CreCENT: Creatinine and Chloride based Electrochemical Non-faradaic renal health mapping Technology

Abstract

In this work, we propose a novel biosensing scheme to enable the stratification of kidney diseases based on severity and progression for timely triage and efficient management. Here, using Creatinine and Chloride as target analytes for the biosensor, we have discussed an arbitrary three-class stratification mapping renal health for Chronic Kidney Disease (CKD) management. Our method is fully quantitative, fast (<5 min turn-around time), and can work with any combination of disease biomarkers to categorize diseases by subtypes and severity. At its core, the biosensor relies on electrochemical impedance spectroscopy to transduce subtle changes at the input Creatinine and Chloride levels in a drop of neat, unprocessed urine. It can operate over a wide dynamic range of 0.15-5 mg/mL for Creatinine and 15-105 mM for Chloride. Further, as proof of concept, the biosensing scheme utilizes a simple Support Vector Machine-based supervised machine learning model for 3-state output disease state classification (corresponding to low, medium, and high disease severity) with a 97.96% accuracy. This scheme is versatile and can be extended to more complex scenarios with more biomarker input stimuli for improved diagnostics and precision therapy for other chronic urological diseases.