Image Quality Improvement in Kidney Stone Detection on Computed Tomography Images
Kidney-Urine-Belly computed tomography (KUB CT) analysis is an imaging modality that has the potential to enhance kidney stone screening and diagnosis. This study explored the development of a semi-automated program that used image processing techniques and geometry principles to define the boundary, and segmentation of the kidney area, and to enhance kidney stone detection. It marked detected kidney stones and provided an output that identifies the size and location of the kidney based on pixel count. The program was tested on standard KUB CT scan slides from 39 patients at Imam Reza Hospital in Iran who were divided into two groups based on the presence and absence of kidney stones in their hospital records. Of these, the program generated six inconsistent results which were attributed to the poor quality of the original CT scans. Results showed that the program has 84.61 per cent accuracy, which suggests the program’s potential in diagnostic efficiency for kidney stone detection.