Automated measurement of muscle architecture from ultrasound images is important due to the unique ability of ultrasound to image muscle in vivo and in real-time. Many automated algorithms track single muscle fascicles or small regions of interest which can be problematic. This algorithm uses object detection methods from MATLAB to identify most muscle fascicles in an image, producing a more complete measure of muscle architecture.
Updated to allow user to select video file with uigetfile function, automatically determines fascicle orientation, and example video is now included. Video was cropped and every other frame selected to get within upload limits, email for full video.
Code has been updates to include individual fascicle length calculations as opposed to an average that was calculated in previous versions.
Version 2.0 includes bug fixes, filter to reduce frame to frame variance, and a more complete check image.