Hamamatsu uses near-IR to measure blood glucose levels

12 Mar 2024

New approach relating glucose to spectral response of hemoglobin could lead to improved diabetes monitoring.

Using optical techniques to assess blood sugar levels is a promising route to monitoring diabetes in patients, potentially allowing alternatives to invasive blood sampling techniques.

Recent examples have included using a confocal Raman spectrometer in a device measuring blood glucose when a patient touches their thumb to the device, with artificial intelligence then computing the results.

A project at Hamamatsu Photonics has now designed a device using near-infrared illumination to make the determination. It has also developed a new way to infer the glucose content from optical study of blood flow, a field termed photoplethysmography (PPG), and from analysis of hemoglobin, rather than tackle the difficulties involved in distinguishing glucose in the blood through absorption alone.

Published in Journal of Biomedical Optics, the study could offer an easier route to the incorporation of blood sugar measurements in commercial smartphones and smartwatches, based on the team's discovery that blood glucose levels create a small phase delay between the near-IR responses of oxy- and deoxyhemoglobin (Hb).

"There is still no practical non-invasive blood glucose monitoring device available for daily use," commented the Hamamatsu team in its published paper. "To address this problem, data mining of PPG data using visible and near-infrared light has found a high correlation between the phase delay in deoxyhemoglobin compared to oxyhemoglobin and blood glucose levels."

Metabolic data from smartwatch measurements

This optical phenomenon, an "asynchronicity between the low-frequency and oscillating components of oxy- and deoxy-Hb signals," can be used as a new index for non-invasive blood glucose measurement, utilizing practical and low-cost visible/near-IR spectroscopy instead of more expensive mid-IR spectroscopy. It should also be fundamentally less sensitive to interference from environmental factors.

To test its theory, the team first used the near-IR sensor in a commercial smartwatch to measure the blood glucose of a subject after drinking different beverages. Similar experiments were then conducted using a custom smartphone holder with a high-brightness LED.

Analysis of the optical data showed that the changes in the phase delay (asynchronicity) between the oscillating components of the HbO2 and Hb signals is closely related to the degree of oxygen consumption during each cardiac cycle.

The next steps will include clinical tests on diabetic individuals to confirm the applicability of the metabolic index in a real-world context, but Hamamatsu believes its technique already points to blood glucose measurement becoming less expensive, more energy efficient, and simpler compared to other noninvasive methods using mid-IR or Raman spectroscopy.

"Our approach could be a powerful tool towards portable and accessible blood glucose level monitoring devices in the future," commented Hamamatsu's Tomoya Nakazawa.