Boundary detection of kidney ultrasound image based on vector graphic approach

Abstract

This paper presents a new approach for boundary detection of kidneyfrom three-dimensional ultrasound images.The technique proposed here is based on vector graphic image formation. Before converting the ultrasound image intovector graphic image, the region of interest (ROI) of the kidney for each slice was generated automatically. Some imagesalso needed to be rotated to zero degree depending on the position of thekidney in the images. After the vector graphicformation, the boundary points of the kidney were identified. The error points were removed and the interpolation was thenperformed for contouring the kidney from its background. Experiments had been carried out step by step for validationpurposes. Test result based on 30 kidney ultrasound image slices showed that thedeveloped algorithms were able to detect86.67% true ROIs. When compared to manual contouring, the sensitivity of this boundary detection technique was inbetween 94.95% to 97.75% and the specificity was in between 99.26% to99.92%. Based on the results, it can beconcluded that this new semi-automatic technique is reliable to be used for contouring the kidney from three-dimensionalultrasound images.