Toward a method for automatic grading of microscope human embryo images

Abstract

The precise assessment of embryo viability is an extremely important factor for the optimisation of the in vitro fertilisation procedure. In order to assess embryo viability, several embryo scoring systems have been developed. However, they rely mostly on the subjective visual analysis of the embryo morphological features. For instance, an important feature for evaluation of embryos at the day 5 post-fertilisation is the number of cells in the embryo outer layer. In this paper, we present a new method for automation of embryo grading. Based on a polar coordinate version of the input image, we estimated the number of cells in the selected plane of focus using the fractal dimension. A correlation coefficient of 0.81 (n=25) between fractal dimension and the number of cells was found. We also present first segmentation results and highlight challenges that lie ahead.