Automated Feature Description of Follicle Size in Assisted Reproductive Treatment

In assisted reproductive treatment, monitoring of follicular size using serial ultrasound is essential to access ovarian response. Traditional method requires doctors to measure the follicle size manually which will lead to inaccurate findings. As for more consistent and reliable parameter of follicular growth, an automated feature description may offer better accuracy in estimating to the response. In this study, by using two-dimensional ultrasound to acquire data from the ovaries, the ultrasound result will indicate the feature description automatically without manual calculation. This automated feature description is developed based on image processing technique using canny edge-detection method in MATLAB. It provides the analysis of the features based on area, perimeter, compactness, major and minor axis and centroid dataset to identify the follicle siz