| Sumario: | Background: Programs such as Kinovea, Dvideow, Mokka BTK, FreeMoCap, OpenCap, and previously Visual3D Reader have contributed to the analysis of human movement. Dvideow is now deprecated, and Visual3D Reader, once available as a free tool, is no longer an option after being incorporated by HAS-Motion. These solutions often face limitations related to outdated approaches or restricted open-source flexibility. vailá addresses these challenges by offering a solution that integrates dataset management, video editing, signal processing, and machine learning for biomechanical analysis. It allows researchers to work with multimodal data in a unified environment and use independent modules or customized scripts as needed. Objective: To present vailá, an open-source toolbox for biomechanical analysis with a flexible architecture that supports the integration, management, and processing of diverse datasets. vailá is designed to run across multiple operating systems without requiring high computational resources, thus providing an accessible and efficient solution for researchers. Material and Methods: vailá is implemented in Python 3.12.9 and leverages robust libraries for data processing and visualization. Its modular design is organized into distinct components, including file management, data processing, and interactive visualization. The toolbox incorporates algorithms for two-dimensional and three-dimensional pose estimation, with kinematic calculations performed using Euler angles and Quaternions. Kinetic assessments are executed through methods that analyze force profiles and center-of-pressure trajectories. In addition to traditional biomechanical analysis, vailá offers functionalities for dataset handling and video editing. Importantly, vailá was conceived and developed following sound software engineering principles to ensure a timeless architecture that avoids the pitfalls of software deprecation seen in past applications. Its open-source nature enables the easy integration of custom scripts, and it was designed so that users with or without programming knowledge can effectively operate the tool. Results: Preliminary evaluations indicate that vailá processes multimodal biomechanical data efficiently, generating precise visualizations and reliable analytical outputs. The toolbox includes specialized scripts for two-dimensional and three-dimensional reconstruction, image and data distortion correction, and calibration using Direct Linear Transformation and chessboard techniques. It also supports a wide range of analyses including force plate assessments, electromyography, GNSS/GPS, and advanced data visualization. Its flexible, modular design not only reduces processing time compared to conventional methods but also allows for extensive customization, ensuring the tool meets diverse research requirements in both academic and clinical settings. Conclusions: vailá is a robust, flexible open-source solution for integrated biomechanical analysis. Its unified platform promotes transparency and reproducibility in human movement research, while its accessible design and easily available scripts make it user-friendly even for those without programming experience. The toolbox can be readily found on GitHub by searching "vailá Multimodal Toolbox".
|
|---|