Kidney Stone Detection using Image Segmentation and Radial Transform

Abstract

While identification and detection of stone in kidney, it is imperative to use ultrasound image diagnostic methodology. However, manual analysis of ultrasound image is a tiring and cumbersome process, therefore, an automated method using radial transform is proposed that includes minimum and maximum radii in each quadrant, then area and perimeter in each quadrant gives the good estimate of the stone size. While the center of mass gives the location of the stone in image. All the measurements are taken in pixel units and can be converted to real scale by taking proper scaling factor.