Innovative Non-Invasive Blood Glucose Monitoring by Using Multiband Split Ring Resonator Monopole Antenna (MSRRMA)

doi:10.21203/rs.3.rs-5263082/v1

This paper presents a multiband, compact monopole antenna based on an irregular curved shape of split ring resonator design MSRRMA to enable continuous monitoring of human blood glucose concentration level without requiring invasive blood samples. The proposed sensor resonates at frequencies of 0.94, 1.5, 3, 4.6, 6.3 and 7.2 GHz, with reflection coefficient impedance bandwidths ≤-10 dB at 4%, 3.5%, 65%, and 50%, respectively. Two cells of SRR metamaterial shape are used to create notches at 1.7 GHz and 4.4 GHz. The proposed monopole antenna is constructed using a cost-effective FR4 substrate with dimensions of 35 × 50 × 1.6 mm³. The 3D electromagnetic simulation using computer simulation technology (CST) software is employed to simulate the proposed antenna sensor with a human finger phantom model, measuring the sensor's sensitivity to blood glucose concentrations. The finger phantom model is positioned at various angles around the antenna to detect frequency changes relevant to blood glucose content monitoring. In experimental research, a real human finger is placed around the proposed sensor to measure resonant frequency, magnitude, and phase changes. The proposed sensor is fabricated to validate the results, highlighting its benefits and superiority in sensitivity compared to alternative methods in the literature.

Glucose levels

monopole antenna

non-invasive

multiband and sensor

continuous monitoring

Split ring resonator (SRR)

The main characteristic of diabetes is that it is typically brought on by extremely high blood glucose levels for an extended length of time [1], [2]. Over 60% of the global diabetes population lives in a "risk zone," accounting for 463 million cases of the disease, according to the World Health Organization (WHO) and the International Diabetes Federation (IDF) [2]. To increase the accuracy of the findings, it is advised that this identification be done frequently. Finger prick tests and glucose monitoring are two methods that can be used to identify diabetics [3], [4] albeit they are seen as intrusive. Nevertheless, because blood samples need to be drawn frequently, intrusive monitoring are harmful to the patient.

A band-stop filter with a defective ground structure (DGS) was used to monitor intravenous glucose levels, as reported in [5]. Moreover, it is less sensitive than previously reported to variations in blood sample glucose levels and penetration depth. However, there are a number of options presented when it comes to check blood sugar levels [6]-[12]. While in [13] is presented a regular shaped of SRR sensor to monitor the glucose level. Moreover Satish, proposed a different pressure sensor in [14], and Lukas, provided several creative microwave sensors in [14]. Deshmukh and Ghongade, devised a method for measuring glucose levels by estimating the glucose levels and computing the dielectric characteristics using a ring resonator [15]. The detection of glucose in the presence of animal tissues was addressed by the authors of [16] by using mm-wave techniques. Furthermore, a variety of novel techniques for more precisely determining the glucose content in blood samples were described in [17–23]. Microwave sensors are used to measure blood glucose levels in [24–25]. This is a non-invasive technology that has been developed. The most popular non-invasive glucose monitoring technology is the antenna system because of its great sensitivity to even small fluctuations in blood glucose levels, ease of deployment, resistance to noise, temperature, and other disturbances, and deep penetration. In [27] is presented the building of a patch antenna with 2.4 and 5.8 GHz operating frequencies that is loaded with glucose and deionized water medical applications. The development of phantom-loaded patch antenna at 4.75 GHz is described in [28]. A variety of liquid phantoms, such as physiological solutions and pig blood, were assessed; the samples under analysis had glucose concentrations ranging from 150 mg/dL to 550 mg/dL. It was found that the solution caused a linear shift in frequency of 5 MHz, with no correlation between the readings and samples' varying volumes and temperatures. In [29], meta-materials can also be used to identify glucose. The current status of research on glucose level monitoring sensors and the need for more sensitivity sensor development are well explained in recent review studies in [30]. While in [31] the simulated result sensitivity in differences on a small range of glucose changes from 0 mg/dL to 200 mg/dL is 0.017 dB/mg/dL at three operating bands.

The paper introduces a proposed system for a compact monopole antenna sensor based on an irregularly shaped split ring resonator (MSRRMA) for monitoring blood glucose levels. This design enables multiple resonance frequencies (0.94, 1.5, 3, 4.6, 6.3, and 7.2 GHz), which is beneficial for monitoring blood glucose levels across a range of applications. The proposed sensor allows for continuous monitoring of blood glucose levels without the need for invasive blood sampling. This represents a significant improvement over traditional glucose monitoring methods that require finger pricks or other invasive techniques. The incorporation of two cells of SRR metamaterial to create specific notches at 1.7 GHz and 4.4 GHz is innovative and enhances the antenna's ability to selectively respond to changes in blood glucose levels.


Figure 1: Proposed sensor model.

The proposed design could be improved the sensitivity and specificity of the monitoring system. Constructed on a commercial FR4 substrate with compact dimensions (35 × 50 × 1.6 mm³), the design is practical for real-world applications. The compactness of the antenna makes it suitable for wearable or portable devices. Moreover, the sensor is tested on real human fingers, measuring resonant frequency, magnitude, and phase changes. This method of empirical validation provides strong evidence of the sensor's effectiveness, enhancing the credibility of the research findings.

The organization of the paper is as follows: The design process, parametric analysis, and microstrip antenna structure are all covered in Section II. Results and comments are presented in Section III, along with a modelling of a study on frequency changes related to a finger phantom. Results and discussion are presented in Section IV and V, respectively along with a modelling of a finger phantom-related investigation on frequency shifts for various finger phantom positions and glucose levels. Section VI concludes the paper.

The analysis will be conducted using the CST electromagnetic field simulator, along with optimization of the proposed antenna. The development of the antenna progresses through various stages, as illustrated in Figs. 2(a)–(c), leading to the final design shown in Fig. 2(d). The antenna is constructed on a low-cost FR4 substrate with a thickness of 1.6 mm, a relative permittivity of 4.4, and a loss tangent of 0.02. Antenna (1) is designed as a traditional elliptical sector shape, fed by a rectangular transmission line with step impedance to provide an input of 50 Ω and improve impedance matching, resonating at 2 GHz. The reflection coefficient magnitude and phase responses at each design stage are shown in Fig. 3.

As depicted in Fig. 2(f), the antenna consists of a rectangular ground plane and an elliptical patch with major radius W2 and minor radius L1 [32]. In the second design step, a gap is etched in the monopole shape to increase the current path and reduce the resonant frequency from 2 GHz to 1.3 GHz, resulting in a size reduction of approximately 41%, as shown in Fig. 2(b). The |S11| response for this step is shown as a red dashed line in Fig. 3.

In the next step, an irregular curved shape is introduced to form the first cell of the proposed split-ring resonator (SRR), as seen in Fig. 3(c). The |S₁₁| response is represented by the black line in Fig. 3(a). The resonant frequency is further reduced to 0.95 GHz, and a notch is created at 1.7 GHz, with a 44% reduction in size from the original shape. Another irregular curved SRR shape is then added, forming a second cell, which produces a second notch at 4.4 GHz, as shown in Figs. 2(d) and 3.

In the final design step, a U-shaped structure is etched onto the ground plane beneath the transmission line to improve impedance matching. This is represented by the black line in Figs. 2(e) and (f). The |S₁₁| response shows an ultra-wideband (UWB) performance with two notches at 1.7 GHz and 4.4 GHz, achieving a sensor size reduction of about 57.5% from the original.

The proposed sensor antenna offers better matching, reduced losses, and additional bands, maximizing measurement accuracy. The inclusion of SRRs enhances the antenna’s sensitivity by boosting magnetic coupling, which is essential for strong interactions with applied electromagnetic fields in the region of human finger positioning. The simulation parameters for the proposed sensor are listed in Table 1.

Table 1

Dimensions of the Proposed Antenna Sensor.
Parameters	W1	W2	L1	L2	X1	X2	X3	X4	X5	X6	X7	Y1
Dimension(mm)	2	35	5.1	16.6	3	1.92	4	1	5	1	7	11.5
Parameters	Y2	Y3	Y4	Y5	Y6	R1	R2	R3	R4	R5	R6	t
Dimension(mm)	17	2	7	2	5.1	14.5	11	9.5	8	6.5	5	0.035

Initially, the glucose sample with different concentration are characterized using the canonical model to determine the electric properties of the samples. There are two methodologies for identifying the different solutions with various glucose concentrations levels.

A. First Methodology

The evaluation of various glucose concentrations was performed by simulating a plastic container with a dielectric constant of 2.2, as shown in Fig. 4. The container, cylindrical in shape, was filled with equal amounts of water mixed with varying quantities of glucose to achieve specific concentration levels. It was placed over the proposed sensor, directly above the SRR region, where the electromagnetic field coupling is strongest due to the intense interaction between the magnetic field and the container.

The container has a plastic construction, with an inner diameter of 12 mm, a thickness of 1 mm, and a height of 10 mm. Because the electric and magnetic fields are concentrated within the container, it was positioned over the SRR area to maximize interaction. The electrical properties of different glucose concentrations at 2.45 GHz, based on the Cole-Cole and Debye models, are shown in Table 2.

The reflection coefficient response of the proposed sensor for different glucose concentrations is shown in Fig. 5. Table 3 summarizes the results of the simulated frequency shifts and variations in the magnitude and phase of the reflection coefficient for the proposed sensor when a thin layer of glucose solution is applied in the container.

Table 2: The Electrical Properties of Glucose Levels at 2.45 GHz.
Glucose Concentration (mg/dL)	$\:{\varvec{\epsilon\:}}_{\varvec{r}}$	$\:\varvec{\sigma\:}$(S/m)
0	69.8470	3.4518
100	69.8203	3.4405
150	69.8070	3.4350
350	69.7550	3.4130

B. Second Methodology

The creation of an antenna sensor that could monitor changes in blood glucose concentration from the fingertips is the main goal of this research. To validate the proposed operation, a real human finger is mimicked by a finger phantom model created in a CST environment. Dielectric materials with varied dielectric constants and conductivities are used to simulate the different layers of the finger, including skin, fat, muscle, blood, and bone. The thicknesses of the various layers of the finger phantom together with their respective dielectric constants, which are given in Table 4 based on the Cole-Cole and Debye parameter polynomials.

Variations in a person's blood glucose concentration could be detected when the finger is placed along with the proposed design. Blood and fat are two layers from which glucose levels could be sensed. Nevertheless, fat is preferable because it has higher glucose levels than blood. Differences in shift in the resonance ∆F, phase (degrees), and S11 magnitude, could be attributed to differences in weight or finger size for person under test. The location of the antenna design has the greatest impact on the findings, both in terms of magnitude and phase, as the radiation from the antenna changes with respect to the finger's layers at different points. The buffer's design consists of glass bricks with approximate dimensions 5×2.5 cm², and thickness equal to 0.1 cm.

To begin examining the effects of the finger, the shape of the human finger is positioned in three different configurations: without a buffer, with a buffer, and rotating the finger's direction relative to the antenna-based sensor. A buffer must be used since the PCB conductivity has an impact on the sensitivity measurement findings when the finger matches with the antenna directly. The human finger placed over the two cell of split ring resonator shaped sensor as shown in Fig. 6. There are two possible configurations along or perpendicular the current pass of the antenna monopole feed line.

The results of both configurations show that there is similar sensor results obtained in reflection coefficient magnitude and phase as shown in Fig. 6 (c) and (d), respectively. This test mean that the effect of human figure ordination is minimal and not it does not affect the blood glucose reading. Figure 7 shows the simulation response of the sensor antenna with human fingertip with different glucose concentration level with different glucose level as 100, 150 and 350 mg/dl. The finger phantom is placed above the radiating element of the proposed sensor, and changes in frequency, S11 magnitude and phase are used to gauge fluctuations in the glucose concentration.

Placing the finger phantom on the radiating element causes a comparable shift in the reflection coefficient's magnitude and phase. As this setup produces a reasonable degree of frequency shift, it is ideal to place the finger phantom close to the top of the radiating element for the purpose of measuring the glucose content. Table 5 shows the unique frequency shifts obtained for various finger phantom positions on the antenna.

Moreover, to guarantee that non-invasive blood glucose monitoring devices are safe, efficient, and comfortable for users as well as to comply with regulatory criteria and optimize device performance simulation and SAR calculation are essential. The body's rate of electromagnetic energy absorption is measured by SAR.

Specific Absorption Rate (SAR) is strictly regulated by organizations such as the International Commission on Non-Ionizing Radiation Protection (ICNIRP) and the Federal Communications Commission (FCC) in the USA to ensure that electronic devices are safe for human use. Compliance with these limits is essential for devices to receive the necessary approvals. The FCC and ICNIRP set SAR limits at 1.6 W/kg and 2.0 W/kg, averaged over 1 gram and 10 grams of tissue, respectively, with similar regulations in the European Union.

In this study, a finger phantom is placed above the radiating element of the proposed sensor to calculate the absorption power and ensure that the system is safe for human use, as illustrated in Fig. 8. The maximum SAR value is approximately 2.5 W/kg at 6.7 GHz, while the average value over the operating band is around 2 W/kg.

IV. Sensor Measurement Results

The multi band monopole antenna sensor is fabricated as shown in Figure 9 to validate the design results of the MSRRMA sensor. This sensor is fabricated in the Microstrip Dept., of the Electronics Research Institute by using photographic technique to fabricate the printed circuit board (PCB). The MSRRMA sensor is measured on air, then SAR is measured at different resonant frequency and the two technologies are applied as plastic container with different glucose concentration and human fingertip.

A. Sensor Fabrication and Measurement in Air

Printed circuit boards (PCBs) are utilized in the fabrication of the proposed multi-band monopole antenna sensor. To evaluate the design performance, a low-cost FR4 substrate is used in a precise photolithography technique, as shown in Figure 9. The sensor is fabricated in the Microstrip section of the Electronics Research Institute. The reflection coefficient of the proposed sensor is measured in air using the Rohde & Schwarz ZVA 67 Vector Network Analyzer.

There is good agreement between the simulated and measured results; however, some variations are inevitable. Understanding and accounting for these differences can help refine the design and testing processes. The discrepancies may be due to the idealized conditions in simulations compared to real-world factors. Enhancements in soldering techniques, accurate material characterization, manufacturing processes, and regular calibration can significantly reduce the gap between simulated and measured outcomes.

B. SAR Measurement

SAR measurements are essential for ensuring the security of wireless devices since they quantify the exposure to electromagnetic fields. Manufacturers may verify that their devices are safe and reliable in real-world environments by following standard operating procedures and using calibrated equipment, as illustrated in Figure 10 at the Electronics Research Institute, central Lab. The proposed sensor's observed SAR values at 1.5, 3 GHz and 5.2 GHz are displayed in Tables 6. According to Table 6, the SAR values obtained for the suggested sensor at different power levels fall within the accepted safe limits as stated by the IEEE standard, which is 1.6 W/kg. As demonstrated in, the SAR is measured at various resonance frequencies and input powers. The SAR is measured at different resonant frequencies and input power as shown in Table 6.

C. Characterization of Glucose

The glucose concentration was prepared by dissolving the prescribed amounts of glucose in distilled water to yield concentrations of 50 mg/dl, 100 mg/dl, 200 mg/dl, and 350 mg/dl. Then, we entered the various concentrations into a DAK device to measure the properties of the glucose concentration, such as permittivity real value and loss tangent over the operating band as shown in Figure 10.

First, an inverse relationship between the magnitude of the dielectric constant, and the value of glucose level are discovered. When the value of glucose concentration level is large as 350 mg/dl, the real value of the dielectric constant is less than the value of the low glucose level as 0 mg/dl. As well as the variation of the permittivity within the operating band from 3 to 4 GHz and from 5 to 6 GHz are larger than other measured operating bands. Second, on the contrary, the imaginary part of glucose concentration level increased with the value glucose level.

D. Measurements with Glucose

This part contains comparable methodologies to the simulation using the human finger and a container. The container was employed at specific concentration levels during characterization. that a glucometer was used to measure this concentration.

i Measurements with Glucose Solutions with Container

A recommended reflection sensor's performance is measured and given. A dielectric container is attached to the prototype sensor's top in order to release the liquid into the sensing area. A plastic container with a e_r=2.2 component is placed symmetrically over the sensor region using a Rohde & Schwarz VNA, as seen in Figure11(a). The container's base measures 12.57× 2 × 1 mm³. Using a pipette, various concentrations of glucose (0,100, 150, and 350 mg/dL) are administered; the same amount is used for each concentration. The results of different concentrations of the transmission and reflection coefficients are shown in Figure 11. By analysing the shifts in the frequency, magnitude, and phase of the S-parameter, as shown in Figure 11, The difference between the glucose concentration in the container are determined by measurements the difference of S-parameters, the highest shift in frequency, magnitude and phase at 350 Mg/dL concentration.

ii Measurements with Finger

In the second case, using finger tips in measurements are validated. According to recent research, finger measurements might be a useful technique for determining how various blood levels interact with one another. Its appendages are meant to be used by humans. Approval from an Ethical Perspective NILES-EC-CU 24/1/2, issued on August 2024 has been obtained. The measuring process requires participants (healthy & diabetic participants) to fast for six hours prior to the test. After checking their glucose levels, the participants are re-tested after two hours of eating as a second round of testing. Glucose levels of participants are expected to reach higher values compared to prior status while fasting. Before the test began, the participants washed their hands to remove any contaminants, and they are given around fifteen minutes to relax until their bodies stabilized.

In order to get two normal blood glucose levels one for each of the two male and female cases and one for the diabetic case the glucose concentration is measured using the Nano VNA during fasting and after eating for a total of four cases. the concentration of glucose in the blood is initially measured using glucometer as well as the Nano VNA after calibration as shown in Figure 12. The magnitude and phase of S-parameters are recorded at the four cases and presented in Figure 12. The amount of glucose in the blood is first measured with a glucometer as shown in Table 7.

As shown in Figure 12, after measuring the four distinct cases and comparing the results after and before fasting, we can detect the difference between the measurement made after and before fasting by examining the shift and frequency changes. Next section, will be discussed these differences of results.

The multi band monopole antenna sensor is fabricated as shown in Fig. 9 to validate the design results of the MSRRMA sensor. This sensor is fabricated in the Microstrip Dept., of the Electronics Research Institute by using photographic technique to fabricate the printed circuit board (PCB). The MSRRMA sensor is measured on air, then SAR is measured at different resonant frequency and the two technologies are applied as plastic container with different glucose concentration and human fingertip.

A. Sensor Fabrication and Measurement in Air

Printed circuit boards (PCBs) are utilized in the fabrication of the proposed multi-band monopole antenna sensor. To evaluate the design performance, a low-cost FR4 substrate is used in a precise photolithography technique, as shown in Fig. 9. The sensor is fabricated in the Microstrip section of the Electronics Research Institute. The reflection coefficient of the proposed sensor is measured in air using the Rohde & Schwarz ZVA 67 Vector Network Analyzer.

There is good agreement between the simulated and measured results; however, some variations are inevitable. Understanding and accounting for these differences can help refine the design and testing processes. The discrepancies may be due to the idealized conditions in simulations compared to real-world factors. Enhancements in soldering techniques, accurate material characterization, manufacturing processes, and regular calibration can significantly reduce the gap between simulated and measured outcomes.

B. SAR Measurement

SAR measurements are essential for ensuring the security of wireless devices since they quantify the exposure to electromagnetic fields. Manufacturers may verify that their devices are safe and reliable in real-world environments by following standard operating procedures and using calibrated equipment, as illustrated in Fig. 10 at the Electronics Research Institute, central Lab. The proposed sensor's observed SAR values at 1.5, 3 GHz and 5.2 GHz are displayed in Tables 6. According to Table 6, the SAR values obtained for the suggested sensor at different power levels fall within the accepted safe limits as stated by the IEEE standard, which is 1.6 W/kg. As demonstrated in, the SAR is measured at various resonance frequencies and input powers. The SAR is measured at different resonant frequencies and input power as shown in Table 6.

C. Characterization of Glucose

The glucose concentration was prepared by dissolving the prescribed amounts of glucose in distilled water to yield concentrations of 50 mg/dl, 100 mg/dl, 200 mg/dl, and 350 mg/dl. Then, we entered the various concentrations into a DAK device to measure the properties of the glucose concentration, such as permittivity real value and loss tangent over the operating band as shown in Fig. 10.

First, an inverse relationship between the magnitude of the dielectric constant, and the value of glucose level are discovered. When the value of glucose concentration level is large as 350 mg/dl, the real value of the dielectric constant is less than the value of the low glucose level as 0 mg/dl. As well as the variation of the permittivity within the operating band from 3 to 4 GHz and from 5 to 6 GHz are larger than other measured operating bands. Second, on the contrary, the imaginary part of glucose concentration level increased with the value glucose level.

D. Measurements with Glucose

This part contains comparable methodologies to the simulation using the human finger and a container. The container was employed at specific concentration levels during characterization. that a glucometer was used to measure this concentration.

i. Measurements with Glucose Solutions with Container

A recommended reflection sensor's performance is measured and given. A dielectric container is attached to the prototype sensor's top in order to release the liquid into the sensing area. A plastic container with a ε_r = 2.2 component is placed symmetrically over the sensor region using a Rohde & Schwarz VNA, as seen in Fig. 11(a). The container's base measures 12.57× 2 × 1 mm³. Using a pipette, various concentrations of glucose (0,100, 150, and 350 mg/dL) are administered; the same amount is used for each concentration. The results of different concentrations of the transmission and reflection coefficients are shown in Fig. 11. By analysing the shifts in the frequency, magnitude, and phase of the S-parameter, as shown in Fig. 11, The difference between the glucose concentration in the container are determined by measurements the difference of S-parameters, the highest shift in frequency, magnitude and phase at 350 Mg/dL concentration.

ii. Measurements with Finger

In the second case, using finger tips in measurements are validated. According to recent research, finger measurements might be a useful technique for determining how various blood levels interact with one another. Its appendages are meant to be used by humans. Approval from an Ethical Perspective NILES-EC-CU 24/1/2, issued on August 2024 has been obtained. The measuring process requires participants (healthy & diabetic participants) to fast for six hours prior to the test. After checking their glucose levels, the participants are re-tested after two hours of eating as a second round of testing. Glucose levels of participants are expected to reach higher values compared to prior status while fasting. Before the test began, the participants washed their hands to remove any contaminants, and they are given around fifteen minutes to relax until their bodies stabilized.

In order to get two normal blood glucose levels one for each of the two male and female cases and one for the diabetic case the glucose concentration is measured using the Nano VNA during fasting and after eating for a total of four cases. the concentration of glucose in the blood is initially measured using glucometer as well as the Nano VNA after calibration as shown in Fig. 12. The magnitude and phase of S-parameters are recorded at the four cases and presented in Fig. 12. The amount of glucose in the blood is first measured with a glucometer as shown in Table 7.

As shown in Fig. 12, after measuring the four distinct cases and comparing the results after and before fasting, we can detect the difference between the measurement made after and before fasting by examining the shift and frequency changes. Next section, will be discussed these differences of results.

Table 7

Different case study before and after eating
Case Number	Gender	Glucose level during fasting	Glucose level after eating
1	Female	86	93
2	Male	86	336
3	Male	160	231
4	Female	105	244

The frequency band is varied and tested throughout a range to see how variations in glucose concentration affect the impedance pattern utilized to compute the sensor's sensitivity. When blood is added, the interaction between the blood glucose samples and the fields modifies the resonant frequency for the relative permittivity ε_r and the sensitivity of the sensor. The variation of the glucose concentration are changed according to the resonant frequency of operation as shown in Fig. 11. The inter operating band as 3.1 GHz and 5.95 GHz are more sensitive frequency band to the variation of the blood glucose concentration. The following equation in (2) is used to analyse the sensitivity; it shows that any change in the sensor's resonant frequency (Δ f) or magnitude (ΔS₁₁) depends on a change in the glucose concentration (ΔC). One may extract the equations as [33–34].

$$\:{\text{S}}_{fr}=\frac{\varDelta\:F}{\varDelta\:\text{C}}\:\text{M}\text{H}\text{z}/\text{m}\text{g}/\text{d}\text{l}$$

1

$$\:{S}_{dB}=\frac{\varDelta\:S}{\varDelta\:\text{C}}\:\text{d}\text{B}/\text{m}\text{g}/\text{d}\text{l}$$

2

$$\:{\text{S}}_{Ph}=\frac{\varDelta\:{\upvarphi\:}}{\varDelta\:\text{C}}\:\text{d}\text{e}\text{g}\text{r}\text{e}\text{e}/{mgdL}^{-1}$$

3

The sensitivity of the measurement by using container is around 0.015 MHz/mg/dL, 0.003 dB/mg/dL and 0.003degree//mg/dL at resonant 3.1 GHz 4.624. while at resonant frequency 5.9 GHz the sensitivity is 0.5 MHz/mg/dL, 0.005 dB/mg/dL and 0.05degree//mg/dL. While the sensitivity of the measurement by using human finger, the value of sensitivity is around to 0.299 dB/mg/dL for case 1 and 1.859 dB/mg/dL for case 2 as well as the sensitivity is around 0.05951degree/mg/dL for case 1 and 0.510473degree/mg/dl for case 2.

Table 8

The Outcomes of the Four Cases by Using Aforementioned Equations.
	(Case 1)	(Case 2)	(Case 3)	(Case 4)
S_fr $\:\:(\mathbf{M}\mathbf{H}\mathbf{z}/\mathbf{m}\mathbf{g}{\varvec{d}\varvec{L}}^{-1})$	0.2427	24.624	3.4199	1.5387
S_dB $\:(\mathbf{d}\mathbf{B}/{\varvec{m}\varvec{g}\varvec{d}\varvec{L}}^{-1}$)	0.299	1.859	0.012344	0.128014
S_ph ( $\:\mathbf{d}\mathbf{e}\mathbf{g}\mathbf{r}\mathbf{e}\mathbf{e}/{\varvec{m}\varvec{g}\varvec{d}\varvec{L}}^{-1}$ )	0.05951	0.510473	0.895$\:\times\:{10}^{-3}$	3.887$\:\times\:{10}^{-3}$

Moreover, the sensitivity is around 0.2427 MHz/mg/dL for case 1 and 24.624 MHz/mg/dl for case 2, while, the value of sensitivity is around to 0.299 dB/mg/dL for case 1 and 1.859 dB/mg/dL for case 2 as well as the sensitivity is around 0.05951degree/mg/dL for case 1 and 0.510473degree/mg/dl for case 2. The resonant frequency at 172.37 MHz decreases with increasing glucose concentrations between 3.280 and 3.1080 GHz. The frequency moves somewhat downward at lower levels and drastically lowers at higher quantities as shown in Table 8. This alteration causes relative permittivity to drop and sensor sensitivity to increase with glucose concentrations. Table 9 compares the data collected by using the newly developed microwave antenna sensor with previous published before. In order to compare the results of this research with those of previous studies. Table 9 shows the proposed sensor is more compact and good sensitivity than other.


Measurement Technique	Ref	Area (mm²)	Human Trial	Concentration (mg/dL)	Resonant Freq. (GHz)	S (MHz per mg/dL)
Inverted-F Antenna	[35]	38 × 38	NM	0-530	0.53	3.54
CSRR resonator	[36]	59 × 20	NM	70–150	1–6	67 − 11
Microstrip Patch Antenna-Sensor	[37]	42.97 ×43.60	NM	0-400	2.4	25
Distributed MEMS transmission lines (DMTL)	[38]	NM	No.	0-347.8	16	16.4
Single-port Sensor	[39]	55 × 30	No.	100–1000	4.8	14
MSRRMA	Our work	35×50	Yes	0-350	0.5-7	24.6
NM : Not Mention

In future work of our team, to mitigate the impact of sweat, movement, temperature, blood type, and finger pressure on measurement accuracy, it may be necessary to incorporate compensation techniques. These could include signal processing algorithms that account for moisture and other interfering factors, or the development of sensors capable of distinguishing between changes in dielectric properties caused by glucose and those caused by other factors. Without such measures, these interferences can introduce noise into the measurement and reduce the sensor's ability to accurately detect glucose levels. Additionally, this approach involves designing and fabricating a custom packaging case using 3D printing technology, providing the flexibility needed to optimize sensor performance. Since finger size and shape vary, the sensors may require calibration for different finger sizes or the ability to adapt to a wide range of tissue volumes. Advanced signal processing techniques or machine learning algorithms could be employed to account for finger size variations and enhance measurement accuracy across different users. Furthermore, the possibility of integrating the microwave sensor with other non-invasive monitoring technologies, such as optical sensors, could be explored to provide complementary data and further improve overall accuracy.

In this paper, we presented the MSRRMA sensor for non-invasive glucose level measurement via the fingertip, operating at 0.94, 1.5, 3, and 4.6 GHz, with a sensor size reduction of approximately 57.5%. To simulate realistic conditions, a finger phantom model comprising five layers—skin, fat, muscle, blood, and bone—was developed in CST EM simulations. Each layer had variable dielectric properties, with blood's dielectric constant varying according to glucose concentration. The phantom was positioned at different locations along the antenna’s radiating element, with glucose concentrations ranging from 0 to 350 mg/dL. A sensitivity of up to 24 MHz/mg/dL was achieved when the phantom was placed directly above the antenna. Additionally, the Specific Absorption Rate (SAR) remained below the safety limit of 1.6 W/kg at a maximum power output of 20 dBm. The developed prototype was tested on volunteers, including both healthy and diabetic individuals, under fasting conditions and after eating. Significant variations in glucose levels were observed between the simulations and real-world measurements, demonstrating the accuracy of the phantom model in mimicking a human finger. With its compact size, high sensitivity, and cost-effectiveness, the proposed antenna-based sensor proves to be a promising solution for non-invasive blood glucose monitoring.

All authors declare that the submitted research paper is my original work and no part of it has been published anywhere else in the past.
Ethical Approval: NILES-EC-CU 24/1/2

Conflict of Interest: There are no relevant financial interests in the manuscript and no
other potential conflicts of interest to disclose.
Competing Interest: The authors declare no competing interests.
Funding: Not Applicable.
Data Availability: The data that may be required to reproduce the findings of this study
are available upon reasonable request from the corresponding author, who is indicated
with an asterisk (*).

Megdad: Ayman R and Aldhaheri:Sobahi and W. Rabah , "A noninvasive method for measuring the blood glucose level using a narrow band microstrip antenna," Applied Computational Electromagnetics Society Journal, vol. 37, no. 11, p. 1118, 2022.
Hill : Gleadle: Pulvi, M. Jonathan , E. Kathleen, Mariastella :McNaughton and A. Darlene, "The social determinants of health for people with type 1 diabetes that progress to end-stage renal disease," Health Expectations, vol. 18, no. 6, pp. 2513--2521, 2015.
Wilbert : Mobashshe, T. Ahmed, . G. Villena and Amien, "The progress of glucose monitoring—A review of invasive to minimally and non-invasive techniques, devices and sensors," Sensors, vol. 19, no. 4, p. 800, 2019.
Juan: Bronchalo: Enrique, G. Carlos , Potelon: Benjamin, Potelon, Quendo and Benjamin, "Concentration measurement of microliter," IEEE Transactions on, vol. 68, no. 7, pp. 2621-2634, 2019.
Mansour, Esraa, Allam, Ahmed , Rahman and Adel, "A novel approach to non-invasive blood glucose sensing based on a defected ground structure," 2021 15th European Conference on Antennas and Propagation (EuCAP), pp. 1-5, 2021.
N. ,. Anis: Siti, "A portable non-invasive blood glucose monitoring device with IoT," Evolution in Electrical and Electronic Engineering, vol. 2, no. 1, pp. 36-44, 2021.
S. George, "Non-invasive monitoring of glucose," International Journal of Mobile Human Computer, vol. 10, no. 3, pp. 10-29, 2018.
A. B. :. b. M. S. W. H. :. B. :. S. Abd Salam:Nurul, "The evolution of non-invasive blood glucose monitoring system for personal application," Journal of Telecommunication, Electronic and Computer Engineering (JTEC), vol. 8, no. 1, pp. 59-65, 2016.
M. Yadav: Jyoti:Rani: Asha:Singh: Vijander: Murari: Bhaskar, "Prospects and limitations of non-invasive blood glucose monitoring using near-infrared spectroscopy," Biomedical signal processing and control, vol. 12, pp. 214--227, 2015.
P. C. Iris, Jessica, V. Cuevas, Alonso, Silverio and A. Gustavo, "An IoT-based non-invasive glucose level monitoring system using raspberry pi," Applied Sciences, vol. 9, no. 15, p. 3046, 2019.
Mohamed, MS, Hameed, F. Mohamed, Areed, F. Nihal, El-Okr, MM, Obayya and SSA, "Analysis of highly sensitive photonic crystal biosensor for glucose monitoring," Applied Computational Electromagnetics Society Journal, vol. 31, no. 7, p. 836, 2016.
Deshmukh, V. Vidya, Chorage and S. Suvarna, "Non-invasive determination of blood glucose level using narrowband microwave sensor," Journal of Ambient Intelligence and Humanized Computing, pp. 1-16, 2021.
A. Zahran, Y. A. Aly, A. R. Eldamak and D. N. Elsheakh, "Compact Complementary Split Ring Resonator Based Sensor for Non-Invasive Glucose monitoring," 2024 International Telecommunications Conference (ITC-Egypt), Cairo, Egypt, 2024, pp. 1-7, doi: 10.1109/ITC-Egypt61547.2024.10620499.
Satish, Sen, Kushal, Anand and Sneh, "Demonstration of microstrip sensor for the feasibility study of non-invasive blood-glucose sensing," Mapan, vol. 36, pp. 193-199, 2021.
Deshmukh, V. Vidya, Ghongade and B. Rajesh, "Measurement of dielectric properties of aqueous glucose using planar ring resonator," 2016 International conference on microelectronics, computing and communications (MicroCom), pp. 1-5, 2016.
H. a. G. I. a. S. I. a. S. Cano-Garcia, Shimul, Palikaras, George, Kosmas, Panagiotis, Kallos and Efthymios, "Reflection and transmission measurements using 60 GHz patch antennas in the presence of animal tissue for non-invasive glucose sensing," pp. 1-3, 2016.
Liu, W. Louis, Kandwal, Abhishek, Cheng, Qingsha, Shi, Hongjian, Tobore, Igbe, Nie and Zedong, "Non-invasive blood glucose monitoring using a curved Goubau line," Electronics, vol. 8, no. 6, p. 662, 2019.
Karacolak, Tutku, Moreland, C. Elaine, Topsakal and Erdem, "Cole--cole model for glucose-dependent dielectric properties of blood plasma for continuous glucose monitoring," Microwave and Optical Technology Letters, vol. 55, no. 5, pp. 1160-1164, 2013.
Karacolak, Tutku, Moreland, C. Elaine, Topsakal and Erdem, "Cole--cole model for glucose-dependent dielectric properties of blood plasma for continuous glucose monitoring," Microwave and Optical Technology Letters, vol. 50, no. 55, pp. 1160--1164, 2013.
Jiang, Fan, Li, Song, Yu, Yang, Cheng, Qingsha, S, Koziel and Slawomir, "Sensitivity optimization of antenna for non-invasive blood glucose monitoring," 2017 International Applied Computational Electromagnetics Society Symposium (ACES), pp. 1-2, 2017.
Li, Jingzhen, Tobore, Igbe, Liu, Kandwal, Abhishek, Wang, Lei, Nie and Zedong, "Non-invasive monitoring of three glucose ranges based on ECG by using DBSCAN-CNN," IEEE journal of biomedical and health informatics, vol. 25, no. 9, pp. 3340-3350, 2021.
I. M. a. v. d. Z. S. C. a. W. Atzeni, Johanna, Zwerver, Johannes, Smit, J. Andries, Mulder and J. Douwe, "The AGE Reader: A non-invasive method to assess long-term tissue damage," Methods, vol. 203, pp. 533-541, 2022.
Abedeen, Zainul, Agarwal and Poonam, "Microwave sensing technique based label-free and real-time planar glucose analyzer fabricated on FR4," Sensors and Actuators A: Physical, vol. 279, pp. 132-139, 2018.
M. M. a. S. S. a. S. P. A. Alam, M. Md, Saha, Swapnil, Saha, Proshib, Nur, Ferna, Narin, Moon, Nazmun, Nessa, Karim, Asif, Azam and Sami, "D-care: A non-invasive glucose measuring technique for monitoring diabetes patients," Proceedings of International Joint Conference on Computational Intelligence: IJCCI 2018, pp. 443-445, 2020.
Susana, Ernia, Ramli, Kalamullah, Murfi, Hendri and H. Apriantoro:Nursama, "Non-invasive classification of blood glucose level for early detection diabetes based on photoplethysmography signal," Information, vol. 13, no. 2, p. 59, 2022.
Yilmaz, Tuba, Ozturk, Tugce, Joof and Saul, "A comparative study for development of microwave glucose sensors," Proceedings of the 32nd URSI GASS, Montreal, QC, Canada, pp. 19-29, 2017.
Kim, Seungwan, Melikyan, Harutyun, Kim, Jongchel, Babajanyan, Arsen, Lee, Jung-Ha, Enkhtur, Lkhamsuren, Friedman, Barry, Lee and Kiejin, "Noninvasive in vitro measurement of pig-blood d-glucose by using a microwave cavity sensor," Diabetes research and clinical practice, vol. 96, no. 3, pp. 379-384, 2012.
Kim, Seungwan, Melikyan, Harutyun, Kim, Jongchel, Babajanyan, Arsen, Lee, Jung-Ha and Enkhtur, "Noninvasive in vitro measurement of pig-blood d-glucose by using a microwave cavity sensor," Diabetes research and clinical practice, vol. 96, no. 3, pp. 379--384, 2012.
Vrba, Jan, Vrba and David, "Metamaterial sensor for microwave non-invasive blood glucose monitoring," {World Congress on Medical Physics and Biomedical Engineering 2018: June 3-8, 2018, Prague, Czech Republic (Vol. 3), vol. 3, pp. 789--792, 2019.
Khan, N. Ahsan, Cha, Young-Ok, Giddens, Henry, Hao and Yang, "Recent advances in organ specific wireless bioelectronic devices: Perspective on biotelemetry and power transfer using antenna systems," Engineering, vol. 11, pp. 27-41, 2022.
A. Zahran, Y. Ayman, A. R. Eldamak and D. N. Elsheakh, "Microwave Antenna Based For Sensing Glucose Changes In Blood," 2024 41st National Radio Science Conference (NRSC), New Damietta, Egypt, 2024, pp. 59-65, doi: 10.1109/NRSC61581.2024.10510521.
Megdad, R. Ayman, Aldhaheri, W. Rabah, Sobahi and M. Nebras, " ANoninvasive Method for Measuring the Blood Glucose Level Using," Applied Computational Electromagnetics Society Journal, vol. 37, no. 11, p. 1118, 2022.
Marzouk, M. Hala, Hameed, Anwer, S, A. El-, Allam, Ahmed, Pokharel, K. Ramesh, Rahman and B. Adel, " Anewrectangular dielectric resonator sensor for glucose," International Journal of Circuit Theory and Applications, 2023.
Deshmukh, V. Vidya, Chorage and S. Suvarna, "Non-invasive determination of blood glucose level using narrowband microwave sensor," Journal of Ambient Intelligence and Humanized Computing, pp. 1-16, 2021.
Nandgaonkar, R, A and Khadase, "Design of implantable MSA for glucose monitoring," International Conference on Communication and Signal Processing 2016 (ICCASP 2016), 2016.
Omer, E. Ala, Shaker, George, Safavi, Naeini, Safieddin, Ngo, Kieu, Shubair and M. Raed, "Multiple-cell microfluidic dielectric resonator for liquid sensing applications," {IEEE Sensors Journal, vol. 21, no. 5, pp. 6094-6104, 2020.
Satish, Sen, Kushal, Anand and Sneh, "emonstration of microstrip sensor for the feasibility study of non-invasive blood-glucose sensing," Mapan, vol. 36, pp. 193-199, 2021.
Li, Lijie, Uttamchandani and Deepak, "A microwave dielectric biosensor based on suspended distributed MEMS transmission lines," IEEE Sensors Journal, vol. 9, no. 12, pp. 1825-1830, 2009.
Turgul, Volkan, Kale and Izzet, "Simulating the effects of skin thickness and fingerprints to highlight problems with non-invasive RF blood glucose sensing from fingertips," IEEE sensors journal, vol. 17, no. 22, pp. 7553--7560, 2017.

Table 4 and 6 are available in the Supplementary Files section.

No competing interests reported.

Table6and4.docx

Innovative Non-Invasive Blood Glucose Monitoring by Using Multiband Split Ring Resonator Monopole Antenna (MSRRMA)

Status:

Version 1

Abstract

Figures

I. Introduction

II. Antenna configuration And Design

III. Glucose Characterization

IV. Sensor Measurement Results

V. Discussion

VI. CONCLUSION

Declarations

References

Table 4 and 6

Additional Declarations

Supplementary Files

Status:

Version 1

Table 2: The Electrical Properties of Glucose Levels at 2.45 GHz.
Glucose Concentration (mg/dL)	\(\:{\varvec{\epsilon\:}}_{\varvec{r}}\)	\(\:\varvec{\sigma\:}\)(S/m)
0	69.8470	3.4518
100	69.8203	3.4405
150	69.8070	3.4350
350	69.7550	3.4130

	(Case 1)	(Case 2)	(Case 3)	(Case 4)
S_fr \(\:\:(\mathbf{M}\mathbf{H}\mathbf{z}/\mathbf{m}\mathbf{g}{\varvec{d}\varvec{L}}^{-1})\)	0.2427	24.624	3.4199	1.5387
S_dB \(\:(\mathbf{d}\mathbf{B}/{\varvec{m}\varvec{g}\varvec{d}\varvec{L}}^{-1}\))	0.299	1.859	0.012344	0.128014
S_ph ( \(\:\mathbf{d}\mathbf{e}\mathbf{g}\mathbf{r}\mathbf{e}\mathbf{e}/{\varvec{m}\varvec{g}\varvec{d}\varvec{L}}^{-1}\) )	0.05951	0.510473	0.895\(\:\times\:{10}^{-3}\)	3.887\(\:\times\:{10}^{-3}\)