Toilet Sensor for Chronic Kidney Disease

Background

Chronic Kidney Disease (CKD) patients do not experience symptoms until late in the disease's progression. Therefore, many patients do not receive treatment until kidney replacement surgery is necessary. The best way to improve patient outcomes would be to frequently test patients for indicators of CKD so they can start treatment early. 

Aim

Promote early detection of CKD by continuously screening at-risk populations in their homes via automated urinalysis. 

Approach

Create a toilet attachment that tests for a patient’s urinary albumin-creatinine ratio (uACR) once per week and sends data to an app which analyzes the patient’s health status. 

Biomarker

There are several biomarkers that could be used to indicate the presence of CKD. We decided to use uACR, which has several key advantages over other biomarkers. First, it's non-invasive and compatible with urinalysis. Second, it adjusts for differences in urinary concentration due to hydration, meaning it has a very low probability of generating false positives. uACR divides urinary albumin concentration (mg/L) by urinary creatinine concentration (g/L) to obtain a final quotient in mg/g. Two positive uACR tests (>30) spaced 6 months apart indicate a patient has CKD. 

Testing Method

We decided to use dipsticks as our testing mechanism. This is because dipsticks are accurate, cheap, and very easy to visualize. One stick would be dedicated to albumin, and another would be dedicated to creatinine. Each stick would have bands that change color when concentration exceeds a certain threshold. Therefore, the number of bands that change color indicates a potential range of concentration for each reagent. Testing would be performed automatically once per week. 

Toilet Attachment

I designed a CAD model of a toilet attachment that would perform the weekly uACR dipstick tests. The outside contains hooks to attach to the outside of the toilet, covers to protect cups/dipsticks, and contractable holes to release cups/dipsticks for testing. The inside has containers for both the cups and dipsticks. It also has paths to the outside as well as grabbers for mobility. Finally, it also has an image processor to read results and a Raspberry Pi that coordinates grabber movements and sends results to an app. 

The App

I coded an app in Matlab that can interpret the uACR results and monitor patient health over time. First, the app converts the dipstick readings into albumin and creatinine concentrations. It then calculates uACR and graphs results over time. If the patient has two or more positive uACR results spaced 6 months apart, the app will recommend seeing a doctor for further testing for CKD.