Software Development for Neuroscience, Biology, and Biomechanics Applications

Abstract

Understanding locomotion is an important focus of modern science. Our health and well-being are directly linked to our movement. Animal, including human, movement can explain many biological phenomena. Also, it impacts our ability to treat musculoskeletal injuries and neurological disorders, improve prosthetic limbs, and construct agile legged robots. Two fundamental methods used in locomotion research, especially in the field of neuroscience, are 1) quantification of kinematics from videography, and 2) the creation of stable internal neural interfaces using metal electrodes. With the recent explosion of computer vision algorithms for gathering meaning from video, robotic tools for physical interaction, and a bevy of new genetic tools with which to manipulate the nervous system in intact, freely behaving rodents, there is a need for software that applies these advancements to movement science problems. These tools are especially important now as perturbation based research, where internal or external perturbations are applied to a moving animal in order to better dissect the mechanisms of movement, become more common. To address the need in the former area, we present Python-based software to segment and track landmarks from multiple views, high-speed video using color and 3D information, producing and analyzing kinematics in 3D. This software produces kinematics from raw multiple camera video, and furthermore can perform joint angle analyses in 2D or 3D, a standard technique in locomotor biomechanics. The software has been evaluated using 20 animals and under different conditions (e.g., intact, spinal cord injured, and aged animals). To address the need in the area of neural interfacing, we present open source Matlab software to acquire, characterize, and model the impedance spectra of metal electrodes in solution. Requiring only Matlab and standard data acquisition hardware, the software measures the magnitude and phase of the interface, and fits the most commonly used Randles model. The software was evaluated using five custom-made nerve cuffs. The Randles model parameters, including the constant phase element, were calculated and are in good agreement with the literature. Together, these tools will aid researchers in movement science and related fields.