Kinematic analysis of gait in an underwater treadmill using land-based Vicon T 40s motion capture cameras arranged externally

Abstract

Aquatic therapy for rehabilitation can be performed in a variety of environments, which can vary from a traditional swimming pool to a self-contained underwater treadmill. While kinematic analysis has been performed in large volume swimming pools using specific underwater motion capture systems, researchers may only have access to a land-based motion-capture system, which is not waterproof. Additionally, underwater motion capture systems may not fit within the confines of a smaller underwater treadmill. Thus, the purpose of this study was to design and analyze methodology to quantify lower limb kinematics during an aquatic treadmill session, using a land-based motion capture system. Kinematics of lower limb motion at different speeds was studied while walking on an underwater treadmill in comparison to walking on the same treadmill without water (empty tank). The effects of the presence of water on walking kinematics was analyzed and interpreted using parametric and non-parametric testing procedures. The results suggest significant influences of speed on knee and ankle angles (p < 0.05) in both dryland and aquatic scenarios. Knee and ankle angle measures revealed no significant differences between the dryland and water treadmill scenarios (p > 0.05). The increased time requirement in water for the full gait cycle found in this study indicates influence of resistive effects. This finding can be especially suited for muscle strengthening and stabilizing treatments for lower limbs. Also, a framework was developed to realize a potential methodology to use land-based motion capture cameras to successfully analyze the kinematics of gait in constrained aquatic volumes.

Introduction

Physical activity and locomotion in water has become a preferred medium for exercise routines by clinicians and physical therapists in clinical rehabilitation programs (Becker, 2009, Masumoto et al., 2005, Masumoto et al., 2004, Masumoto and Mercer, 2008). Fluid mechanical properties, such as buoyancy and drag play a decisive role in this choice. The buoyancy of water counteracts the force due to gravity while the drag is the resistive force opposing the movement of the limb through the water. The fluid dynamic drag helps in enhanced control of the movement of the limb (Becker, 2009, Masumoto and Mercer, 2008). These drag and buoyancy forces are affected by water density, specific gravity, hydrostatic pressure, and thermodynamics (Becker, 2009). Water is nearly 800 times denser than air (Llana-Belloch et al., 2013) and drag is proportional to the fluid density. Therefore, drag magnitude in water is orders of magnitude more than in air. Water can also effectively transfer heat energy, making it a natural therapeutic medium for rehabilitation for patients with special exercise considerations (Becker, 2009).

The emergence of aquatic-based approaches for rehabilitation establishes the need for researchers to quantify its effects and maximize the benefits of a structured rehabilitation program. This necessitates analysis and understanding of locomotion in water from two important perspectives: the medical standpoint (psychological, metabolic, cardiorespiratory, neurological, muscular etc.,) and the biomechanical standpoint (Andriacchi and Alexander, 2000, Masumoto et al., 2005). Effects of a water environment for clinical rehabilitation have been previously reported from a medical standpoint: psychological (Oda et al., 1999), cardiorespiratory (Hall et al., 1998, Shono et al., 2001), metabolic (Conners et al., 2019, Conners et al., 2014), muscular (Masumoto et al., 2005) and neurological (Conners et al., 2014). This has helped researchers establish the clinical relevance of aquatic therapy for rehabilitation programs. Aquatic therapy not only helps older adults with medical conditions, it also beneficial for younger adults and athletes recovering from lower limb injuries (Abdul Jabbar et al., 2017). Although the clinical relevance of the biomechanical parameters of human locomotion in water has not yet been established, there is a growing body of literature in this area of research (Lu and Chang, 2012, Masumoto et al., 2005, Masumoto et al., 2004, Masumoto and Mercer, 2008).

The biomechanics of human locomotion can be analyzed from two perspectives: kinematics or kinetics. Kinematics describes the movement of body segments neglecting the masses. The kinematic quantities can be a combination of both linear components (position, linear velocity, and linear acceleration) and angular components (orientation, angular velocity, and angular acceleration). On the other hand, kinetics deals with the measurement of linear forces and angular moments. Motion capture is used to evaluate human locomotion for dryland applications (Boudarham et al., 2013, Kang and Dingwell, 2008, Rendos et al., 2013, Russell Esposito et al., 2017). However, evaluating biomechanics in water is challenging, as most instruments are not built for an aquatic environment (Masumoto & Mercer, 2008).

The research on kinematic data collection underwater has seen significant development since the 1990s (Kwon and Casebolt, 2006, Lauder et al., 1998, Raghu et al., 2019, Silvatti et al., 2013). Also, reliability of human locomotion data from motion capture systems has often been examined for evaluating kinematics both in land and underwater (Gourgoulis et al., 2008a, Gourgoulis et al., 2008b, Kaufman et al., 2016, Kwon and Casebolt, 2006, Miller et al., 2016). Significant research studies have been performed to develop a systematic framework for 3-D motion analysis underwater (Gourgoulis et al., 2008a, Kwon and Casebolt, 2006, Silvatti et al., 2013). However, these studies have often been restricted to large volumed spaces.

In the past decade, state-of-the-art motion capture systems have been specifically developed to be submerged in water for motion capture underwater, such as Oqus cameras (Qualisys, Gothenburg, Sweden) (Abdul Jabbar et al., 2017). However, such systems may be too large to be housed in the smaller capture volumes used in many aquatic applications. Furthermore, use of underwater cameras in smaller capture volumes restricts the viewing angles for accurate motion capture. For example, use of an extensive underwater system in a swimming pool with a large volume might be valid, but such an analysis in self-contained units, like an aquatic treadmill, will be limited due to the restrictions of the smaller volume.

An aquatic environment for rehabilitation can be provided by two means: walking in a swimming pool or in a self-contained unit, such as an underwater treadmill. Walking in swimming pool can be beneficial for complete alleviation of weight in an aquatic rehabilitation plan. However, it is difficult to maintain intensity of exercise while walking or running in a swimming pool (Fujishima & Shimizu, 2003). A self-contained underwater treadmill, however, allows the speed at which a patient walks to be controlled and altered (Conners et al., 2019). Therefore, an aquatic treadmill can be an effective exercise unit as it offers precise control over locomotion parameters and it can be customized according to the requirements of the patient.

A potential exists to explore the use of land-based motion capture cameras arranged externally to capture motion underwater through a view window. The use of land-based motion capture systems for constrained aquatic applications also requires evaluation of the accuracy in reconstruction of markers in a submerged environment. Consequently, the potential of Vicon T40s (Vicon Motion Systems, LA, USA) - a land-based motion capture camera used in this study, and its static accuracy for constrained aquatic environments was shown to be satisfactorily (Raghu et al., 2019). Furthermore, the purpose of this study is to gain insight involving a dynamic task in terms of kinematic differences of the knee and ankle angles while walking in an underwater treadmill with and without water. In particular, the markers placed on the lower limb in a gait motion under water are reconstructed and processed for knee and ankle angles at treadmill speeds between 1 mph (0.45 m/s) and 3 mph (1.34 m/s).

To our knowledge, there are only two studies (Abdul Jabbar et al., 2017, Barela et al., 2006) that have measured the kinematic aspects of gait in water compared to dryland. Abdul Jabbar et al. (2017) used Oqus cameras in a large swimming pool, whereas Barela et al. (2006) used digital cameras and digitization techniques to decipher locomotion in water. Additionally, the evaluation of the accuracy of motion capture systems in water is a developing area of research (Eichelberger et al., 2016, Lauder et al., 1998, Windolf et al., 2008) that often depends on a laboratory context and is application specific. While the accuracy is previously reported (Raghu et al., 2019), the possibility of using a land-based optical motion capture system to reconstruct markers submerged in a constrained aquatic environment during a dynamic task have not been explored, which is a novelty addressed in this study.