This repository is no longer maintained. We moved to https://github.com/AIT-brainlab/raman-for-glucose-measurement

Table of Content

• Measure Blood Glucose with Raman Spectroscopy
  • Resources
  • What are we doing?
  • How?
  • The rabbit hole
  • What we have done
  • MILESTONE
  • Why are we sharing this?

Measure Blood Glucose with Raman Spectroscopy

This repository is for the ongoing research of Measuring Glucose in Blood with Raman Spectroscopy.

Resources

• To see the current status of the project
• To see to-do list (issues)
• To discuss
• To read document

What are we doing?

To gold standard way to measure the concentration of glucose in blood is by sampling the blood directly (invasive). This creates a couple of problems; (1) it hurts, (2) no continuous measurement, and (3) risk of infection. All researchers are asking "Can we do this in a non-invasive manner?"

How?

Well, based on our literature review, there are quite a handful of ways. We could use some other samples like saliva and tears. Or we could try to get the blood without taking the blood.

What????

Yes, it seems possible to measure the blood through the skin (transcutaneous) via a family of spectroscopy techniques. Both Near-Infrared Spectroscopy (NIRS) and Raman Spectroscopy are being compared frequently since they are known to complement each other (absorption vs scattering). The decisive factor is the ability to measure an aqueous sample (containing water). This is because the water ($H_2O$) absorbs a large band of the NIR spectrum, so it is difficult to measure the substance existing in the water. Thus, we chose to use Raman.

This decision is backed up by many publications. Raman can be used to measure the concentration in (1) water, and (2) blood. In addition, multiple researches show that a measurement of Raman on (3) skin can also be used to find Glucose concentration in blood (glycemia). The results show a very high correlation between the gold-standard invasive method and the on-skin Raman method, thus, luring us to this rabbit hole.

The rabbit hole

If this is easy then we would not discuss this. Our research group consists of one Ph.D. in Computer Science and a master's student in Data Science (who has experience in Raman measurement). The project kicked off in 2022 but we have not been able to reproduce any publication. So this year (2024), with all the failures we had, we came up with a plan.

What we have done

• We built a lab room for Raman's experiment.
  • We fixed the problem with noise in our measurement.
• We 3D print the tube that encloses the Raman scattering.
• We conducted several on-skin Raman experiments.
  • We did an OGTT experiment in August 2023.
  • We experimented on different measuring sites (fingertip, nail fold, thenar, forearm) Sep-Oct 2023.
  • We did an extensive experiment (because of the noise mentioned earlier) in November 2023.
  • We experimented with the stretch method to change the volume of blood in November 2023.
• We studied Hilbert Vibration Decomposition (HVD) as a preprocessing for on-skin Raman in December 2023.

MILESTONE

• Reproduce measuring glucose concentration in water by February 2024
• Reproduce measuring glucose concentration in blood by February 2024
• Implement Hilbert Vibration Decomposition (HVD) for extracting the Raman feature by February 2024
• If all the above works, Reproducing measuring glucose concentration on-skin Raman by April 2024

Why are we sharing this?

Well, we do need help. So this is the way we think we can find collaboration.

And, our objective is to publicize what we learned and found in the hope it would benefit someone who wants to jump into this field of research.

We might not be able to make a journal/paper but we do believe our experience is valuable to someone....