Automated quantitative assessment of perifollicular vascularization

Automated quantitative assessment of perifollicular vascularisation is important in the research if the application of hormonal therapy after follicle puncture in natural cycles is really always needed. Our algorithm works with power Doppler ultrasound images. At the first step the position and shape of the dominant follicle is determined with the procedure based on the continuous wavelet transform. Then vessels around the follicle are categorized according to their diameter and the results are statistically evaluated by the histograms of vessel diameters. To improve the visual results the vessel reconstruction, based on minimal spanning trees, is done at the end. Key-Words: 3D ultrasound image segmentation, vessel thickness assessment, vessel reconstruction