COMPARATIVE ANALYSIS OF SOURCE WAVELENGTH COMBINATION IN MULTI WAVELENGTH PHOTOPLETHYSMOGRAPHY METHOD FOR NON-INVASIVE HBA1C MEASUREMENT
Diabetes, a chronic condition marked by impaired glucose metabolism, resulted in 1.5 million deaths globally in 2019 and had a prevalence of 10.6% among Indonesian adults in 2021. With no cure available, effective management depends on maintaining healthy lifestyle and regularly monitoring blood...
Saved in:
Main Author: | |
---|---|
Format: | Theses |
Language: | Indonesia |
Online Access: | https://digilib.itb.ac.id/gdl/view/84835 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Institut Teknologi Bandung |
Language: | Indonesia |
Summary: | Diabetes, a chronic condition marked by impaired glucose metabolism, resulted
in 1.5 million deaths globally in 2019 and had a prevalence of 10.6% among
Indonesian adults in 2021. With no cure available, effective management depends
on maintaining healthy lifestyle and regularly monitoring blood glucose levels.
Glycated hemoglobin (HbA1c) is a key marker for assessing average blood
glucose levels and associated complications. However, traditional HbA1c
measurement methods are invasive, which poses challenges in terms of patient
comfort, cost, and accessibility, especially in regions with limited clinical
resources.
This study explores the potential of non-invasive HbA1c measurement using
photoplethysmography (PPG) technology, which has been successfully employed
for non-invasive blood oxygen level monitoring through light absorption
characteristics of hemoglobin. We investigated the use of multiple wavelengths
(465, 525, 615, and 880 nm) for measuring HbA1c in 37 research subjects
consisting of 13 normal, 10 prediabetic, and 14 diabetic individuals. Our results
show that combining blue-red and green-red wavelengths yields the highest
correlation, with Pearson’s r values of 0.921 and 0.845, respectively. Clarke's
Error analysis indicates that these wavelength combinations achieve accuracy
rates of 94.6% and 78.38% in region A, with data primarily falling in regions A
and B. |
---|