Relationships of phototextural characteristics of ovine presumptive zygotes to their developmental potential in vitro

In vitro embryo production (IVP) has widespread implications for animal husbandry and conservation programs. The overarching goal of achieving pregnancy with IVP systems calls for accurate methods of embryo viability screening. The present study set out to examine whether phototextural attributes of ovine presumptive zygotes were reliable markers of embryo developmental competence in vitro. Sheep oocytes were obtained post-mortem from nine cycling Polish Longwool ewes using the ovarian cutting and sieving method. Following in vitro maturation of oocytes and their fertilization with fresh ram semen, the development of embryos (n = 37) was monitored using time-lapse imaging. The embryos were retrospectively classified as non-arresting (attaining the blastocyst stage; n = 6) or arresting (non-dividing presumptive zygotes or embryos arresting before the 7th mitotic division; n = 31). Commercially available image analytical software ImageProPlus® was used to generate bitmaps of the regions of interest comprising zygotic cytoplasm. Subsequently, a proprietary computer program r-Algo 2.0 was used to determine if there existed clusters of pixels for which first-order phototextural characteristics of presumptive zygotes' (t0) cytoplasm (i.e., mean pixel intensity (MPI), heterogeneity (MPH), and concentration (MPC)) differed significantly between the two subsets of embryos. Within the algorithmically identified pixel intensity ranges, future non-arresting embryos at t0 exceeded their arresting counterparts in MPI (pixel range of 73-75; 74.06 ± 0.01 vs. 73.98 ± 0.008; P = 0.0007) and MPC (pixel range: 105-106; 1.25 ± 0.22% vs. 0.56 ± 0.07%; P = 0.0004) but had lower mean MPH (pixel range: 36-42; 1.81 ± 0.04 vs. 1.97 ± 0.02; P = 0.0005). Computerized segmentation of ovine zygotic microphotographs is a promising non-invasive method to predict their developmental potential in vitro.