Publisher's Synopsis
Kidney stone disease (renal calculi) is a very common disease, and its incidence rate increases every year worldwide. The recurrence rate, emergency room visits, and treatment expenses have drastically increased during the past two decades. The kidney stones have to be diagnosed and treated with minimal pain and cost. Medical ultrasound is a non-invasive, radiation-free, and inexpensive imaging modality. However, the problem with ultrasound imaging is the poor image quality due to the presence of speckle noise. The speckle noise is visualized as a bright spot, which increases the sharpness of the ultrasound image. In kidney stone detection applications, the presence of speckle noise affects the detection of kidney stones. The kidney stone is also a small feature that appears as a bright spot with or without shadow, like a speckle. Hence, it is hard to distinguish between the speckle noise and a kidney stone. The increased sharpness due to speckle can lead to over-segmentation and false segmentation. Hence, a robust kidney stone detection algorithm is required for ultrasound kidney image segmentation. As the detection of kidney stones largely depends on image quality, this research focused on an effective reduction of speckle noise and the improved segmentation of kidney stones. The Bilateral Filter (BF) was discovered to provide better image quality than other existing filters while conducting experiments with ultrasound kidney images.
