The knee joint, one of the most complex and heavily loaded joints in the human body, plays a crucial role in daily activities and sports, especially in high-impact exercises like jumping. Among the structures of the knee, the patellar tendon (PT) is particularly important as it transmits forces generated by the quadriceps muscle to the tibia, facilitating knee extension and absorbing impact during dynamic movements (Magnusson et al., 2008). Despite its critical role, the PT is frequently injured due to overuse, particularly in athletes who engage in repetitive jumping and landing activities. The prevalence of PT injuries is notably high, reaching up to 45 % in volleyball players and 32 % in basketball players (Albers et al., 2016). This condition, known as “jumper’s knee,” is closely linked to the magnitude of the load on the PT, which is a key factor in the development of patellar tendinopathy (Richards et al., 1996, Janssen et al., 2013, Lin et al., 2022).

Understanding the biomechanical properties of the PT during landing, including the magnitude and distribution of the loads it bears, is therefore essential for developing effective prevention strategies and therapeutic interventions. Various biomechanical modelling approaches, including musculoskeletal modelling and finite element (FE) analysis have been used in previous studies to investigate knee joint mechanics and PT stress. The FE Method is a critical tool in biomechanical research, enabling precise and cost-effective simulations of complex anatomical structures. It accelerates the research process, delivers quantitative insights, and supports the design and optimization of medical treatments and interventions (Ammarullah et al., 2022b, Jamari et al., 2022a, Ammarullah et al., 2023b).

For instance, Lu et al. (2023) integrated musculoskeletal models with FE analysis to explore the impact of Achilles tendon rupture on knee stress during a counter-movement jump. Yan et al. (2024) further advanced this approach by using musculoskeletal simulations to drive FE models. However, these methods were limited to simple structures and could not simulate more complex joint mechanics. As a result, these earlier studies often relied on simplified models and did not account for changes in the model throughout the movement cycle, making it difficult to capture the complex interactions between structures or the dynamic nature of high-impact activities like landing. Furthermore, many previous studies have used static knee joint models at fixed angles, neglecting the changes in knee joint angles during the landing movement. However, some studies have employed advanced imaging techniques to provide more accurate in vivo assessments of knee biomechanics during dynamic activities. For example, Peng et al. (2023) used the dual fluoroscopic imaging system (DFIS) to visualize. In vivo, knee biomechanics during badminton lunges at different distances and different foot positions. Li et al. (2009) employed DFIS to investigate six degrees of freedom (6DOF) knee kinematics during treadmill gait. By using DFIS to adjust the 3D spatial position of the FE models, it is possible to analyze the load-bearing state of in vivo tissues at multiple time points throughout a movement cycle.

Given these limitations in previous research, our study aims to achieve a more comprehensive and realistic analysis of PT stress during landing. As for the type of stress, this study emphasizes von Mises stress due to its reliability in predicting yielding under multiaxial loading conditions in ductile materials like biological tissues. Unlike Tresca stress, which focuses on shear stress, von Mises stress accounts for distortional energy, making it more suitable for modeling complex stress states in heterogeneous tissues, such as those encountered during landing motions. Additionally, contact stress is more localized, offering limited insight into the overall stress distribution in dynamic motion (Ammarullah et al., 2023a, Hidayat et al., 2024a, Muchammad et al., 2024).

This study integrates musculoskeletal modelling, FE analysis, and DFIS. The musculoskeletal model is used to obtain the loads applied to the FE model, which is then aligned in 3D using dynamic X-ray images to construct individualized knee joint models at different angles during the landing phase. Given the high computational cost of FE modelling, we also aim to identify simple and accessible indicators that can predict the peak equivalent stress in the PT. To achieve this, we employ ridge regression to investigate the relationship between applied loads and the peak equivalent stress in the PT. Through this individualized, continuous analysis of in vivo PT biomechanics across the entire landing phase, we aim to uncover the underlying mechanisms of PT injury, establish a foundation for prevention and treatment strategies, and explore the potential for real-time monitoring of PT health.