The BOB human biomechanics analysis software, combined with the Rokoko Smartsuit Pro motion capture equipment, enables biomechanical analysis of police training techniques.

Abstract:

(1) Background: A set of ergonomic parameters related to risk assessment methods for occupational risk prevention, such as REBA or NIOSH, have been measured using inertial sensors that can capture human movement. These methods are based on the assessment of many postural and dynamic parameters. In the case of police physical intervention techniques, the angles of the trunk, legs, arms, forearms, and wrists, joint contact forces and shear forces at the L5 pelvic junction, asymmetries (angles and factors), and muscle strength are the more relevant parameters to consider.

(2) Methods: Data were collected using a motion capture suit equipped with 19 inertial sensors. A large amount of data and 3D graphics were managed by a powerful software package specifically designed for ergonomic analysis. The physical intervention technique used by the police is OTP.

(3) Results: Five ergonomic parameters in traditional police physical intervention techniques were analyzed. REBA scores and ergonomic indicators were recorded and discussed in relation to some preventive risk thresholds in the literature.

(4) Conclusion: Capturing movements in OTP using inertial sensors provides a new and very effective perspective for occupational risk research.

Keywords: ergonomics, occupational risk prevention, police physical intervention, use of force, operational tactical procedures, motion capture technology

1. Introduction

In recent years, the use of inertial sensor-based methodologies in medical applications has increased. For example, the use of wearable sensor platforms to analyze human movement using inertial measurement units (IMUs) has become a widely used tool for rehabilitation or injury prevention.

There is also a wide range of literature on biomechanical studies of different movements based on IMUs. For instance, the compilation conducted by Van der Kruk et al. is noteworthy as they conducted an in-depth analysis of the measurement accuracy of IMUs for application in fast-moving activities such as sports. However, there is not much focus on physical interventions by police or security forces in these studies, which is the scope of this research.

Among the technical publications we found on police interventions, a study by Mavor based on optical motion and inertial measurement unit sensors (IMUs) examined the impact of load transport captured during military-like movements on military or defense forces. In this study, several parameters such as full-body joint angles were determined through movements like running, walking, kneeling, and lying prone. Based on this data, the IMU system was shown to be suitable for capturing and reconstructing full-body movements and their variability in military activities. However, the movements covered were quite simple, necessitating a more detailed analysis of a broader range of physical intervention techniques, such as upper body movements.

Conducting biomechanical analysis of the most important actions during police physical interventions is very significant for the development of risk prevention research.

This paper aims to lay the foundation for reducing occupational risks associated with police physical intervention techniques. It will particularly focus on an operational tactical procedure (OTP) that police use to control opponents based on methods and techniques taught by police training academies. OTP differs from traditional procedures, which are based on martial arts and combat sports aimed at immobilizing opponents. These types of interventions involve a high risk of injury to both the opponent and the police. In contrast, OTP avoids striking any critical areas and inflicting serious harm on the opponent.

Our research is based on ergonomic assessments for prevention. Therefore, it is essential to accurately determine certain types of biomechanical parameters, such as the frontal or sagittal plane angles of the knees, hips, or spine, for proper assessment. In this regard, IMU-based 3D motion capture tools have proven their level of accuracy in determining these parameters.

The Rokoko 19 IMU suit has proven to be a reliable cross-platform option (iOS, MS Windows, etc.). In addition, the modeling software developed by Shippen has also been used.Human Biomechanics (BoB)建模软件包。

这种硬件和软件组合已被证明是分析人体生物力学人体工程学因素的一种强大设备，因此它们已被用于计算关节的运动范围、轨迹、关节扭矩、肌肉力、地面反作用力和关节接触力。

当前已经开发了许多不同的人体工程学评估方法，如REBA[9,10]、NIOSH[11]、RULA[12]、OWAS[13]、LEST[14]、JSI[15]、NMQ-E[16]、Snook和Ciriello表[17]、OCRA[18]、OCRA检查表[19]和Chaffin生物力学模型[20]。其中，本研究将考虑REBA方法（作为评估姿势负荷的方法）和NIOSH方法（旨在评估体重处理）的主要特征和参数。这两种方法将成为我们分析作战战术程序或OTP的警察物理干预技术（PIT）中的人机工程学的基础（Vera Jiménez等人，2020）[6]。

可以认为，作为人体工程学评估方法，REBA和NIOSH不是评估物理干预技术的最合适方法。然而，尽管它们存在不可否认的缺陷，但它们评估的一些参数对于本研究而言可能非常有用，特别有趣。

西班牙劳动和社会经济部下属的国家工作安全、健康和福利研究所收集了REBA和NIOSH方法作为预防技术注释（NTP是其西班牙语首字母缩写）。这些非关税壁垒是良好做法的指南和建议，除非纳入现行法规中，否则不是强制性的。在考虑特定NTP中所含建议的适用性时，应考虑其发布日期。

1.1.REBA（快速全身评估）方法中的相关参数

REBA方法由Nogareda[10]设计，包含在NTP 601中，旨在评估与人体解剖几何结构相关的一系列参数，如四肢和身体其他部分的角度和相对位置，根据这些参数，根据这些几何参数的特定间隔和限值，分配一系列分数。

通过应用该方法，评估上肢、下肢、躯干和颈部的姿势，同时区分身体的右侧和左侧。因此，身体的每一部分都被分配了一个分数，随着其位置远离其最放松的姿势，分数会更高。REBA方法还考虑了握力的类型或形式，并且握力越强，得分越高。

The maximum total score of REBA is 15 points, which includes five ranges of values. Based on different risk levels, a risk index and corresponding recommendations or interventions are established. The risk index can range from negligible to very high risk, as follows: negligible risk, no action recommended; low risk, some changes should be considered; moderate risk, some changes are needed; high risk, changes should be made as soon as possible; and very high risk, changes need to be implemented immediately.

In the case of police physical intervention techniques, the trunk (Table 1), legs (Table 2), arms (Table 3), forearms (Table 4), and wrist angles (Table 5) are the more relevant parameters, as shown in the figure below.

1.2. Relevant parameters in the NIOSH method

This study will mainly focus on risks associated with low back pain or back problems; it has been considered that the method based on the NIOSH lifting equation [11] will be the most appropriate. Therefore, based on the NIOSH equation, we determined the compression force limit at the L5 pelvic junction, measured in Newtons (N) and asymmetry angle (A).

Asymmetry angle and asymmetry factor

The asymmetry angle (A) is an indicator of the worker's trunk twisting/bending during lifting or loading tasks. Therefore, it represents asymmetric movement relative to the sagittal plane of the body.

According to NIOSH, lifting that requires trunk twisting is penalized, meaning that asymmetric lifting should be avoided. To this end, the asymmetry coefficient (AM) is calculated using the formula: AM=1− (0.0032 ∗ A), where A is the rotation angle (in degrees), as shown in Figures 1 and 2.

Figure 1: Recommended lifting weight percentage according to the asymmetry angle in the NIOSH method

Figure 2: 3D vector of clamping force, representing the compression force at the L5 pelvic junction

When there is no asymmetry, the AM factor is 1, and its value decreases as the angle increases. It should also be noted that when A>135°, the AM value is 0 (Figure 1).

The latter is the worst case, and lifting is not recommended (RWL=0 indicates this).

In the most extreme cases, individuals affected by sagittal imbalance typically exhibit significant changes in gait patterns, joint range of motion, and lower limb strength and muscle work capacity [21].

Sagittal imbalance is significantly associated with reduced activation of all major muscle groups in the lower limbs (gluteus medius, lumbar muscles, hamstrings, etc.), except for the quadriceps.

Additionally, the study by González-Míguez [21] indicates that through three-dimensional motion analysis, the system used in this study can avoid the usual underestimation detected by conventional radiological diagnostics.

1.3. Joint contact forces at the L5 pelvic junction

The most relevant compressive forces to consider are those acting on the spinal intervertebral discs, including body weight and load behavior. According to NIOSH [11] standards, the 3400 N (Figure 2) at the L5 pelvic junction has been identified as the compression force limit for low back pain risk.

1.4. Shear forces at the L5 pelvic junction

Shear forces are those that cause vertebrae to slide or displace parallel to adjacent lower or upper vertebrae. Therefore, the force is perpendicular to the compressive force. When the human trunk exerts pulling and/or pushing forces, greater forces are typically generated (Figure 3).

Figure 3: Shear forces at the L5 pelvic junction represented by a 3D vector

According to a comprehensive biomechanical engineering study on the most common causes of low back pain [22], these forces typically range from 600 to 3200 N, with an average of 1700 N. This paper concludes that in low-frequency tasks, absolute forces up to 1000 N (≤100 loads/day) are acceptable for 90% of the working-age population, while in higher frequency tasks (around 1000 loads/day), an absolute force of 700 N is tolerable.

Clearly, all these dynamic parameters will strongly depend on the weight, height, age, and gender of the officer performing each type of physical intervention technique.

1.5. Total muscle strength

Muscle strength is another dynamic parameter considered of interest in this study. Using the samebiomechanical calculation software Bobto retrieve the corresponding data. The calculation of muscle strength is based on the energy used to perform specific work over a given time, expressed in watts.

Over the years, this parameter has often been analyzed in sports training. It is specifically used in certain sports, such as tennis, baseball, or golf [23, 24, 25, 26]. However, it is also the subject of research in certain jobs where muscle strength is related to a range of typical diseases. In this sense, the National Institute for Scientific Research (L‘Assicurazione contro gli infortuni sul lavoro) [26] conducted a study on EU construction workers who suffer from musculoskeletal disorders (MSD) due to performing tasks that put workers at risk (such as manual handling of heavy objects, forced postures, etc.).

In this study, the most common injuries include ulnar collateral ligament (UCL) tears, flexor-pronator tendonitis or tears, ulnar nerve entrapment, posterior impingement, little head muscle separation chondromalacia, and extensor tendonopathy, with tendinopathy being the most common injury.

One of the main objectives of this study is to outline the injuries of the upper limbs (shoulder, elbow, and wrist) and their prevention based on current biomechanical knowledge, as well as how loads and muscle movements should contribute to a better understanding of the pathophysiology of injuries.

2. Materials and Methods

This is a biomechanical modeling software package that uses data corresponding to position, velocity, and acceleration provided by sensors to determine linear velocity and acceleration, angular velocity and acceleration (from muscle stretching or rotation), and other dynamic measurements such as muscle tension/compression force, energy, and the power they exert. All this data is valuable for the purposes of this study.BoBAlso provides three-dimensional graphics and a user interface where analysis results can be displayed (Figure 4).

Figure 4: Graphics generated by the BoB software package.

The software provides various options for registering values related to REBA analysis. Therefore, for grip quality, there are the following four options: suitable handle and medium power grip (most suitable for OTP); acceptable but not ideal handle; unacceptable, but may have armrests; no handle, awkward or unsafe. It also offers four options for "Activity Level?" as follows: one or more body parts remain still for more than 1 minute; small range movements repeated more than 4 times per minute; actions that lead to rapid changes in posture (most suitable for OTP); none of the above.

The device consists of a set of 19 active (three-axis accelerometers, gyroscopes, and geomagnetic) wireless inertial sensors that can determine the position, velocity, acceleration, and magnetic field within the human body. It also has a specific software application that displays body positions as avatars on the screen (Figure 5). The sensors can record data up to 100 times per second.

Figure 5: (a) Avatar of human posture during the execution of intervention techniques. (b) Arrangement of wireless sensors on the human body.

2.3. OTP Police Physical Intervention Techniques

Police physical intervention techniques are based on Operational Tactical Procedures (OTP), which include a series of Physical Intervention Techniques (PIT) that use defensive mechanisms such as blocking, redirecting, and grabbing upper and lower limbs to avoid striking vulnerable and critical areas, thereby minimizing the risk of serious harm.

In fact, the techniques described in the OTP program rely on applying controllable pain, for example, striking the triceps tendon or quadriceps may cause pain sufficient to completely block and control the opponent. These procedures not only avoid harm to the opponent but also to the police.

2.4. Research Collaborators

Measurements of all the aforementioned biomechanical parameters were recorded when two professional male officers wearing sensor suits executed the entire process of classic physical intervention techniques.

3. Results and Discussion

Due to the biomechanical analysis conducted using this new capture technology based on wireless sensors, a large number of biomechanical parameters were identified. This ergonomics study based on human factors should allow us to determine the probability of injury.

The software application allows the use of avatars corresponding to the body position at specific moments in the OTP. Figure 6 shows four moments of a female officer facing an opponent during OTP. The officer is applying restraint techniques to a civilian.

Figure 6: Snapshots of the four steps in OTP implementation: (a) 0.0 seconds, (b) 1.5 seconds, (c) 2.6 seconds, and (d) 3.6 seconds. Left: Anatomical avatar corresponding to the officer's posture in the right-hand image.

Figure 6a shows the initial moment of the process, which is a safe position against the attacker. The left image shows the anatomical avatar corresponding to the officer's posture.

The image in Figure 6b shows the officer moving forward, with the leg closest to her moving position while grabbing the opponent's elbow with her forearm (from the same side).

In the image of Figure 6c, the opponent is being dragged to the officer's hip, which is a step before he is restrained. Finally, Figure 6d shows how the opponent is restrained in a vertical position, where he should be completely under control.

Throughout the OTP implementation process, the sensors record biomechanical parameters, which are then converted into a set of useful scores used by REBA or NIOSH to assess the situation, thereby preventing any possible occupational risks.

3.1 Assessment of REBA

Figure 7 shows the evolution of the REBA scores in the OTP implemented in this study. REBA values are registered 100 times per second, with the four snapshots in Figure 6 marked as (a) to (d). It can be seen from the figure that the OTP score selected in this study did not reach the "very high risk" level, while its average score falls within the "medium risk" range. This is an unusual score that corresponds only to this specific case.

Figure 7: REBA scores of OTP. The moments corresponding to the four snapshots in Figure 6 are shown as (a–d).

3.2. Asymmetry

Figures 8, 9, 10, and 11 show the blue curves corresponding to asymmetry angles and coefficients, joint contact forces, shear forces at the L5 pelvic junction, and muscle strength, respectively.

Figure 8: Asymmetry angle values throughout the OTP. The moments corresponding to the four snapshots in Figure 6 are shown as (a–d).

Figure 9: Joint contact force values at the L5 pelvis during the entire OTP. The moments corresponding to the four snapshots in Figure 6 are shown as (a–d).

Figure 10: Lower force values at the L5 pelvic joint during the entire OTP. The moments corresponding to the four snapshots in Figure 6 are shown as (a–d).

Figure 11: Total muscle strength throughout the OTP. The moments corresponding to the four snapshots in Figure 6 are shown as (a–d).

The range of variation of the asymmetry angle (Figure 8) is from −35° to 25°, with twisting starting from the first second of the intervention, and the asymmetry coefficient decreases by no more than 15% of the lifted weight.

3.3. Joint Contact Force

As shown in Figure 9, the joint contact force or compressive force at the L5 pelvic junction changes very rapidly. Its maximum value (2146 N) is reached when the police force the opponent to bend over (Figure 9c).

One of the parameters of particular interest in this article is the shear force on the segment between the L5 vertebra and the pelvis, as previous studies have reported a significant difference between experts and novices, which is noteworthy from a risk prevention perspective.

When the torso bends forward, greater forces are typically generated due to gravity acting on the upper body, but in some jobs involving pushing or pulling objects, this force can also be very significant. Shear force is typically defined as the force acting parallel to the plane determined by the intervertebral disc within the specific segment of interest.

3.4. Shear Force

当躯干向前弯曲时，由于重力作用于上身，通常会产生更大的力，但在一些涉及推或拉物体的工作中，这种力也会非常重要[27]。剪切力通常被定义为平行于特定感兴趣节段内椎间盘确定的平面作用的力[28]。

3.4.剪切力

From Figure 10, it can be seen that during the entire transition process, the shear force at the L5 pelvic joint is moderate (not very high) until the officer takes a step forward. Then, when the police lean forward towards the opponent, it increases from Figure 10b to Figure 10c. In this example, the maximum shear force (418 N) occurs at 1.6 seconds.

3.5. Muscle Strength

From the perspective of occupational risk, another very important parameter is the muscle strength exerted by the police. Figure 11 represents the total muscle strength with positive and negative values, where positive values are assigned to eccentric movements (i.e., muscle strength is applied in the opposite direction of muscle movement). Conversely, concentric movements are associated with negative values. If considering the absolute value of power (the rate of energy consumption per unit time), the highest muscle power consumption occurs between the steps of Figure 10c and Figure 11b.

4. Conclusion

Inertial Measurement Unit (IMU) sensors are used to capture motion, providing a new and efficient method for OTP occupational risk research.

Through this study, the IMU-based 3D motion capture tool has proven to be suitable for measuring the biomechanical parameters used in the REBA and NISOH assessment methods, such as the trunk's frontal and sagittal angles; additionally, other dynamic parameters have been identified, such as joint contact forces and shear forces at the L5 pelvic connection, asymmetries (angles and coefficients), and muscle strength.

IMU technology is also easy to use outside of laboratory environments.

In summary, we conclude that these are suitable tools for determining the risk levels of other physical intervention techniques.

Abbreviation Glossary:

BoB Biomechanics of Body
IMU Inertial Measurement Units
JSI Job Strain Index
LEST Laboratoire d’Economie et Sociologie du Travail (Labor Economics and Sociology Laboratory)
Sociology Laboratory)
NIOSH National Institute for Occupational Safety and Health
NMQ-E Nordic Musculoskeletal Questionnaire (Extended)
OCRA Occupational Repetitive Action
OTP operational tactical procedure
OWAS Ovako Working Analysis System
PIT Physical Intervention Techniques
REBA Rapid Entire Body Assessment
RULA Rapid Upper Limb Disorders
References:
1. Laidig D., Schauer T., Seel T. Exploiting kinematic constraints to compensate magnetic disturbances when calculating joint angles of approximate hinge joints from orientation estimates of inertial sensors; Proceedings of the International Conference on Rehabilitation Robotics (ICORR); London, UK. 17–20 July 2017; pp. 971–976. [PubMed] [CrossRef] [Google Scholar]

2. Kok M., Hol J.D., Schön T.B. Using inertial sensors for position and orientation estimation. Found. Trends Signal Process. 2017;11:1–153. doi: 10.1561/2000000094. [CrossRef] [Google Scholar]

3. Weygers I., Kok M., De Vroey H., Verbeerst T., Versteyhe M., Hallez H., Claeys K. Drift-free inertial sensor-based joint kinematics for long-term arbitrary movements. IEEE Sens. J. 2020;20:7969–7979. doi: 10.1109/JSEN.2020.2982459. [CrossRef] [Google Scholar]

4. Van der Kruk E., Reijne M.M. Accuracy of human motion capture systems for sport applications; state-of-the-art review. Eur. J. Sport Sci. 2018;18:806–819. doi: 10.1080/17461391.2018.1463397. [PubMed] [CrossRef] [Google Scholar]

5. Mavor M.P., Ross G.B., Clouthier A.L., Karakolis T., Graham R.B. Validation of an IMU suit for military-based tasks. Sensors. 2020;20:4280. doi: 10.3390/s20154280. [PMC free article] [PubMed] [CrossRef] [Google Scholar]

6. Vera-Jiménez J.C., Lorente J.A., González-Herrera L., Álvarez J.A., Ferreiro-González M., Ayuso J. A legal and forensic medicine approach to police physical intervention techniques in high-risk situations. Int. J. Environ. Res. Public Health. 2020;17:2809. doi: 10.3390/ijerph17082809. [PMC free article] [PubMed] [CrossRef] [Google Scholar]

7. Tak I., Wiertz W.-P., Barendrecht M., Langhout R. Validity of a new 3-D Motion analysis tool for the assessment of knee, hip and spine joint angles during the single leg squat. Sensors. 2020;20:4539. doi: 10.3390/s20164539. [PMC free article] [PubMed] [CrossRef] [Google Scholar]

8. Shippen J., May B. BoB—Biomechanics in MATLAB; Proceedings of the 11th International Conference Biomdlore; Druskininkai, Lithuania. 20–23 October 2016; [CrossRef] [Google Scholar]

9. Hita-Gutiérrez M., Gómez-Galán M., Díaz-Pérez M., Callejón-Ferre J.A. An overview of REBA method applications in the world. Int. J. Environ. Res. Public Health. 2020;17:2635. doi: 10.3390/ijerph17082635. [PMC free article] [PubMed] [CrossRef] [Google Scholar]

10. Cuixart S.N. NTP 601: Evaluación de las Condiciones de Trabajo: Carga Postural. Método REBA (Rapid Entire Body Assessment). Instituto Nacional de Seguridad e Higiene en el Trabajo (INSHT) 2001. [(accessed on 29 March 2022)]. Available online: https://www.studocu.com/ca-es/document/universitat-oberta-de-catalunya/prevencion-de-riesgos-derivados-de-la-organizacion-y-la-carga-de-trabajo/ntp-601-evaluacion-de-las-condiciones-de-trabajo-carga-postural-metodo-reba-rapid-entire-body-assessment/20730183

11. Work Practices Guide for Manual Lifting. National Institute for Occupational Safety and Health; Cincinnati, OH, USA: 2017. NIOSH Technical Report 1981, n° 81–122. [Google Scholar]

12. Lim C.-M., Jung M.-C., Kong Y.-K. Evaluation of upper-limb body postures based on the effects of back and shoulder flexion angles on subjective discomfort ratings, heart rates and muscle activities. Ergonomics. 2011;54:849–857. doi: 10.1080/00140139.2011.600777. [PubMed] [CrossRef] [Google Scholar]

13. Karhu O., Härkönen R., Sorvali P., Vepsäläinen P. Observing working postures in industry: Examples of OWAS application. Appl. Ergon. 1981;12:13–17. doi: 10.1016/0003-6870(81)90088-0. [PubMed] [CrossRef] [Google Scholar]

14. Guélaud F., Roustang G., Beauchessne M., Gautrat J. For an Analysis of Workers' Working Conditions in the Company. LEST Method. Laboratoire d’Économie et de Sociologie de travail; Aix en Provence, France: 1975. [Google Scholar]

15. Restuputri D.R., Masudin I., Putri A.R.C. The comparison of ergonomic risk assessment results using job strain index and OCRA methods. IOP Conf. Ser. Mater. Sci. Eng. 2020;821:012029. doi: 10.1088/1757-899X/821/1/012029. [CrossRef] [Google Scholar]

16. Dawson A.P., Steele E.J., WHodges P.W.P., Stewart S. Development and Test–Retest Reliability of an Extended Version of the Nordic Musculoskeletal Questionnaire (NMQ-E): A screening instrument for musculoskeletal pain. J. Pain. 2009;10:517–526. doi: 10.1016/j.jpain.2008.11.008. [PubMed] [CrossRef] [Google Scholar]

17. Snook S.H., Ciriello V.M. The design of manual handling tasks: Revised tables of maximum acceptable weights and forces. Ergonomics. 1991;34:1197–1213. doi: 10.1080/00140139108964855. [PubMed] [CrossRef] [Google Scholar]

18. Occhipinti E. OCRA: A concise index for the assessment of exposure to repetitive movements of the upper limbs. Ergonomics. 1998;41:1290–1311. doi: 10.1080/001401398186315. [PubMed] [CrossRef] [Google Scholar]

19. Rhén I.M., Forsman M. Inter-and intra-rater reliability of the OCRA checklist method in video-recorded manual work tasks. Appl. Ergon. 2020;84:103025. doi: 10.1016/j.apergo.2019.103025. [PubMed] [CrossRef] [Google Scholar]

20. Bean J.C., Chaffin D.B., Schultz A.B. Biomechanical model calculation of muscle contraction forces: A double linear programming method. J. Biomech. 1988;21:59–66. doi: 10.1016/0021-9290(88)90192-3. [PubMed] [CrossRef] [Google Scholar]

21. González-Míguez P. Ph.D. Thesis. Universitat Autònoma de Barcelona; Barcelona, Spain: 2017. Biomechanical Analysis of Gait by Means of an Analytical Model of Inverse Dynamics in Patients with Alteration of the Sagittal Balance of the Spine. [Google Scholar]

22. Gallagher S., Marras W.S. Tolerance of the lumbar spine to shear; A review and recommended exposure limits. Clin. Biomech. 2012;27:973–978. doi: 10.1016/j.clinbiomech.2012.08.009. [PubMed] [CrossRef] [Google Scholar]

23. Davidson P.A., Pink M., Perry J., Jobe F.W. Functional anatomy of flexor pronator muscle group in relation to the MCL of the elbow. Am. J. Sports Med. 1995;23:245–250. doi: 10.1177/036354659502300220. [PubMed] [CrossRef] [Google Scholar]

24. Pluim B.M., Staal J.B., Windler G.E., Jayanthi N. Tennis injuries: Occurrence, aetiology, and prevention. Br. J. Sports Med. 2006;40:415–423. doi: 10.1136/bjsm.2005.023184. [PMC free article] [PubMed] [CrossRef] [Google Scholar]

25. Eygendaal D., Rahussen F.T.G., Diercks R.L. Biomechanics of the elbow joint in tennis players and relation to pathology. Br. J. Sports Med. 2017;4:820–823. doi: 10.1136/bjsm.2007.038307. [PMC free article] [PubMed] [CrossRef] [Google Scholar]

26. Campo G., Cegolon L., De Merich D., Fedeli U., Pellicci M., Heymann W.C., Pavanello S., Guglielmi A., Mastrangelo G. The Italian national surveillance system for occupational injuries: Conceptual framework and fatal outcomes, 2002–2016. Int. J. Environ. Res. Public Health. 2020;17:7631. doi: 10.3390/ijerph17207631. [PMC free article] [PubMed] [CrossRef] [Google Scholar]

27. Knapik G.G., Marras W.S. Spine loading at different lumbar levels during pushing and pulling. Ergonomics. 2009;52:60–70. doi: 10.1080/00140130802480828. [PubMed] [CrossRef] [Google Scholar]

28. Adams M.A., Bogduk N., Burton K., Dolan P. The Biomechanics of Back Pain. 2nd ed. Churchill Livingstone; Edinburgh, UK: 2006. [Google Scholar]