Challenges Applying NMR & MRS to Consumer Health Tracking

Consumers have become accustomed to instruments such as wearables that allow for measurement of biomarkers. Thus far these capabilities, sometimes referred to as “quantified self”, rely on quantification of motion and vital signs and often rely on superficial detectors such as optical sensors. There is a desire and need for more capable devices which enable consumers to be more aware of their health state. One clear area of pressing need is in pre-diabetes and diabetes: a global problem with an estimated 1 in 3 Americans categorized as pre-diabetic and 80% of those unaware of their status and undiagnosed. Because pre-diabetes can be reversed and prevent the development of diabetes in most patients, there is a strong incentive and benefit to detecting these “pre-patients”. Our collaboration has been developing technology and methods that will be commercialized as a benchtop, consumer facing, non-invasive in vivo health-tracking device. The approach relies on Nuclear Magnetic Resonance (NMR) and Magnetic Resonance Spectroscopy (MRS) to quantify a set of metabolites. Such data, acquired repeatedly over time, can be used to infer alterations in various user health states. The final device, DigiScan, is intended for public spaces like pharmacies, gyms, workplaces, and homes, as well as clinical settings. After insertion of a digit, metabolite spectra are read from multiple locations along a finger. The results are displayed on a health dashboard app, offering an overall health score and metrics for factors like pre-diabetes or diabetes, cholesterol, obesity, fitness and recovery, which are linked to variability in such metabolites as glucose, lactate, and lipids. Here we present the project status, including the benchmarking and ground truth experiments, the application of signal processing pipelines, and other challenges faced in developing this novel device.