Computer Vision Methods for Assessing Ovarian Reserve

The ovarian reserve reflects a woman's ability to successfully realize her reproductive function. The assessment of the ovarian reserve is an actual task for clinical practice. The use of methods of computer processing of diagnostic images, which can speed up and facilitate the performance of routine tasks, increase the objectivity of clinical ultrasound diagnostics, allows for a retrospective analysis. The paper considers the application of machine learning methods for the analysis of ultrasound images. In this paper, we propose the use of modern methods of deep learning in the task of segmentation. Creating special augmentations with the transformation of objects in an ultrasound image. The correct methodology for dividing data into training and validation for use in the medical field. And we also made our dataset publicly available, which contains complex examples from real ultrasound.