
In UterUS: Automatic segmentation and alignment of uterine shapes from 3D ultrasound data, we aim to develop computational methods leveraging volumetric ultrasound imaging for standardized analysis of uterine morphology. We are building a growing dataset of annotated 3D transvaginal ultrasound volumes to help us and other researchers explore methods for automatic segmentation and alignment of uterine shapes. We are trying to establish baseline models of the normal uterus. This ongoing effort will support future investigations into shape variations associated with fertility issues and recurrent miscarriages. We believe that the gathered data and developed algorithms can facilitate classification studies and improved diagnostic tools to advance women's reproductive health research and clinical care.
