A Method for Markerless Tracking of the Strain Distribution of Actively Contracting Cardiac Muscle Preparations

Abstract

Current techniques for tracking the displacement and strain patterns within contracting cardiac specimens often require ‘invasive’ markers to be fixed on the sample, offer only modest spatial resolution, and are computationally inefficient. We present a new technique for determining, without requiring the addition of markers, the dynamic displacement and strain distribution within isolated cardiac muscle samples. We use an accurate and robust new digital image correlation algorithm to measure deformations from a stitched-together time-series of images of a contracting trabecula, and a video recording of a contracting myocyte, captured by an optical transmission microscope. This technique yielded a 2D map of local deformation and strain throughout each specimen, over the time-course of a twitch, at a spatial resolution far exceeding that of previous techniques. In the trabecula we examined, our method enabled us to quantify its non-uniform strain and sarcomere length dynamics even when it was contracting with its length fixed. We noted the non-uniform relaxation of the specimen, where part of the trabecula appeared to be in contraction while the rest was in extension. The images of the myocyte revealed a strain distribution that ranged from −7% to −18%. By applying this method, it is now possible to accurately quantify internal deformation in isolated muscle samples to subpixel resolution, which is essential for interpreting any motion-dependent measurements such as sarcomere length and change in cross-sectional area, all of which are preparation-specific.