Awardees of summer 2022 chronic disease pilot research grants report their progress

Wearable Sweat Sensors for Smartphone-Enabled Diabetes Monitoring

Investigators
Mohammad H. Malakooti, Department of Mechanical Engineering
Miqin Zhang, Department of Materials Science & Engineering

Project update
We are developing a cost-effective colorimetric biosensor that can be integrated with smartphone data collection and analysis. Over the past five months, we have synthesized and tested various textile-based colorimetric sensors under different conditions. We were able to successfully produce the first generation of pH sensors in our lab and demonstrate a clear relationship between the sensor’s color and the pH of the solution applied to it.

Our next step was to create machine learning (ML) algorithms for analyzing the captured images and determining the pH. However, we discovered that increasing the color contrast of the sensors would significantly reduce the error and improve the accuracy of the analysis. In a series of experiments, we studied the role of chemicals and processing parameters on the sensitivity and stability of the pH sensors. As a result, the second generation of our pH sensors show a much higher sensitivity (i.e., significant color change) when applying artificial sweat with different pH levels. This led to about 20% higher accuracy of the developed machine learning model when predicting the pH of sweat based on the captured photos using a smartphone. Using this dataset, we will employ other ML algorithms to obtain important parameters that influence color measurement and identify the pH of the samples.

In the next phase of the project, we will synthesize glucose sensors, increase their stability and sensitivity, and employ machine learning and artificial intelligence for image analysis and prediction of glucose level based on sensor’s color change.

Evaluating a novel, portable, self-administered device (“Beacon”) that measures critical flicker frequency toward at-home testing for minimal hepatic encephalopathy in cirrhosis

Investigators
George N. Ioannou, Department of Medicine
James Fogarty, Paul G. Allen School of Computer Science & Engineering
Sean A. Munson, Department of Human Centered Design & Engineering
Philip Vutien, Department of Medicine, Veterans Affairs Puget Sound Healthcare System
Ravi Karkar, Paul G. Allen School of Computer Science & Engineering
Richard Li, Paul G. Allen School of Computer Science & Engineering

Project update
Critical flicker frequency is one of the best screening tests for minimal hepatic encephalopathy, but measuring critical flicker frequency currently requires specialized clinical equipment that is bulky, expensive and not designed for self-measurement or at-home measurement. The overall goals of the Beacon project are to develop a portable and inexpensive device that can support at-home self-measurement of and longitudinal monitoring of critical flicker frequency in patients with chronic disease. We are focused on developing Beacon in the context of patients with chronic liver disease, building upon initial in-clinic testing with 153 patients that found Beacon measurements are nearly identical to measurements from an existing clinical device.

With pilot award support, we have now completed a total of 11 at-home deployments with patients with chronic liver disease. Each six-week patient deployment includes daily measurements over two weeks and then weekly measurements over four additional weeks. Patients have successfully obtained at-home self-measurements of critical flicker frequency (e.g., demonstrating 100% compliance with collecting study self-measurements). Initial inspection of collected data finds participant critical flicker frequency measurements are stable over a six-week measurement period. We are continuing data collection, are improving documentation and support for at-home self-measurement, and are planning for future additional longitudinal studies.