We present a time-lapse collection and reconstruction technique that allows following embryonic heart development at any computationally halted heart contraction state. The central idea is to image at least one full heartbeat at a fast frame rate, resulting in a two-dimensional plus time (2D+T) data set, and repeat this operation every few minutes over several hours for multiple axial positions. The acquired data are five dimensional (X, Y , and Z in space, `fast' and `slow' dimensions in time). The (2D+T) image series are then synchronized to their neighbors in the axial and development time dimensions using a non-rigid registration algorithm (constrained such as to leave all but the fast time dimension unchanged). The algorithm proceeds recursively over the different axial positions and developmental stages. We successfully applied this procedure to image the development of the embryonic zebrafish heart between 32 and 44 hours post fertilization (hpf).
Michael Liebling, Julien Vermot, Scott E. Fraser