Noninvasive In Vivo Estimation of Blood-Glucose Concentration by Monte Carlo Simulation

Abstract

Continuous monitoring of blood-glucose concentrations is essential for both diabetic and nondiabetic patients to plan a healthy lifestyle. Noninvasive in vivo blood-glucose measurements help reduce the pain of piercing human fingertips to collect blood. To facilitate noninvasive measurements, this work proposes a Monte Carlo photon simulation-based model to estimate blood-glucose concentration via photoplethysmography (PPG) on the fingertip. A heterogeneous finger model was exposed to light at 660 nm and 940 nm in the reflectance mode of PPG via Monte Carlo photon propagation. The bio-optical properties of the finger model were also deduced to design the photon simulation model for the finger layers. The intensities of the detected photons after simulation with the model were used to estimate the blood-glucose concentrations using a supervised machine-learning model, XGBoost. The XGBoost model was trained with synthetic data obtained from the Monte Carlo simulations and tested with both synthetic and real data (n = 35). For testing with synthetic data, the Pearson correlation coefficient (Pearson’s r) of the model was found to be 0.91, and the coefficient of determination (R2) was found to be 0.83. On the other hand, for tests with real data, the Pearson’s r of the model was 0.85, and R2 was 0.68. Error grid analysis and Bland–Altman analysis were also performed to confirm the accuracy. The results presented herein provide the necessary steps for noninvasive in vivo blood-glucose concentration estimation.

Cuffless Blood Pressure Estimation Based on Monte Carlo Simulation Using Photoplethysmography Signals

“…In the public dataset, the PPG signal was collected at a wavelength of 905 nm, and in the self-made dataset, the PPG signal was collected at a wavelength of 940 nm. In the proposed hardware, the IR wavelength is 940 nm, and this device is well established and used in our previous studies [ 5 , 43 ]. It is difficult to change the IR LED in the hardware to match the wavelength of the public dataset.…”

Commercial and Scientific Solutions for Blood Glucose Monitoring—A Review

“…8.1 mg/dL was reported on 12 subjects. Although the results for optical-based sensing systems of [141] and of [142], where 35 patients were evaluated, were located mainly in Zone A and B of the Clarke error grid, the strong dependency on the temperature and environmental aspects, such as scattering light, has to be mentioned. Moreover, their relative bulky structure makes it difficult to integrate into a wearable glucose sensing system.…”

“…Therefore, they are not suitable for a wearable glucose monitoring device considering the state of the art. Moreover, in [142], photoplethysmography (PPG) was used for non-invasive glucose sensing. The reported accuracy was slightly worse than that of commercial sensors with 96.85% in zones A and B and with 3.15% in the dangerous zone D. The limitation of this study was that heterogeneous finger models were considered.…”

“…Most of the systems proposed in Table 12 cover this required range approximately. However, approaches such as that of [104,114,141,142] are also limited in their analyzed detection range. The Clarke error grid results of those have to be interpreted with caution, since it is easier to design a precise sensor system for a narrow detection range than for a broad one.…”

Recent applications of machine learning and deep learning models in the prediction, diagnosis, and management of diabetes: a comprehensive review

“…A Monte Carlo (MC) photon simulation-based model was developed for assessing the concentration of blood glucose through PPG on the fingertip [ 103 ]. The MC method was selected for the simulation of a photon in the finger model due to its flexibility in calculating the optical interplays with the biological tissues.…”