo P; Institute of Biomedical and Neural Engineering, Reykjavik University, Reykjavik, Iceland., Amato F; Department of Information Technology and Electrical Engineering, University of Naples Federico II, Naples, Italy., Donisi L; Department of Advanced Medical and Surgical Sciences, University of Campania Luigi Vanvitelli, Naples, Italy. leandro.donisi@unicampania.it.
Source:
Physical and engineering sciences in medicine [Phys Eng Sci Med] 2026 Mar; Vol. 49 (1), pp. 101-113. Date of Electronic Publication: 2025 Oct 09.
Publication Type:
Journal Article
Language:
English
Journal Info:
Publisher: Springer Country of Publication: Switzerland NLM ID: 101760671 Publication Model: Print-Electronic Cited Medium: Internet ISSN: 2662-4737 (Electronic) Linking ISSN: 26624729 NLM ISO Abbreviation: Phys Eng Sci Med Subsets: MEDLINE
Imprint Name(s):
Original Publication: Switzerland : Springer, [2020]-
MeSH Terms:
References:
Trask C (2012) Modeling costs of exposure assessment methods in industrial environments. Work 41(Supplement 1): 6079–6086. https://doi.org/10.3233/WOR-2012-1064-6079. (PMID: 10.3233/WOR-2012-1064-6079)
Grieco A (1998) Epidemiology of musculoskeletal disorders due to biomechanical overload. Ergonomics 41(9):1253–1260. https://doi.org/10.1080/001401398186298. (PMID: 10.1080/0014013981862989754030)
Radwin RG (2002) Biomechanical aspects of work-related musculoskeletal disorders. Theor Issues Ergon Sci Vol 2(2):153–217. https://doi.org/10.1080/14639220110102044. (PMID: 10.1080/14639220110102044)
Marras WS (1995) Biomechanical risk factors for occupationally related low back disorders. Ergonomics 38(2):377–410. https://doi.org/10.1080/00140139508925111. (PMID: 10.1080/001401395089251117895740)
Karhu O (1981) Observing working postures in industry: examples of OWAS application. Appl Ergon 12(1):13–17. https://doi.org/10.1016/0003-6870(81)90088-0. (PMID: 10.1016/0003-6870(81)90088-015676393)
Battevi N (2026) MAPO index for risk assessment of patient manual handling in hospital wards: avalidation study. Ergonomics 49(7):671–687. June https://doi.org/10.1080/00140130600581041. (PMID: 10.1080/00140130600581041)
Hignett S (2000) Rapid entire body assessment (REBA). Appl Ergon 31(2):201–205. https://doi.org/10.1016/S0003-6870(99)00039-3. (PMID: 10.1016/S0003-6870(99)00039-310711982)
Lynn M (1993) RULA: a survey method for the investigation of work-related upper limb disorders. Appl Ergon 24(2):91–99. https://doi.org/10.1016/0003-6870(93)90080-S. (PMID: 10.1016/0003-6870(93)90080-S)
Waters TR (1994) Applications manual for the revised NIOSH lifting equation. Ergonomics 36(7):749–776. (PMID: 10.1080/00140139308967940)
Snyder K, Thomas B, Lu M, Jha R, Barim M, Hayden M, Werren S (2021) A deep learning approach for lower back pain risk prediction during manual lifting. PLoS ONE 16(2):e0247162. (PMID: 10.1371/journal.pone.0247162336067837894914)
Thomas B, Lu M, Rashmi J. Machine learning for detection and risk assessment of lifting action 2022. https://doi.org/10.1109/THMS.2022.3212666.
Chen J, Qiu J, Ahn C (2017) Construction worker’s awkward posture recognition through supervised motion tensor decomposition. Autom Constr 77:67–81. (PMID: 10.1016/j.autcon.2017.01.020)
Alwasel A, Sabet A, Nahangi M, Haas CT, Abdel-Rahman E (2017) Identifying poses of safe and productive masons using machine learning. Autom Constr 84:345–355. (PMID: 10.1016/j.autcon.2017.09.022)
Conforti I, Mileti I, Del Prete Z, Palermo E (2020) Measuring biomechanical risk in lifting load tasks through wearable system and machine-learning approach. Sensors 20:1557. (PMID: 10.3390/s20061557321688447146543)
Selvaraj V, Nagaraj A, Whiffen BG, Min S (2024) Development of a wireless smart sensor system and case study on lifting risk assessment. Manuf Lett 41:229–240. (PMID: 10.1016/j.mfglet.2024.09.027)
Donisi L (2021) Work-related risk assessment according to the revised NIOSH lifting equation: a preliminary study using a wearable inertial sensor and machine learning. Sensors 21(8):2593. https://doi.org/10.3390/s21082593. (PMID: 10.3390/s21082593339172068068056)
Donisi L (2022) A logistic regression model for biomechanical risk classification in lifting tasks. Diagnostics 12(11):2624 https://doi.org/10.3390/diagnostics12112624. (PMID: 10.3390/diagnostics12112624363594689689567)
Morris R (2019) Validity of mobility lab (version 2) for gait assessment in young adults, older adults and parkinson’s disease. Physiol Meas 40(9):8. https://doi.org/10.1088/1361-6579/ab4023. (PMID: 10.1088/1361-6579/ab4023)
Schmitz-Hübsch T (2016) Accuracy and repeatability of two methods of gait analysis–GaitRite™ und mobility lab™–in subjects with cerebellar ataxia. Gait Posture 48:194–201. https://doi.org/10.1016/j.gaitpost.2016.05.014. (PMID: 10.1016/j.gaitpost.2016.05.01427289221)
Donisi L (2021) Benchmarking between two wearable inertial systems for gait analysis based on a different sensor placement using several statistical approaches. Measurement. https://doi.org/10.1016/j.measurement.2020.108642. (PMID: 10.1016/j.measurement.2020.108642)
Lu ML (2016) Evaluation of the impact of the revised National Institute for occupational safety and health lifting equation. Hum Factors 58:667–682. https://doi.org/10.1177/0018720815623894. (PMID: 10.1177/0018720815623894268227954991821)
Spector JT (2014) Automation of workplace lifting hazard assessment for musculoskeletal injury prevention. Ann Occup Environ Med 26(15):1–8.
Fan B, Li Q, Tan T, Kang P, Shull PB (2021) Effects of IMU sensor-to-segment misalignment and orientation error on 3-D knee joint angle estimation. IEEE Sens J 22(3):2543–2552. https://doi.org/10.1109/JSEN.2021.3137305. (PMID: 10.1109/JSEN.2021.3137305)
Jiang C, Yang Y, Mao H, Yang D, Wang W (2022) Effects of dynamic IMU-to-segment misalignment error on 3-DOF knee angle Estimation in walking and running. Sensors 22(22):9009. https://doi.org/10.3390/s22229009. (PMID: 10.3390/s22229009364336089697725)
Gleadhill S, James D, Lee J (2018) February validating temporal motion kinematics from clothing attached inertial sensors. In Proceedings, MDPI, vol 2, no 6, p 304. https://doi.org/10.3390/proceedings2060304.
Jayasinghe U, Harwin WS, Hwang F (2019) Comparing clothing-mounted sensors with wearable sensors for movement analysis and activity classification. Sensors 20(1):82. https://doi.org/10.3390/s20010082. (PMID: 10.3390/s20010082318777806983049)
Schafer RW (2011) What is a Savitzky-Golay filter? IEEE Signal Process Mag 28(4):111–117. https://doi.org/10.1109/MSP.2011.941097. (PMID: 10.1109/MSP.2011.941097)
Barandas M (2020) TSFEL: time series feature extraction library. SoftwareX. https://doi.org/10.1016/j.softx.2020.100456. (PMID: 10.1016/j.softx.2020.100456)
Fan S (2019) A feature selection and classification method for activity recognition based on an inertial sensing unit. Information. https://doi.org/10.3390/info10100290. (PMID: 10.3390/info10100290)
Geethanjali P (2012) Time domain feature extraction and classification of EEG data for brain computer interface. In 9th international conference on fuzzy systems and knowledge discovery, Chongqing, China, pp 1136–1139.
Pan YN (2008) Spectral entropy: a complementary index for rolling element bearing performance degradation assessment. J Mech Eng Sci 223(5), 1223–1231. https://doi.org/10.1243/09544062JMES1224. (PMID: 10.1243/09544062JMES1224)
Mallat S (2008) Time meets frequency. A wavelet tour of signal processing, 2nd edn. Elsevier, San Diego, California, USA.
Zhou ZH (2021) Introduction in machine learning, 1st edn. Springer nature, Singapore.
Zhang Z (2016) Introduction to machine learning: k-nearest neighbors. Ann Transl Med 4(11), 218. https://doi.org/10.21037/atm.2016.03.37. (PMID: 10.21037/atm.2016.03.37273864924916348)
Suthaharan S (2016) Support vector machine. Machine learning models and algorithms for big data classification: thinking with examples for effective learning, 1st edn. Springer, New York, USA, pp 207–235. (PMID: 10.1007/978-1-4899-7641-3_9)
Rish I (2001) An empirical study of naive Bayes classifier. In: IJCAI 2001 workshop on empirical methods in artificial intelligence, Seattle, Washington, USA, pp 41–46.
Rokach L (2010) Decision trees in data mining and knowledge discovery handbook. Springer, New York, NY, USA, pp 165–192.
Popescu MC (2009) Multilayer perceptron and neural networks. WSEAS Trans Circuits Syst 8(7):579–588.
Ververidis D, Kotropoulos C (2005) September sequential forward feature selection with low computational cost. In: 2005 13th European signal processing conference, IEEE, pp 1–4.
Snoek J (2012) Practical bayesian optimization of machine learning algorithms. In: Advances in neural information processing systems, Nevada, USA, p 25.
Singh D (2020) Investigating the impact of data normalization on classification performance. Appl Soft Comput https://doi.org/10.1016/j.asoc.2019.105524. (PMID: 10.1016/j.asoc.2019.105524335193237834168)
Aksoy S (2001) Feature normalization and likelihood-based similarity measures for image retrieval. Pattern Recognit Lett 22(5):563–582. https://doi.org/10.1016/S0167-8655(00)00112-4. (PMID: 10.1016/S0167-8655(00)00112-4)
Sola J (1997) Importance of input data normalization for the application of neural networks to complex industrial problems. IEEE Trans Nucl Sci 44(3):1464–1468. https://doi.org/10.1109/23.589532. (PMID: 10.1109/23.589532)
Hsu CW (2003) A practical guide to support vector classification. pp 1396–1400.
Anguita D (2012) The’K’in K-fold cross validation. In: 20th European symposium on artificial neural networks (ESANN), Bruges, Belgium, pp 441–446.
Pauli MP (2021) Balanced leave-one-subject-out cross-validation for microsleep classification. Curr Dir Biomedical Eng 7(2):147–150. https://doi.org/10.1515/cdbme-2021-2038. (PMID: 10.1515/cdbme-2021-2038)
Rainio O (2024) Evaluation metrics and statistical tests for machine learning. Sci Rep. https://doi.org/10.1038/s41598-024-56706-x. (PMID: 10.1038/s41598-024-56706-x3897776511231319)
Kumar R (2011) Receiver operating characteristic (ROC) curve for medical researchers. Indian Pediatr 48:277–287. https://doi.org/10.1007/s13312-011-0055-4. (PMID: 10.1007/s13312-011-0055-421532099)
Moyen-Sylvestre B (2022) Power spectrum of acceleration and angular velocity signals as indicators of muscle fatigue during upper limb low-load repetitive tasks. Sensors. https://doi.org/10.3390/s22208008. (PMID: 10.3390/s22208008362983579608815)
Kohavi R (1995) A study of cross-validation and bootstrap for accuracy estimation and model selection. In International joint conference on AI (Ijcai), Montreal, Quebec, Canada, pp 1137–1145.
Wong TT (2015) Performance evaluation of classification algorithms by k-fold and leave-one-out cross validation. Pattern Recogn 48(9):2839–2846. https://doi.org/10.1016/j.patcog.2015.03.009. (PMID: 10.1016/j.patcog.2015.03.009)
Matijevich ES, Volgyesi P, Zelik KE (2021) A promising wearable solution for the practical and accurate monitoring of low back loading in manual material handling. Sensors 21(2):340. (PMID: 10.3390/s21020340334191017825414)
Grieco A (1998) Epidemiology of musculoskeletal disorders due to biomechanical overload. Ergonomics 41(9):1253–1260. https://doi.org/10.1080/001401398186298. (PMID: 10.1080/0014013981862989754030)
Radwin RG (2002) Biomechanical aspects of work-related musculoskeletal disorders. Theor Issues Ergon Sci Vol 2(2):153–217. https://doi.org/10.1080/14639220110102044. (PMID: 10.1080/14639220110102044)
Marras WS (1995) Biomechanical risk factors for occupationally related low back disorders. Ergonomics 38(2):377–410. https://doi.org/10.1080/00140139508925111. (PMID: 10.1080/001401395089251117895740)
Karhu O (1981) Observing working postures in industry: examples of OWAS application. Appl Ergon 12(1):13–17. https://doi.org/10.1016/0003-6870(81)90088-0. (PMID: 10.1016/0003-6870(81)90088-015676393)
Battevi N (2026) MAPO index for risk assessment of patient manual handling in hospital wards: avalidation study. Ergonomics 49(7):671–687. June https://doi.org/10.1080/00140130600581041. (PMID: 10.1080/00140130600581041)
Hignett S (2000) Rapid entire body assessment (REBA). Appl Ergon 31(2):201–205. https://doi.org/10.1016/S0003-6870(99)00039-3. (PMID: 10.1016/S0003-6870(99)00039-310711982)
Lynn M (1993) RULA: a survey method for the investigation of work-related upper limb disorders. Appl Ergon 24(2):91–99. https://doi.org/10.1016/0003-6870(93)90080-S. (PMID: 10.1016/0003-6870(93)90080-S)
Waters TR (1994) Applications manual for the revised NIOSH lifting equation. Ergonomics 36(7):749–776. (PMID: 10.1080/00140139308967940)
Snyder K, Thomas B, Lu M, Jha R, Barim M, Hayden M, Werren S (2021) A deep learning approach for lower back pain risk prediction during manual lifting. PLoS ONE 16(2):e0247162. (PMID: 10.1371/journal.pone.0247162336067837894914)
Thomas B, Lu M, Rashmi J. Machine learning for detection and risk assessment of lifting action 2022. https://doi.org/10.1109/THMS.2022.3212666.
Chen J, Qiu J, Ahn C (2017) Construction worker’s awkward posture recognition through supervised motion tensor decomposition. Autom Constr 77:67–81. (PMID: 10.1016/j.autcon.2017.01.020)
Alwasel A, Sabet A, Nahangi M, Haas CT, Abdel-Rahman E (2017) Identifying poses of safe and productive masons using machine learning. Autom Constr 84:345–355. (PMID: 10.1016/j.autcon.2017.09.022)
Conforti I, Mileti I, Del Prete Z, Palermo E (2020) Measuring biomechanical risk in lifting load tasks through wearable system and machine-learning approach. Sensors 20:1557. (PMID: 10.3390/s20061557321688447146543)
Selvaraj V, Nagaraj A, Whiffen BG, Min S (2024) Development of a wireless smart sensor system and case study on lifting risk assessment. Manuf Lett 41:229–240. (PMID: 10.1016/j.mfglet.2024.09.027)
Donisi L (2021) Work-related risk assessment according to the revised NIOSH lifting equation: a preliminary study using a wearable inertial sensor and machine learning. Sensors 21(8):2593. https://doi.org/10.3390/s21082593. (PMID: 10.3390/s21082593339172068068056)
Donisi L (2022) A logistic regression model for biomechanical risk classification in lifting tasks. Diagnostics 12(11):2624 https://doi.org/10.3390/diagnostics12112624. (PMID: 10.3390/diagnostics12112624363594689689567)
Morris R (2019) Validity of mobility lab (version 2) for gait assessment in young adults, older adults and parkinson’s disease. Physiol Meas 40(9):8. https://doi.org/10.1088/1361-6579/ab4023. (PMID: 10.1088/1361-6579/ab4023)
Schmitz-Hübsch T (2016) Accuracy and repeatability of two methods of gait analysis–GaitRite™ und mobility lab™–in subjects with cerebellar ataxia. Gait Posture 48:194–201. https://doi.org/10.1016/j.gaitpost.2016.05.014. (PMID: 10.1016/j.gaitpost.2016.05.01427289221)
Donisi L (2021) Benchmarking between two wearable inertial systems for gait analysis based on a different sensor placement using several statistical approaches. Measurement. https://doi.org/10.1016/j.measurement.2020.108642. (PMID: 10.1016/j.measurement.2020.108642)
Lu ML (2016) Evaluation of the impact of the revised National Institute for occupational safety and health lifting equation. Hum Factors 58:667–682. https://doi.org/10.1177/0018720815623894. (PMID: 10.1177/0018720815623894268227954991821)
Spector JT (2014) Automation of workplace lifting hazard assessment for musculoskeletal injury prevention. Ann Occup Environ Med 26(15):1–8.
Fan B, Li Q, Tan T, Kang P, Shull PB (2021) Effects of IMU sensor-to-segment misalignment and orientation error on 3-D knee joint angle estimation. IEEE Sens J 22(3):2543–2552. https://doi.org/10.1109/JSEN.2021.3137305. (PMID: 10.1109/JSEN.2021.3137305)
Jiang C, Yang Y, Mao H, Yang D, Wang W (2022) Effects of dynamic IMU-to-segment misalignment error on 3-DOF knee angle Estimation in walking and running. Sensors 22(22):9009. https://doi.org/10.3390/s22229009. (PMID: 10.3390/s22229009364336089697725)
Gleadhill S, James D, Lee J (2018) February validating temporal motion kinematics from clothing attached inertial sensors. In Proceedings, MDPI, vol 2, no 6, p 304. https://doi.org/10.3390/proceedings2060304.
Jayasinghe U, Harwin WS, Hwang F (2019) Comparing clothing-mounted sensors with wearable sensors for movement analysis and activity classification. Sensors 20(1):82. https://doi.org/10.3390/s20010082. (PMID: 10.3390/s20010082318777806983049)
Schafer RW (2011) What is a Savitzky-Golay filter? IEEE Signal Process Mag 28(4):111–117. https://doi.org/10.1109/MSP.2011.941097. (PMID: 10.1109/MSP.2011.941097)
Barandas M (2020) TSFEL: time series feature extraction library. SoftwareX. https://doi.org/10.1016/j.softx.2020.100456. (PMID: 10.1016/j.softx.2020.100456)
Fan S (2019) A feature selection and classification method for activity recognition based on an inertial sensing unit. Information. https://doi.org/10.3390/info10100290. (PMID: 10.3390/info10100290)
Geethanjali P (2012) Time domain feature extraction and classification of EEG data for brain computer interface. In 9th international conference on fuzzy systems and knowledge discovery, Chongqing, China, pp 1136–1139.
Pan YN (2008) Spectral entropy: a complementary index for rolling element bearing performance degradation assessment. J Mech Eng Sci 223(5), 1223–1231. https://doi.org/10.1243/09544062JMES1224. (PMID: 10.1243/09544062JMES1224)
Mallat S (2008) Time meets frequency. A wavelet tour of signal processing, 2nd edn. Elsevier, San Diego, California, USA.
Zhou ZH (2021) Introduction in machine learning, 1st edn. Springer nature, Singapore.
Zhang Z (2016) Introduction to machine learning: k-nearest neighbors. Ann Transl Med 4(11), 218. https://doi.org/10.21037/atm.2016.03.37. (PMID: 10.21037/atm.2016.03.37273864924916348)
Suthaharan S (2016) Support vector machine. Machine learning models and algorithms for big data classification: thinking with examples for effective learning, 1st edn. Springer, New York, USA, pp 207–235. (PMID: 10.1007/978-1-4899-7641-3_9)
Rish I (2001) An empirical study of naive Bayes classifier. In: IJCAI 2001 workshop on empirical methods in artificial intelligence, Seattle, Washington, USA, pp 41–46.
Rokach L (2010) Decision trees in data mining and knowledge discovery handbook. Springer, New York, NY, USA, pp 165–192.
Popescu MC (2009) Multilayer perceptron and neural networks. WSEAS Trans Circuits Syst 8(7):579–588.
Ververidis D, Kotropoulos C (2005) September sequential forward feature selection with low computational cost. In: 2005 13th European signal processing conference, IEEE, pp 1–4.
Snoek J (2012) Practical bayesian optimization of machine learning algorithms. In: Advances in neural information processing systems, Nevada, USA, p 25.
Singh D (2020) Investigating the impact of data normalization on classification performance. Appl Soft Comput https://doi.org/10.1016/j.asoc.2019.105524. (PMID: 10.1016/j.asoc.2019.105524335193237834168)
Aksoy S (2001) Feature normalization and likelihood-based similarity measures for image retrieval. Pattern Recognit Lett 22(5):563–582. https://doi.org/10.1016/S0167-8655(00)00112-4. (PMID: 10.1016/S0167-8655(00)00112-4)
Sola J (1997) Importance of input data normalization for the application of neural networks to complex industrial problems. IEEE Trans Nucl Sci 44(3):1464–1468. https://doi.org/10.1109/23.589532. (PMID: 10.1109/23.589532)
Hsu CW (2003) A practical guide to support vector classification. pp 1396–1400.
Anguita D (2012) The’K’in K-fold cross validation. In: 20th European symposium on artificial neural networks (ESANN), Bruges, Belgium, pp 441–446.
Pauli MP (2021) Balanced leave-one-subject-out cross-validation for microsleep classification. Curr Dir Biomedical Eng 7(2):147–150. https://doi.org/10.1515/cdbme-2021-2038. (PMID: 10.1515/cdbme-2021-2038)
Rainio O (2024) Evaluation metrics and statistical tests for machine learning. Sci Rep. https://doi.org/10.1038/s41598-024-56706-x. (PMID: 10.1038/s41598-024-56706-x3897776511231319)
Kumar R (2011) Receiver operating characteristic (ROC) curve for medical researchers. Indian Pediatr 48:277–287. https://doi.org/10.1007/s13312-011-0055-4. (PMID: 10.1007/s13312-011-0055-421532099)
Moyen-Sylvestre B (2022) Power spectrum of acceleration and angular velocity signals as indicators of muscle fatigue during upper limb low-load repetitive tasks. Sensors. https://doi.org/10.3390/s22208008. (PMID: 10.3390/s22208008362983579608815)
Kohavi R (1995) A study of cross-validation and bootstrap for accuracy estimation and model selection. In International joint conference on AI (Ijcai), Montreal, Quebec, Canada, pp 1137–1145.
Wong TT (2015) Performance evaluation of classification algorithms by k-fold and leave-one-out cross validation. Pattern Recogn 48(9):2839–2846. https://doi.org/10.1016/j.patcog.2015.03.009. (PMID: 10.1016/j.patcog.2015.03.009)
Matijevich ES, Volgyesi P, Zelik KE (2021) A promising wearable solution for the practical and accurate monitoring of low back loading in manual material handling. Sensors 21(2):340. (PMID: 10.3390/s21020340334191017825414)
Contributed Indexing:
Keywords: Biomechanical risk assessment; Inertial measurement units; Machine learning; Physical ergonomics; Revised NIOSH lifting equation; Weight lifting
Entry Date(s):
Date Created: 20251009 Date Completed: 20260314 Latest Revision: 20260314
Update Code:
20260314
DOI:
10.1007/s13246-025-01655-6
PMID:
41066018
Database:
MEDLINE
*Further Information*
*Work-related musculoskeletal disorders represent a significant occupational health issue. These disorders encompass a range of conditions resulting from specific risk factors associate to manual material handling such as: intensity, repetition, and duration. Over the years, several observational methodologies have been developed to assess biomechanical risk, but their limits depend mainly on clinicians' subjective assessment. For this reason, wearable sensors coupled with artificial intelligence have recently been integrated in the occupational ergonomic field. This study aimed to develop a new technological methodology-based on machine learning algorithms and inertial wearable sensors-able to automatically discriminate biomechanical risk associated with lifting loads. Ten healthy volunteers were enrolled in this study performing specific weight-lifting tasks wearing two inertial measurement units on the sternum and lumbar region. The acquired inertial signals were appropriately processed to extract several features in the time-domain and frequency-domain which have been used as input to several machine learning algorithms. Excellent results in discriminating biomechanical risk classes were obtained reaching accuracies and areas under the receiver operating characteristic curve above 86% and 95%, respectively. In addition, the sternum emerged as the most informative body landmark, while the mean absolute value was identified as the most informative feature. Future investigations on a larger study population could confirm the potential of the proposed automatic procedure to be used in the workplace in combination with well-established methodologies.
(© 2025. Australasian College of Physical Scientists and Engineers in Medicine.)*
*Declarations. Conflict of interest: The authors have no relevant financial or non-financial interests to disclose. Ethics approval: This study was performed in line with the principles of the Declaration of Helsinki. Consent to participate: Informed consent was obtained from all individual participants included in the study. Consent to publish: The authors affirm that human research participants provided informed consent for publication of the images in Figs. 3 and 4.*