Joint Torque Prediction via Hybrid Neuromusculoskeletal Modelling during Gait Using Statistical Ground Reaction Estimates: An Exploratory Study

Joint torques of lower extremity are important clinical indicators of gait capability. This parameter can be quantified via hybrid neuromusculoskeletal modelling that combines electromyography-driven modelling and static optimisation. The simulations rely on kinematics and external force measurements, for example, ground reaction forces (GRF) and the corresponding centres of pressure (COP), which are conventionally acquired using force plates. This bulky equipment, however, hinders gait analysis in real-world environments. While this portability issue could potentially be solved by estimating the parameters through machine learning, the effect of the estimation errors on joint torque prediction with biomechanical models remains to be investigated. This study first estimated GRF and COP through feedforward artificial neural networks, and then leveraged them to predict lower-limb sagittal joint torques via (i) inverse dynamics and (ii) hybrid modelling. The approach was evaluated on five healthy subjects, individually. The predicted torques were validated with the measured torques, showing that hip was the most sensitive whereas ankle was the most resistive to the GRF/COP estimates for both models, with average metrics values being 0.70 < R2 < 0.97 and 0.069 < RMSE < 0.15 (Nm/kg). This study demonstrated the feasibility of torque prediction based on personalised (neuro)musculoskeletal modelling using statistical ground reaction estimates, thus providing insights into potential real-world mobile joint torque quantification.

A review of musculoskeletal modelling of human locomotion

Locomotion through the environment is important because movement provides access to key resources, including food, shelter and mates. Central to many locomotion-focused questions is the need to understand internal forces, particularly muscle forces and joint reactions. Musculoskeletal modelling, which typically harnesses the power of inverse dynamics, unites experimental data that are collected on living subjects with virtual models of their morphology. The inputs required for producing good musculoskeletal models include body geometry, muscle parameters, motion variables and ground reaction forces. This methodological approach is critically informed by both biological anthropology, with its focus on variation in human form and function, and mechanical engineering, with a focus on the application of Newtonian mechanics to current problems. Here, we demonstrate the application of a musculoskeletal modelling approach to human walking using the data of a single male subject. Furthermore, we discuss the decisions required to build the model, including how to customize the musculoskeletal model, and suggest cautions that both biological anthropologists and engineers who are interested in this topic should consider.

Muscle Length of the Hamstrings Using Ultrasonography Versus Musculoskeletal Modelling

Muscle morphology is an important contributor to hamstring muscle injury and malfunction. The aim of this study was to examine if hamstring muscle-tendon lengths differ between various measurement methods as well as if passive length changes differ between individual hamstrings. The lengths of biceps femoris long head (BFlh), semimembranosus (SM), and semitendinosus (ST) of 12 healthy males were determined using three methods: Firstly, by identifying the muscle attachments using ultrasound (US) and then measuring the distance on the skin using a flexible ultrasound tape (TAPE-US). Secondly, by scanning each muscle using extended-field-of view US (EFOV-US) and, thirdly, by estimating length using modelling equations (MODEL). Measurements were performed with the participant relaxed at six combinations of hip (0°, 90°) and knee (0°, 45°, and 90°) flexion angles. The MODEL method showed greater BFlh and SM lengths as well as changes in length than US methods. EFOV-US showed greater ST and SM lengths than TAPE-US (p < 0.05). SM length change across all joint positions was greater than BFlh and ST (p < 0.05). Hamstring length predicted using regression equations is greater compared with those measured using US-based methods. The EFOV-US method yielded greater ST and SM length than the TAPE-US method. SM showed the highest change in length at different hip and knee joint positions.