Categorizing Kidney Stones Using Region Properties and Pixel Intensity Matrix

Abstract

Kidney stones or renal calculi are crystals which are formed with in the kidney or in the urinary tract. When there is a decrease in urine volume or if there are more crystalline forming substances in urine, kidney stones are formed. The risk of getting more kidney stones is reduced by finding out the type of kidney stones that helps in identifying the cause for the formation of the stones. In most cases kidney stones larger than 5 mm in size are treated surgically and those lesser than 5 mm in diameter usually pass spontaneously in up to 98% of cases. Kidney stones form in the ureter, bladder, or in urethra. Based on the information obtained from patient history, physical examination, urine analysis, radiographic studies the kidney stones are diagnosed. Kidney stones if small in size are excreted out through urine and if they grow to be larger, they become lodged in the ureter and block the urine flow from that kidney and causes pain. We may need pain medication when there is discomfort. Treatment for the kidney stones is by medication, stone removal by surgery. This paper proposes a novel method for identification of three types of kidney stones namely stag horn, struvite, and calcium type based on Euler number using region properties and contrast which is calculated from pixel intensity matrix. The results obtained show that the Euler number is efficient in identifying calcium stones. Whereas the parameter contrast is calculated using the pixel intensity matrix of the kidney stone images is useful in segregating struvite/stag horn stones. The method proposed show 100% accuracy in the experimental case with limited number of samples; however the accuracy could be confirmed after experimenting the method with huge number of samples.