Non-invasive blood glucose estimation method based on the phase delay between oxy- and deoxyhemoglobin using visible and near-infrared spectroscopy.

Significance: Many researchers have attempted to estimate blood glucose levels (BGLs) noninvasively using near-infrared (NIR) spectroscopy. However, the optical absorption change induced by blood glucose is weak in the NIR region and often masked by interference from other components such as water and hemoglobin. Aim: Instead of using direct optical absorption by glucose, this study proposes an index calculated from oxy- and deoxyhemoglobin signals that shows a good correlation with BGLs while using conventional visible and NIR spectroscopy. Approach: The metabolic index, which is based on tissue oxygen consumption, was derived through analytical methods and further verified and reproduced in a series of glucose challenge experiments. Blood glucose estimation units were prototyped by utilizing commercially available smart devices. Results: Our experimental results showed that the phase delay between the oxy- and deoxyhemoglobin signals in near-infrared spectroscopy correlates with BGL measured by a conventional continuous glucose monitor. The proposed method was also confirmed to work well with visible spectroscopy systems based on smartphone cameras. The proposed method also demonstrated excellent repeatability in results from a total of 19 oral challenge tests. Conclusions: This study demonstrated the feasibility of non-invasive glucose monitoring using existing photoplethysmography sensors for pulse oximeters and smartwatches. Evaluating the proposed method in diabetic or unhealthy individuals may serve to further increase its practicality.

Development of Portable, Wireless and Smartphone Controllable Near-Infrared Spectroscopy System.

We have developed portable near-infrared tissue oxygenation monitoring systems, called the "PocketNIRS Duo" and the "PocketNIRS HM", which features wireless data communication and a sampling rate of up to 60 data readings per second. The systems can be controlled by smartphone or personal computer. We demonstrate the efficacy of the systems for monitoring changes in brain and arm muscle hemodynamics and oxygenation in breath-holding and cuff-occlusion tests, respectively.Our systems should prove to be useful as an oxygenation monitor not only in research but also in healthcare applications.