Discussion on the validity of NIR spectral data in non-invasive blood glucose sensing

Abstract

In this paper, the effects of two-dimensional correlation spectroscopy (2DCOS) on chance correlations in the spectral data, generated from the correlations between glucose concentration and some undesirable experimental factors, such as instrument drift, sample temperature variations, and interferent compositions in the sample matrix, are investigated. The aim is to evaluate the validity of the spectral data set, instead of assessing the calibration models, and then to provide a complementary procedure for better verifying or rejecting the data set. It includes tracing back to the source of the chance correlation on the chemical basis, selecting appropriate preprocessing methods before building multivariate calibration models, and therefore may avoid invalid models. The utility of the proposed analysis is demonstrated with a series of aqueous solutions using near-infrared spectra over the overtone band of glucose. Results show that, spectral variations from chance correlations induced by those experimental factors can be determined by the 2DCOS method, which develops avenues for prospectively accurate prediction in clinical application of this technology.

Challenges in the noninvasive detection of body composition using near-infrared spectroscopy

“…39,40 As illustrated in Fig. 2, when the concentration of an interfering component happens to have a similar change as that of the interesting component, chance correlation occurs and the multivariate regression model will take both changes as useful signal induced by the glucose.…”

“…41 Furthermore, chance correlations can also result from instrument shift and temperature, which decreases the prediction precision. 39,40 To solve the problem, we propose that the random sampling is the best way to avoid the chance correlation induced by the instrument shift, while in the in vitro multi-component experiments, chance correlation induced by the interference components should be averted by suitable design of the steps. As to the in vivo experiments, such chance correlations cannot easily be removed since measurement conditions and physiological status cannot be intentionally controlled.…”

“…Conclusions made on such basis can be more condently applied to the in vitro and in vivo experiments, such as the°oating reference method. 23,[37][38][39][40][41][42][43] Moreover, we attempt to develop new methods of reconstructing the parameters based on the di®use-re°ectance spectroscopy. 44 Chenxi Li put forward a novel parameter reconstruction method in the following steps, as shown in Fig.…”

Monte Carlo simulation-based thinning and calculating method for noninvasive blood glucose sensing by near-infrared spectroscopy

“…However, this method has not been widely used in clinical use since it is limited to series of factors, such as the changes of physiological background and the indeterminate experimental conditions, which may a®ect the signals of blood glucose to a great extent. [5][6][7] Here,°o ating reference theory (FRT) 4,8 was applied to address the aforementioned limitation. It can be pointed out that at the°oating reference point (FRP), the di®use re°ection was independent of the change of blood glucose concentration.…”

“…Min 7 and Wang 10 put emphasis on developing mathematical models for modifying the in°uence of instrument drifts. Min 6 conducted in vitro aqueous glucose solution experiments with the following three models, single-beam measurement, singlebeam measurement with reference correction and double-beam measurement with double detectors. It demonstrated that the predictive ability of the PLS model based on double-beam measurement performed more stably with an average RMSEP of 13.76 mg/dL.…”