Patients with Chronic Kidney Disease(CKD) need to periodically monitor their creatinine concentration levels for kidney health assessment. Current methods of conducting creatinine assessment require clinical settings, high reagent volumes, automated instruments and are time consuming. There is a need for testing tools that enable CKD-patients to monitor their serum creatinine levels at home. We propose to develop a novel inexpensive paper microfluidic device for colorimetric assay of serum creatinine. Combined with a smartphone camera and machine learning algorithms, our microfluidic device can be used as a convenient and accurate monitoring platform. This project aims to improve upon an initial design developed in our research group  by improving the transport and separation of blood constituents, to increase the quality of the images collected for machine learning-based analysis. In our project, we simulated the performance of our microfluidic device using COMSOL. The goal of the improved device is to accurately measure serum creatinine levels while reducing the time required to measure eGFR: an indicator of CKD severity. This could be attained by designing, modeling, and optimizing the flow of a Lateral Flow Assay on the aforementioned paper microfluidic device.