SYSTR : A Machine Learning based Smart Urine Analysis Tool

The proposed system presents a Computer Aided Diagnosis (CAD) for the early detection of diseases such as diabetes, urinary tract infections, kidney disorders using urine sample by adopting the principle of photoelectric comparison to test the number of biochemical components in the urine. Color changes in accordance with the reaction of the biochemical components in the dipstick by employing Colorimetric technique. This Technique is used to determine the concentration of compounds in fluids by the application of Beer-Lambert’s Law. Each bud in the test strip has its principle of value, on which they react to the components present in the fluid. The parameters of the urine are analyzed by using a tool with test strips affixed to the lid through the colorimetric method or reflectance method. The steps that are performed by the proposed system are image acquisition, grid identification, RGB color feature extraction, data mapping followed by machine learning algorithm which compares the proportions of various parameters. Hence abnormalities in color are determined and disease is predicted at the earliest to provide a point of care (POC) diagnosis