Feasibility of Raman spectroscopy as a potential in vivo tool to screen for pre‐diabetes and diabetes

Abstract

In this article, we investigated the feasibility of using Raman spectroscopy and multivariate analysis method to noninvasively screen for prediabetes and diabetes in vivo. Raman measurements were performed on the skin from 56 patients with diabetes, 19 prediabetic patients and 32 healthy volunteers. These spectra were collected along with reference values provided by the standard glycated hemoglobin (HbA1c) assay. A multiclass principal component analysis and support vector machine (PCA‐SVM) model was created from the labeled Raman spectra and was validated through a two‐layer cross‐validation scheme. Classification accuracy of the model was 94.3% with an area under the receiver operating characteristic curve AUC of 0.76 (0.65–0.84) for the prediabetic group, 0.86 (0.71–0.93) for the diabetic group and 0.97(0.93–0.99) for the control group. Our results suggest the feasibility of using Raman spectroscopy for the classification of prediabetes and diabetes in vivo.

Comment on “Feasibility of Raman spectroscopy as a potential in vivo tool to screen for pre‐diabetes and diabetes”

“…The data that support the findings of this study are available at [doi.org/10.1002/jbio.202200055], reference number [9].…”

“…Creation of methodology for in vivo analysis of tissues by means of spectroscopy is a challenging task as it requires complex studies combining optics and spectroscopy, biomedicine, biochemistry and machine learning. In the recent studies published in the Journal of Biophotonics [9] the authors presented the results of Raman spectroscopy application for detecting diabetes (T2DM) and pre‐diabetes (preT2DM). The authors utilized PCA‐SVM (principal component analysis—support vector machine) complemented with a dual‐layer cross‐validation (CV) scheme to classify healthy volunteers, diabetes and pre‐diabetes patients.…”

“…(D) ROC curve analysis for the proposed model with five PCs. (E) Processed Raman spectra (mean ± SD) of control group (black), pre‐diabetic patients—preT2DM (blue), diabetic patients—T2DM (red), Raman spectra from HbA1c in vitro (maroon), loadings from the first principal component (cyan), loadings from the second principal component (magenta) and loadings from the third principal component (green) (data from E. Guevara et al [9]).…”

“…In addition, the authors demonstrate performance of the proposed classifier (Figure 1C) with the data for all enrolled individuals (n = 107). However, the authors divided their data into train and test datasets, and it seems reasonable (E) Processed Raman spectra (mean ± SD) of control group (black), prediabetic patients-preT2DM (blue), diabetic patients-T2DM (red), Raman spectra from HbA1c in vitro (maroon), loadings from the first principal component (cyan), loadings from the second principal component (magenta) and loadings from the third principal component (green) (data from E. Guevara et al [9]).…”

“…Today Raman spectroscopy becomes a very promising tool for in vivo noninvasive analysis of skin tissues [1][2][3], including such applications as cancer detection [4][5][6] (including rare diseases [7]), diabetes screening [8,9], kidney failure monitoring [10], and so on. Creation of methodology for in vivo analysis of tissues by means of spectroscopy is a challenging task as it requires complex studies combining optics and spectroscopy, biomedicine, biochemistry and machine learning.…”