[11] In 2018 a paper in PLOS ONE showed independent validation data from a clinical investigation comprising 15 subjects with diabetes mellitus type 1 with a mean absolute relative difference (MARD) of 25.8%.
[13] In 2020, German Institute for Diabetes-Technology published data from 15 subjects with type 1 diabetes on a new prototype GlucoBeam based on Raman spectroscopy from RSP Systems Denmark, showing a MARD of 23.6% on independent validation in out-patient setup with up till 8 days without recalibration.
[16] From approximately 2020, onwards there has been increased R&D activity in the space of new NIGM solutions (particularly CGM ones) with renewed focus on approaches that had already been explored, and new ones altogether.
These methods typically involve shining a specific wavelength of light (near-infrared, mid-infrared, or Raman) onto the skin, where it interacts with the glucose molecules.
[18] DiaMonTech AG is a Berlin, Germany-based privately-held company developing the D-Pocket,[19] a medical device that uses infrared laser technology to scan the tissue fluid in the skin and detect glucose molecules.
[21] DiaMonTech has announced that its envisioned follow-up product D-Sensor, will feature continuous measurements, making it a CGM though no release date has been given.
[25] Masimo has also filed new patents through its subsidiary Cercacor (pending as of September 2023) covering a joint continuous glucose monitoring and pump-closed loop delivery system.
[30] The company in 2020 published literature regarding the aforementioned (see above) non-invasive method it had developed with MIT scientists to engage in continuous glucose monitoring using spectroscopy.
In January of 2024, Liom declared it had developed a prototype, with a claimed MARD (Mean Absolute Relative Difference) value to a reference glucose measurement of approximately 9%.
Data from a subset of subjects with type 2 diabetes shoved 99.8% of measurements within A + B zones in the consensus error grid and a mean absolute relative difference of 14.3%.
The measurements form the type 1 diabetes subgroup, showed 96.5% of the points in zones A + B, while the typical indices of accuracy, the mean absolute relative difference (MARD) and RMSE, over the 15 days were 19.9% and 1.9 mmol/L, respectively.
These techniques typically involve applying a specific radio frequency or microwave signal to the skin, which then penetrates the underlying tissues.
The presence of glucose alters the dielectric properties (permittivity and conductivity) of the tissue, leading to changes in the amplitude, phase, or other characteristics of the transmitted or reflected electromagnetic waves.
Hydrogen peroxide is further oxidized at the electrode, producing free electrons, resulting in an electrical current proportional to the glucose concentration in an area of interest.
[47] Synex Medical (based in Boston, US and Toronto, Canada) uses portable magnetic resonance spectroscopy (MRS) for non-invasive glucose monitoring.
Movano said in 2021 that it was building the smallest ever custom radio frequency (RF)-enabled sensor designed for simultaneous blood pressure and glucose monitoring.
[50] SugarBeat, built by Nemaura Medical, is a wireless non-invasive blood glucose monitoring system using a disposable skin patch.
[60] Its device is a combination of magnetohydrodynamic (MHD) technology, advanced algorithms and highly-sensitive biosensors which link to a smartphone app for data collection and reporting.
[62] The company states it is developing the Occuity Indigo,[63] which aims to measure the change in refractive index of the eye to determine the concentration of glucose in the blood.
[64] BOYDSense is a French-based startup developing a noninvasive glucose monitoring device that analyzes breath-based volatile organic compounds (VOCs).
Early clinical trials have demonstrated that these VOCs can reliably indicate blood glucose levels in individuals with type 2 diabetes.