Noninvasive Monitoring of Glucose Using Near-Infrared Reflection Spectroscopy of Skin-Constraints and Effective Novel Strategy in Multivariate Calibration

A comprehensive review on electromagnetic wave based non-invasive glucose monitoring in microwave frequencies

Diabetes is a chronic disease that affects millions of humans worldwide. This review article provides an analysis of the recent advancements in non-invasive blood glucose monitoring, detailing methods and techniques, with a special focus on Electromagnetic wave microwave glucose sensors. While optical, thermal, and electromagnetic techniques have been discussed, the primary emphasis is focussed on microwave frequency sensors due to their distinct advantages. Microwave sensors exhibit rapid response times, require minimal user intervention, and hold potential for continuous monitoring, renders them extremely potential for real-world applications. Additionally, their reduced susceptibility to physiological interferences further enhances their appeal. This review critically assesses the performance of microwave glucose sensors by considering factors such as accuracy, sensitivity, specificity, and user comfort. Moreover, it sheds light on the challenges and upcoming directions in the growth of microwave sensors, including the need for reduction and integration with wearable platforms. By concentrating on microwave sensors within the broader context of non-invasive glucose monitoring, this article aims to offer significant enlightenment that may drive further innovation in diabetes care.

In Vitro Glucose Measurement from NIR and MIR Spectroscopy: Comprehensive Benchmark of Machine Learning and Filtering Chemometrics

The quantitative analysis of glucose using spectroscopy is a topic of great significance and interest in science and industry. One conundrum in this area is deploying appropriate preprocessing and regression tools. To contribute to addressing this challenge, in this study, we conducted a comprehensive and novel comparative analysis of various machine learning and preprocessing filtering techniques applied to near-infrared, mid-infrared, and a combination of near-infrared and mid-infrared spectroscopy for glucose assay. Our objective was to evaluate the effectiveness of these techniques in accurately predicting glucose levels and to determine which approach was most optimal. Our investigation involved the acquisition of spectral data from samples of glucose solutions using the three aforementioned spectroscopy techniques. The data was subjected to several preprocessing filtering methods, including convolutional moving average, Savitzky-Golay, multiplicative scatter correction, and normalisation. We then applied representative machine learning algorithms from three categories: linear modelling, traditional nonlinear modelling, and artificial neural networks. The evaluation results revealed that linear models exhibited higher predictive accuracy than nonlinear models, whereas artificial neural network models demonstrated comparable performance. Additionally, the comparative analysis of various filtering methods demonstrated that the convolutional moving average and Savitzky-Golay filters yielded the most precise outcomes overall. In conclusion, our study provides valuable insights into the efficacy of different machine learning techniques for glucose measurement and highlights the importance of applying appropriate filtering methods in enhancing predictive accuracy. These findings have important implications for the development of new and improved glucose quantification technologies.

Development of sensor system and data analytic framework for non-invasive blood glucose prediction

Periodic quantification of blood glucose levels is performed using painful, invasive methods. The proposed work presents the development of a noninvasive glucose-monitoring device with two sensors, i.e., finger and wrist bands. The sensor system was designed with a near-infrared (NIR) wavelength of 940 nm emitter and a 900-1700 nm detector. This study included 101 diabetic and non-diabetic volunteers. The obtained dataset was subjected to pre-processing, exploratory data analysis (EDA), data visualization, and integration methods. Ambiguities such as the effects of skin color, ambient light, and finger pressure on the sensor were overcome in the proposed 'niGLUC-2.0v'. niGLUC-2.0v was validated with performance metrics where accuracy of 99.02%, mean absolute error (MAE) of 0.15, mean square error (MSE) of 0.22 for finger, and accuracy of 99.96%, MAE of 0.06, MSE of 0.006 for wrist prototype with ridge regression (RR) were achieved. Bland-Altman analysis was performed, where 98% of the data points were within ± 1.96 standard deviation (SD), 100% were under zone A of the Clarke Error Grid (CEG), and statistical analysis showed p < 0.05 on evaluated accuracy. Thus, niGLUC-2.0v is suitable in the medical and personal care fields for continuous real-time blood glucose monitoring.

A joint evaluation method of dynamic spectrum extraction methods for non-invasive blood component measurement based on stability coefficient, data point adoption rate, and smoothness of the spectrum

BACKGROUND AND OBJECTIVE

Dynamic spectrum (DS) theory is a new non-invasive detection method of human blood components that can theoretically eliminate individual differences in static tissues and the influence of other measurement conditions to achieve blood component analysis with high precision. In order to obtain a high signal-to-noise ratio dynamic spectrum, researchers have proposed various dynamic spectrum extraction methods.

METHODS

In this article, we propose three indexes: stability coefficient (SC), data point adoption rate (DAR), and smoothness of spectrum (SS). These solve the difficulty in evaluating different dynamic spectrum extraction methods without establishing mathematical models.

RESULTS

In this study, DS is extracted using different dynamic spectrum extraction methods from the experimental data of 677 volunteers. Then three indexes, SC, DAR, and SS, are calculated. The trends in the scatter plot of the relationship between the three indexes and modeling results of hemoglobin, red blood cell count, and white blood cell count and the related coefficients demonstrate that SC, DAR, and SS are feasible and effective for evaluation. The results show that the root mean square extraction performs best, while the peak-to-peak value and the fast Fourier transform extraction are the worst.

CONCLUSIONS

This study proposes feasible and effective indexes for evaluating dynamic spectrum extraction methods, providing a possibility for further research on high-precision dynamic spectrum extraction methods.

Non-Invasive Glucose Sensing Technologies and Products: A Comprehensive Review for Researchers and Clinicians

Diabetes Mellitus incidence and its negative outcomes have dramatically increased worldwide and are expected to further increase in the future due to a combination of environmental and social factors. Several methods of measuring glucose concentration in various body compartments have been described in the literature over the years. Continuous advances in technology open the road to novel measuring methods and innovative measurement sites. The aim of this comprehensive review is to report all the methods and products for non-invasive glucose measurement described in the literature over the past five years that have been tested on both human subjects/samples and tissue models. A literature review was performed in the MDPI database, with 243 articles reviewed and 124 included in a narrative summary. Different comparisons of techniques focused on the mechanism of action, measurement site, and machine learning application, outlining the main advantages and disadvantages described/expected so far. This review represents a comprehensive guide for clinicians and industrial designers to sum the most recent results in non-invasive glucose sensing techniques' research and production to aid the progress in this promising field.

A novel multipurpose device for dataset creation and on-device immediate estimation of blood glucose level from reflection ppg

We propose a completely non-invasive and highly accurate portable blood glucose estimator, which is simple hardware that anyone, regardless of their prior experience or knowledge, can use to obtain an immediate reading of their blood sugar level. Glucose levels can be monitored in real time and displayed on the device thanks to its infrared LED light source, sensor with built-in amplification unit, processing unit with blood glucose regression model, power management unit for autonomous operation, and display. The device was initially used to create a dataset of photoplethysmography (PPG) signals collected from the fingertip of 50 subjects. The extracted signal features were correlated with the subject's glucose level, which was measured at the same time using a commercial glucometer, and several regression models were constructed. The models were evaluated using signals from up to 110 subjects across three datasets, and the most optimized model was implemented in the device to predict the subject's blood glucose level solely based on PPG in real-time. The device with the built-in model has been subjected to extensive testing to gauge its efficacy. The device's clinical accuracy is encouraging. The pricey strips and needles that must be purchased along with the hardware in the conventional method will no longer be necessary with this device.

An approach to detecting diphenylamine content and assessing chemical stability of single-base propellants by near-infrared reflectance spectroscopy

Diphenylamine (DPA) as a stabilizer component plays an important role in maintaining the chemical stability of single-base propellants (SBPs). This work investigated the feasibility of rapidly detecting the content of DPA in SBP by near-infrared reflectance spectroscopy (NIRS). The quantitative NIR model was developed by intervals selection, spectral pretreatment and factor number optimization. The optimal spectral intervals were determined to be 1081 nm ∼ 1280 nm and 1378 nm ∼ 1602 nm based on the characteristic spectral peaks of DPA. By comparing the performance of the developed models with different preprocessing methods, the best preprocessing method was standard normal variate transformation (SNV) + de-trending (Dr) + Smoothing. The optimal number of factors was 6 for DPA model. Partial least squares (PLS) regression was used to establish the calibration models of DPA. For the developed model, the determination coefficients of calibration and prediction (R_c², R_p²) were 0.9907 and 0.9884, respectively. The root mean square errors of calibration and prediction (RMSEC, RMSEP) were 0.0310 and 0.0342, respectively. The samples in the prediction set were predicted by the developed model, and the average absolute error of the proposed and reference method was only 0.0265. The developed model can be applied in rapid monitor the content of DPA in SBP. In addition, vieille test have demonstrated that the chemical stability of SBP became worse with the decrease of DPA content. The content of DPA contained in the SBP with qualified chemical stability is not less than 0.8753%. Thus, the developed model can be used to judge whether the chemical stability of SBP is qualified or unqualified.

Quantification of glycated hemoglobin and glucose in vivo using Raman spectroscopy and artificial neural networks

Undiagnosed type 2 diabetes (T2D) remains a major public health concern. The global estimation of undiagnosed diabetes is about 46%, being this situation more critical in developing countries. Therefore, we proposed a non-invasive method to quantify glycated hemoglobin (HbA1c) and glucose in vivo. We developed a technique based on Raman spectroscopy, RReliefF as a feature selection method, and regression based on feed-forward artificial neural networks (FFNN). The spectra were obtained from the forearm, wrist, and index finger of 46 individuals. The use of FFNN allowed us to achieve an error in the predictive model of 0.69% for HbA1c and 30.12 mg/dL for glucose. Patients were classified according to HbA1c values into three categories: healthy, prediabetes, and T2D. The proposed method obtained a specificity and sensitivity of 87.50% and 80.77%, respectively. This work demonstrates the benefit of using artificial neural networks and feature selection techniques to enhance Raman spectra processing to determine glycated hemoglobin and glucose in patients with undiagnosed T2D.

Is Raman the best strategy towards the development of non-invasive continuous glucose monitoring devices for diabetes management?

Diabetes has no well-established cure; thus, its management is critical for avoiding severe health complications involving multiple organs. This requires frequent glycaemia monitoring, and the gold standards for this are fingerstick tests. During the last decades, several blood-withdrawal-free platforms have been being studied to replace this test and to improve significantly the quality of life of people with diabetes (PWD). Devices estimating glycaemia level targeting blood or biofluids such as tears, saliva, breath and sweat, are gaining attention; however, most are not reliable, user-friendly and/or cheap. Given the complexity of the topic and the rise of diabetes, a careful analysis is essential to track scientific and industrial progresses in developing diabetes management systems. Here, we summarize the emerging blood glucose level (BGL) measurement methods and report some examples of devices which have been under development in the last decades, discussing the reasons for them not reaching the market or not being really non-invasive and continuous. After discussing more in depth the history of Raman spectroscopy-based researches and devices for BGL measurements, we will examine if this technique could have the potential for the development of a user-friendly, miniaturized, non-invasive and continuous blood glucose-monitoring device, which can operate reliably, without inter-patient variability, over sustained periods.

Review of Ethanol Intoxication Sensing Technologies and Techniques

The field of alcohol intoxication sensing is over 100 years old, spanning the fields of medicine, chemistry, and computer science, aiming to produce the most effective and accurate methods of quantifying intoxication levels. This review presents the development and the current state of alcohol intoxication quantifying devices and techniques, separated into six major categories: estimates, breath alcohol devices, bodily fluid testing, transdermal sensors, mathematical algorithms, and optical techniques. Each of these categories was researched by analyzing their respective performances and drawbacks. We found that the major developments in monitoring ethanol intoxication levels aim at noninvasive transdermal/optical methods for personal monitoring. Many of the "categories" of ethanol intoxication systems overlap with each other with to a varying extent, hence the division of categories is based only on the principal operation of the techniques described in this review. In summary, the gold-standard method for measuring blood ethanol levels is through gas chromatography. Early estimation methods based on mathematical equations are largely popular in forensic fields. Breath alcohol devices are the most common type of alcohol sensors on the market and are generally implemented in law enforcement. Transdermal sensors vary largely in their sensing methodologies, but they mostly follow the principle of electrical sensing or enzymatic reaction rate. Optical devices and methodologies perform well, with some cases outperforming breath alcohol devices in terms of the precision of measurement. Other estimation algorithms consider multimodal approaches and should not be considered alcohol sensing devices, but rather as prospective measurement of the intoxication influence. This review found 38 unique technologies and techniques for measuring alcohol intoxication, which is testament to the acute interest in the innovation of noninvasive technologies for assessing intoxication.

In Vivo Near-Infrared Noninvasive Glucose Measurement and Detection in Humans

In optical noninvasive glucose detection, how to detect the glucose-caused signals from the constant human variations and disturbed probing conditions is always the biggest challenge. Developing effective measurement strategies is essential to realize the detection. A near-infrared (NIR) spectroscopy-based strategy is studied to effectively solve the in vivo measurement issues, obtaining clean blood glucose-caused signals. Two solutions composing our strategy are applied to the NIR spectroscopy-based measurement system to acquire clean raw signals in the data collection, which are a customized high signal-to-noise ratio multi-ring InGaAs detector to reduce the influence of human variations, and a fixing and aiming method to reproduce a consistent measurement condition. Seventeen cases of glucose tolerance test (GTT) on healthy and diabetic volunteers were conducted to validate the strategy. The human experiment results clearly show that the expected blood glucose changes have been detected at 1550 nm. The average correlation coefficient of the 17 cases of GTT between light signal and glucose reference reaches 0.84. The proposed measurement strategy is verified feasible for the glucose detecting in vivo. The strategy provides references to further studies and product developments for the NIR spectroscopy-based glucose measurement and references to other optical measurements in vivo.

Products for Monitoring Glucose Levels in the Human Body With Noninvasive Optical, Noninvasive Fluid Sampling, or Minimally Invasive Technologies

BACKGROUND

Conventional home blood glucose measurements require a sample of blood that is obtained by puncturing the skin at the fingertip. To avoid the pain associated with this procedure, there is high demand for medical products that allow glucose monitoring without blood sampling. In this review article, all such products are presented.

METHODS

In order to identify such products, four different sources were used: (1) PubMed, (2) Google Patents, (3) Diabetes Technology Meeting Startup Showcase participants, and (4) experts in the field of glucose monitoring. The information obtained were filtered by using two inclusion criteria: (1) regulatory clearance, and/or (2) significant coverage in Google News starting in the year 2016, unless the article indicated that the product had been discontinued. The identified bloodless monitoring products were classified into three categories: (1) noninvasive optical, (2) noninvasive fluid sampling, and (3) minimally invasive devices.

RESULTS

In total, 28 noninvasive optical, 6 noninvasive fluid sampling, and 31 minimally invasive glucose monitoring products were identified. Subsequently, these products were characterized according to their regulatory, technological, and consumer features. Products with regulatory clearance are described in greater detail according to their advantages and disadvantages, and with design images.

CONCLUSIONS

Based on favorable technological features, consumer features, and other advantages, several bloodless products are commercially available and promise to enhance diabetes management. Paths for future products are discussed with an emphasis on understanding existing barriers related to both technical and non-technical issues.