Chronoamperometry Data Analysis for Non-invasive Glucose Monitoring

The client was building a non-invasive glucose monitoring system. One of the first steps was to calibrate the physical device in order to transform the raw signal measurements to glucose levels. The preliminary work involved analyzing chronoamperometry data in order to find optimal parameters for the device and a way to calibrate the device.

The client provided raw data collected with the device in an internal data format. I developed a raw data parser for the internal format in order to process and visualize data in Python. I created a set of visualizations and analyses in order to help domain experts understand how stable and accurate the device is. Moreover, the visualizations and analyses helped the experts to find optimal parameters for the device, e.g., voltage, sampling.

These preliminary data analyses and the device tuning assured high data quality. The following steps are assumed to use collect more data and to use this data for ML-backed device calibration transforming raw signal to glucose concentration.

Selected resources:

Data Scientist | Machine Learning Engineer | AI Advisor

20 years of experience in data processing from BigData to AI