Optical Coherence Tomography and Optical Polarimetry for Non-Invasive Glucose Monitoring: Principles, Recent Advances, Research Gaps, and Future Directions

Abdullah Al Maimun; Sirajam Munira; Md Mahbubur Rahman Akash

doi:10.20944/preprints202606.1368.v1

Submitted:

16 June 2026

Posted:

17 June 2026

You are already at the latest version

Abstract

Diabetes mellitus is a major global public health challenge, with prevalence rising due to aging populations, urbanization, sedentary lifestyles, and dietary changes. Effective glucose monitoring is essential for diagnosis, treatment, and long-term disease management, driving significant research into improved sensing technologies. Conventional invasive and minimally invasive glucose monitoring methods often cause discomfort and reduce patient compliance, motivating the development of non-invasive alternatives. Recent advances in photonics, biomedical engineering, nanotechnology, wearable devices, and artificial intelligence have accelerated the emergence of innovative glucose sensing approaches capable of improving comfort, safety, and monitoring frequency. This review presents a comprehensive overview of recent non-invasive glucose monitoring technologies reported in the literature, including optical, electromagnetic, nanotechnology-based, and physiological sensing methods evaluated through human studies, biological samples, or tissue-equivalent models. The underlying sensing principles, measurement sites, performance characteristics, and practical implementation challenges of these technologies are discussed. Particular attention is given to the integration of machine learning algorithms, which have demonstrated significant potential for enhancing glucose prediction accuracy and supporting real-time monitoring applications. The review also critically examines the advantages, limitations, clinical feasibility, and commercialization prospects of existing technologies, highlighting the key barriers that continue to impede widespread adoption. By consolidating recent developments across multiple scientific and engineering disciplines, this work provides researchers, clinicians, and technology developers with a concise assessment of the current state of non-invasive glucose sensing and identifies future research directions necessary for advancing reliable, accurate, and user-friendly next-generation diabetes management systems.

Keywords:

non-invasive glucose sensing

;

optical coherence tomography (OCT)

;

optical polarimetry (OP)

;

diabetes mellitus

;

machine learning

;

optical biosensors

;

continuous glucose monitoring

Subject:

Engineering - Bioengineering

I. Introduction

Diabetes mellitus (DM) is a chronic metabolic disorder characterized by elevated blood glucose levels resulting from abnormalities in insulin production, insulin action, or both. The prevalence of DM has increased substantially worldwide and continues to pose a significant public health challenge [1]. According to recent global estimates, hundreds of millions of individuals are currently affected by diabetes, making it one of the leading causes of morbidity and mortality worldwide [2,3]. Furthermore, the global diabetic population is projected to rise considerably in the coming decades, driven by factors such as population aging, urbanization, sedentary lifestyles, obesity, and dietary changes [4]. Despite advances in medical care, effective diabetes diagnosis and management remain challenging due to limitations associated with existing glucose monitoring techniques, including patient discomfort, limited compliance, and measurement inaccuracies [5,6,7].

The diagnosis and monitoring of diabetes currently rely primarily on invasive glucose sensing technologies. The specific diagnostic approach varies depending on the type of diabetes and patient characteristics. Type 1 diabetes mellitus, which is more frequently diagnosed during childhood and adolescence, is commonly identified through random blood glucose (RBG) testing and glycated hemoglobin (HbA1c) measurements. Additional confirmation may be obtained through the detection of autoimmune biomarkers, including islet-cell antibodies (ICA), insulinoma-associated protein-2 (IA-2), glutamic acid decarboxylase 65 (GAD65), and zinc transporter 8 (ZnT8). Long-term management of Type 1 diabetes typically involves self-blood glucose monitoring (SBGM), continuous glucose monitoring (CGM) systems, and insulin administration through injections, pens, or insulin pumps to maintain blood glucose levels within the recommended range.

Gestational diabetes mellitus (GDM) is generally diagnosed using a 75-g oral glucose tolerance test (OGTT) according to established clinical guidelines. Following diagnosis, glucose levels are monitored regularly through self-monitoring practices while patients are encouraged to adopt dietary modifications and increased physical activity. Pharmacological interventions, including insulin therapy or metformin, may be introduced when lifestyle-based approaches alone are insufficient to achieve adequate glycemic control.

Figure 1. Overview of diabetes mellitus types, diagnostic methods, and management strategies. Type 1, gestational, and Type 2 diabetes are shown with their corresponding diagnostic criteria and treatment approaches. ICA Islet Cell Autoantibodies; IA-2 Insulinoma-Associated Protein 2; GAD65Glutamic Acid Decarboxylase 65; ZnT8 Zinc Transporter 8; CGM Continuous Glucose Monitoring; SBGM Self-Blood Glucose Monitoring; FBG Fasting Blood Glucose; RBG Random Blood Glucose; HbA1cGlycated Hemoglobin; OGTT Oral Glucose Tolerance Test.

Type 2 diabetes mellitus, the most prevalent form of diabetes worldwide, is commonly diagnosed using fasting blood glucose (FBG), random blood glucose (RBG), glycated hemoglobin (HbA1c), or oral glucose tolerance testing [8]. Management strategies generally begin with lifestyle interventions such as dietary modification, weight reduction, and regular physical activity. When these measures fail to maintain target glucose levels, oral antihyperglycemic medications and, in some cases, insulin therapy are prescribed. Although these approaches have significantly improved diabetes management, their invasive nature and dependence on repeated blood sampling continue to motivate the development of accurate, convenient, and patient-friendly non-invasive glucose monitoring technologies.

Despite significant advances in diabetes management, currently available glucose monitoring approaches remain limited in terms of sensitivity, specificity, patient comfort, and long-term usability. Although Continuous Glucose Monitoring (CGM) systems have improved glycemic assessment, particularly for individuals with Type 1 diabetes, many conventional diagnostic and monitoring methods still rely on intermittent measurements that provide only a limited representation of an individual’s glucose fluctuations throughout the day [9]. Several studies have highlighted the shortcomings of single-point glucose measurements, including fasting blood glucose, random blood glucose, and oral glucose tolerance tests, for accurately capturing dynamic glycemic variations [10]. Similar concerns have been reported for self-blood glucose monitoring (SBGM), which depends on a limited number of daily measurements and may fail to detect significant glucose excursions occurring between testing intervals [3,4,5,6,7,11].

To address these limitations, considerable research efforts have focused on the development of non-invasive glucose sensing technologies (NIGSTs) capable of providing accurate, convenient, and continuous glucose monitoring. This review presents a comprehensive assessment of recently developed and emerging NIGSTs, examining their underlying sensing principles, operational mechanisms, advantages, limitations, and clinical applicability. In addition, the review discusses the design methodologies, technological developments, and translational challenges associated with the development of next-generation glucose sensing systems.

Previous reviews have broadly classified glucose sensing approaches into invasive, minimally invasive, and non-invasive categories [12,13,14,15]. While minimally invasive technologies often employ enzyme-based sensing mechanisms, their performance can be affected by environmental conditions such as temperature, humidity, and pH variations, which may compromise stability and long-term monitoring performance [16,17,18]. Therefore, this review focuses specifically on non-invasive glucose sensing technologies and categorizes them into four major groups: optical, nanotechnology-based, electrical/electromagnetic, and physiological sensing approaches (Figure 2). Furthermore, commercially available devices, emerging prototypes, and technologies currently undergoing clinical evaluation are critically discussed to provide a comprehensive overview of the current state and prospects of non-invasive glucose monitoring.

II. Methods

A comprehensive literature survey was conducted to identify and evaluate recent advances in non-invasive glucose sensing technologies (NIGSTs), including optical, nanotechnology-based, electrical/electromagnetic, and physiological sensing approaches. Relevant publications were retrieved from major scientific databases using combinations of keywords related to non-invasive glucose monitoring, optical sensing, electromagnetic sensing, nanotechnology, physiological sensing, and machine learning-assisted glucose detection. The search focused primarily on studies published between 2018 and 2023.

To ensure comprehensive coverage, backward and forward citation tracking was performed by examining the reference lists of selected articles as well as subsequently published studies that cited these works. All identified publications were screened for relevance based on their contribution to non-invasive glucose monitoring, sensing methodology, experimental validation, and potential clinical applicability. Only studies directly related to glucose sensing technologies were included in the final review.

The selected literature was categorized into four major groups: optical sensing technologies, nanotechnology-based sensing approaches, electrical/electromagnetic techniques, and physiological sensing methods. Although nanotechnology is often treated as an independent category, considerable overlap exists between nanotechnology and optical sensing platforms. Therefore, nanoplasmonic approaches, including Surface-Enhanced Raman Scattering (SERS), Surface Plasmon Resonance (SPR), Plasmon-Enhanced Fluorescence (PEF), and Carbon Quantum Dot (CQD)-based fluorescence sensing, are discussed within the broader context of advanced optical sensing technologies.

Each technology was critically evaluated based on its sensing principle, operating mechanism, performance characteristics, advantages, limitations, clinical feasibility, and commercialization potential. In addition, emerging applications of machine learning and artificial intelligence for signal processing, feature extraction, calibration, and glucose prediction were examined. Commercial products, prototype devices, and technologies currently undergoing clinical validation were also reviewed to provide a comprehensive overview of the current state of non-invasive glucose monitoring.

III. Skin Layers with Properties

The effectiveness of non-invasive glucose sensing technologies (NIGSTs) largely depends on their interaction with biological tissues, particularly the skin. Human skin consists of three primary tissue layers the epidermis, dermis, and hypodermis, which exhibit distinct optical, electrical, and electromagnetic characteristics that influence the propagation, penetration depth, and absorption of incident signals [19,20]. These tissue-specific properties govern how different sensing modalities interact with biological media and determine their ability to detect physiological biomarkers, including glucose, water, proteins, lipids, and electrolytes. A schematic representation of the skin layers and their corresponding thickness is presented in Figure 3.

Most non-invasive glucose sensing approaches estimate blood glucose concentration indirectly by analyzing changes in optical or electromagnetic signals. Consequently, the selection of operating frequency, wavelength, and spectral band plays a critical role in determining sensing performance. The frequency ranges and wavelength bands associated with various sensing techniques are summarized in Table 1.

Both optical and electromagnetic sensing methods employ non-ionizing radiation; however, their interaction mechanisms and penetration depths differ significantly. Optical techniques operating in the terahertz (THz) region primarily interact with superficial skin layers, particularly the epidermis, where signal penetration is influenced by factors such as wavelength, polarization state, tissue composition, and optical power density [21]. In contrast, lower-frequency electromagnetic techniques, including radio-frequency (RF), microwave (MW), and millimeter-wave (mmW) sensing, can penetrate deeper into subcutaneous tissues. This enhanced penetration enables the collection of information related to glucose distribution and tissue composition, although measurement accuracy may be affected by the presence of other biological constituents and the heterogeneous nature of skin tissues.

When optical or electromagnetic waves propagate through biological tissues, four fundamental interaction mechanisms may occur: reflection, scattering, absorption, and transmission [22]. These interactions form the basis of most non-invasive glucose sensing technologies and determine how physiological information can be extracted from biological media.

In reflection-based sensing, incident light or electromagnetic waves are directed toward the tissue surface and the reflected signal is analyzed. Variations in the reflected intensity, phase, or spectral characteristics can provide information about tissue composition and physiological parameters. Scattering-based techniques rely on the interaction of incident waves with cellular structures and tissue components, causing the waves to deviate from their original propagation paths. The resulting scattered signals contain valuable information regarding tissue morphology and biochemical composition.

Absorption-based methods exploit the wavelength-dependent absorption characteristics of biological molecules. When light or electromagnetic radiation passes through tissue, specific constituents absorb energy at characteristic wavelengths. By quantifying the absorbed energy, the concentration of target analytes, including glucose, can be estimated. In transmission-based approaches, light or electromagnetic waves traverse the tissue and the transmitted signal is measured. During propagation, interactions with biological molecules modify the signal through absorption and scattering processes, enabling the extraction of physiological information from the transmitted spectrum [20].

The analysis of these interaction mechanisms enables the assessment of tissue properties and the estimation of glucose concentration within biological systems. The dominant interaction depends on several factors, including the operating wavelength, tissue characteristics, penetration depth, and sensing modality employed. Table 2 summarizes the four primary wave–tissue interaction mechanisms, the information obtained from each interaction, representative glucose sensing techniques, and their associated challenges.

IV. Techniques

Optical and Nanotechnology-Based Glucose Sensing Techniques

Optical sensing technologies represent one of the most extensively investigated approaches for non-invasive glucose monitoring due to their high sensitivity, rapid response, and potential for continuous measurement. These techniques rely on the interaction of light with biological tissues, particles, and molecular constituents, where incident optical signals may undergo reflection, scattering, absorption, or transmission. The resulting changes in optical properties can be analyzed to directly or indirectly estimate glucose concentration.

The underlying principles of optical glucose sensing are governed by several fundamental physical laws. Reflection-based techniques are derived from the law of reflection, whereby the angle of reflected light is related to the angle of incidence. Scattering phenomena are commonly described by Rayleigh scattering theory, which states that shorter wavelengths experience greater scattering than longer wavelengths, resulting in wavelength-dependent variations in signal intensity. Absorption and transmission mechanisms are primarily governed by the Beer–Lambert law, which relates the attenuation of light passing through a medium to the concentration of absorbing molecules, including glucose [23].

A major advantage of optical sensing methods is their capability to support real-time and continuous glucose monitoring without the need for repeated blood sampling [18]. Consequently, considerable research efforts have been devoted to the development of advanced optical sensing platforms. The principal optical techniques discussed in this review include Optical Coherence Tomography (OCT), Optical Polarimetry (OP), Photoplethysmography (PPG), and various spectroscopic approaches. The spectroscopic methods encompass Near-Infrared (NIR), Mid-Infrared (MIR), Far-Infrared (FIR), combined NIR/MIR spectroscopy, diffuse reflectance spectroscopy, scattering and occlusion spectroscopy, and photoacoustic spectroscopy. Furthermore, recent advances in nanotechnology have enabled the integration of plasmonic nanostructures, nanomaterials, and nanoscale optical components, leading to enhanced sensing performance, improved sensitivity, and greater selectivity for non-invasive glucose detection.

Optical Coherence Tomography (OCT)

Optical Coherence Tomography (OCT) is a high-resolution, non-invasive imaging technique that was initially developed for ophthalmic diagnostics and has subsequently been extended to a wide range of biomedical applications, including skin imaging and glucose sensing [14,24,25]. The technique is based on low-coherence interferometry, in which backscattered and reflected light from different tissue layers is analyzed to generate depth-resolved structural information. By measuring variations in optical scattering and refractive index characteristics within biological tissues, OCT can provide valuable insights into tissue composition and physiological changes associated with glucose concentration. A schematic illustration of the OCT operating principle is presented in Figure 4.

In OCT-based glucose sensing, near-infrared light reflected and backscattered from biological tissues is analyzed to reconstruct high-resolution images of tissue microstructures. Variations in glucose concentration influence tissue optical properties, including refractive index and optical rotation, enabling indirect estimation of glucose levels through OCT signal analysis [26]. Previous studies have demonstrated a strong correlation between changes in OCT signal characteristics and blood glucose concentration, highlighting the potential of OCT as a non-invasive monitoring modality [18].

To enhance glucose measurement accuracy, several signal-processing and hardware optimization strategies have been proposed. For example [27] employed a differential Mueller matrix model to evaluate polarization-related parameters, including optical rotation and depolarization characteristics of incident light. Similarly [28] introduced a low-magnification optical coherence tomography (LM-OCT) configuration that improved measurement reliability by optimizing imaging conditions and reducing sensitivity to spatial variations within the tissue.

OCT offers several advantages, including high spatial resolution, excellent signal-to-noise ratio (SNR), and reduced susceptibility to physiological factors such as heart rate, blood pressure fluctuations, osmolyte concentration, and red blood cell variability [25,28]. Nevertheless, several challenges continue to limit its widespread adoption for continuous glucose monitoring. These include relatively low glucose sensitivity, motion-induced artifacts, sensitivity to skin temperature variations, and difficulties in achieving precise quantitative glucose estimation under real-world conditions. Consequently, further research is required to improve system robustness, enhance sensitivity, and develop advanced signal-processing techniques capable of accurately extracting glucose concentration from OCT measurements [25,28].

2.: Optical Polarimetry (OP)

Optical Polarimetry (OP) is a non-invasive sensing technique that exploits the optically active nature of glucose molecules. Glucose possesses the ability to rotate the plane of polarized light by an amount proportional to its concentration, making polarization-based measurements a promising approach for glucose detection [29,30]. In OP systems, polarized light is directed toward biological media, most commonly the aqueous humor of the eye, and the resulting changes in polarization state, optical rotation, and light absorption are measured using photodetectors. These optical changes can then be correlated with glucose concentration to enable non-invasive glucose estimation.

A conceptual illustration of a wearable OP-based continuous glucose monitoring device is presented in Figure 5(a). Among recent developments, Hwang et al. [13] proposed a non-contact glucometer consisting of a polarized light source, beam splitters, photodetectors, and a signal-processing unit. In their experimental study, polarized light with a wavelength of 1650 nm and an optical power of 5 mW was directed into the eyes of four rabbits. By analyzing the polarization rotation and absorption characteristics of the reflected light and comparing the results with serum glucose measurements, the authors reported strong agreement between the optical measurements and glucose concentrations. The system achieved mean measurement differences of approximately 8 mg/dL for in vitro experiments and 29.2 mg/dL for in vivo evaluations.

Optical polarimetry offers several advantages, including high specificity to glucose and the relatively low concentration of proteins and red blood cells within the aqueous humor, which reduces potential optical interference. However, practical implementation remains challenging due to the stringent alignment requirements, susceptibility to physiological motion, and safety concerns associated with prolonged ocular illumination. In particular, the use of high-intensity optical sources and photothermal detection mechanisms may increase the risk of retinal damage, highlighting the need for further research focused on improving measurement safety, sensitivity, and long-term clinical applicability [13].

Further advancements in optical polarimetry have focused on improving measurement accuracy, reducing device size, and enhancing practicality for wearable applications. Li et al. [31] developed a compact and cost-effective glucose sensing prototype capable of estimating glucose concentration through the analysis of weak optical rotation signals generated within palm and finger tissues, as illustrated in Figure 5(b). The system incorporated a Gradient Boosted Trees Regression model to process polarization signals acquired at multiple wavelengths and light intensities, thereby reducing the effects of inter-subject variability and improving prediction performance. The results demonstrated the feasibility of performing optical polarimetric glucose measurements through skin tissue, highlighting the potential of wearable OP-based sensing platforms.

Optical polarimetry offers several attractive features, including high measurement accuracy, excellent spatial resolution, and the potential for miniaturized sensor implementations suitable for portable and wearable healthcare devices [13,14,25]. However, the technique faces significant challenges when detecting low glucose concentrations because the resulting optical rotation signals are often extremely weak. To address this limitation, [32] proposed a Spatial Polarization Modulation System (SPMS) that employs a rotating polarizer and an optical phase retarder to amplify weak polarization signals. By analyzing a single digital image, the system was able to improve sensitivity and achieve glucose measurements with a reported resolution of approximately 100 mg/dL.

Another important challenge arises from the presence of interfering biological constituents within interstitial fluid. In particular, albumin can exhibit optical properties that overlap with glucose-induced polarization effects, potentially reducing measurement accuracy. To overcome this issue,[33] developed a broadband polarimetric sensing platform integrated with a Partial Least Squares (PLS) regression algorithm capable of distinguishing glucose signals from albumin-related optical responses. The proposed approach demonstrated improved prediction accuracy and enhanced robustness in complex biological environments.

Despite these advances, several obstacles continue to hinder the widespread adoption of optical polarimetry for continuous glucose monitoring. These include physiological lag times between blood and interstitial glucose levels, sensitivity to body motion, temperature fluctuations, pH variations, and interference from other optically active compounds present in biological tissues [25]. Therefore, further research is required to improve signal stability, enhance sensitivity under low-glucose conditions, and develop robust compensation algorithms capable of maintaining measurement accuracy under real-world operating conditions.

V. Research Gaps and Future Directions

Despite substantial advances in non-invasive glucose sensing technologies, significant scientific and engineering challenges remain before these systems can achieve widespread clinical adoption. Existing approaches, including Optical Coherence Tomography (OCT), Optical Polarimetry (OP), Photoplethysmography (PPG), Near-Infrared (NIR) spectroscopy, electromagnetic sensing, and nanotechnology-assisted platforms, have demonstrated promising results under controlled conditions; however, their accuracy and robustness remain insufficient for routine clinical implementation in many real-world scenarios [18,25].

One of the primary research gaps is the influence of physiological variability on sensing performance. Factors such as skin thickness, tissue hydration, blood perfusion, temperature, pigmentation, and body composition can significantly affect optical and electromagnetic signals, leading to variations in glucose prediction accuracy [20,25]. Furthermore, physiological delays between blood glucose and interstitial glucose concentrations introduce additional uncertainties that complicate continuous monitoring. Future research should focus on adaptive calibration methods and personalized sensing models capable of compensating for individual-specific physiological differences.

Signal interference from biological constituents remains another critical challenge. Water, proteins, lipids, electrolytes, and albumin often produce optical and electromagnetic responses that overlap with glucose signatures, reducing sensor selectivity and measurement accuracy [25,33]. This limitation is particularly evident in optical polarimetry and spectroscopy-based approaches. The development of advanced signal-processing techniques, selective sensing materials, and multimodal sensing architectures may help overcome these challenges and improve glucose-specific detection.

Although optical sensing techniques have emerged as some of the most promising candidates for non-invasive glucose monitoring, each technology possesses distinct limitations. OCT provides high spatial resolution and excellent signal-to-noise characteristics but suffers from relatively low glucose sensitivity and susceptibility to motion artifacts and temperature variations [25,28]. Optical polarimetry benefits from the intrinsic optical activity of glucose molecules; however, weak polarization signals, interference from other optically active compounds, and alignment challenges continue to limit performance [25,32]. Similarly, PPG and NIR spectroscopy have demonstrated encouraging results and favorable user acceptability, yet their accuracy remains highly dependent on physiological conditions and environmental factors [18].

Machine learning has recently emerged as a powerful tool for improving non-invasive glucose sensing performance. Several studies have demonstrated that algorithms such as Partial Least Squares (PLS) regression, Boosted Trees Regression, Random Forests, and other artificial intelligence approaches can improve glucose prediction by extracting complex patterns from multidimensional sensing data [31,33,36]. Nevertheless, most current models are trained using relatively small datasets collected under controlled laboratory conditions. Future investigations should prioritize large-scale clinical datasets, explainable artificial intelligence (XAI), transfer learning, and personalized machine learning frameworks capable of maintaining robust performance across diverse populations.

Another important gap concerns the development of wearable and continuous monitoring systems. While several prototype devices have been reported, including ear-based, fingertip-based, and wrist-worn sensing platforms, few technologies have demonstrated the long-term stability, reliability, and accuracy required for commercial deployment [35,39]. Future research should focus on miniaturization, low-power operation, wireless communication, flexible electronics, and user-centered design to improve device usability and adoption.

From a translational perspective, a major limitation of current research is the lack of large-scale clinical validation. Many sensing technologies have been evaluated using small participant groups, animal models, tissue phantoms, or laboratory environments, limiting their generalizability [18]. Comprehensive clinical studies involving diverse populations and extended monitoring periods are required to establish clinical reliability and facilitate regulatory approval. Additionally, challenges related to manufacturing scalability, cybersecurity, data privacy, and healthcare integration must be addressed before widespread commercialization can be achieved.

Looking forward, the convergence of advanced photonic sensing technologies, nanomaterials, wearable electronics, and artificial intelligence is expected to drive the next generation of non-invasive glucose monitoring systems. Emerging approaches based on photonic crystal fibers, surface plasmon resonance sensors, nanoplasmonic structures, and multimodal sensing platforms may significantly improve sensitivity and selectivity while reducing the influence of physiological and environmental interference. The integration of these technologies with intelligent data analytics and personalized healthcare platforms has the potential to transform diabetes management and enable truly continuous, accurate, and non-invasive glucose monitoring.

VI. Conclusion

Non-invasive glucose sensing technologies have emerged as a promising alternative to conventional invasive and minimally invasive glucose monitoring methods, offering the potential to improve patient comfort, compliance, and long-term disease management. This review presented a comprehensive assessment of the principal non-invasive glucose sensing approaches, including optical, nanotechnology-based, electrical/electromagnetic, and physiological techniques. The fundamental operating principles, advantages, limitations, and clinical applicability of each sensing modality were critically discussed, together with recent advances in machine learning-assisted glucose prediction and wearable healthcare technologies.

Among the reviewed approaches, optical sensing technologies, particularly Optical Coherence Tomography (OCT), Optical Polarimetry (OP), Photoplethysmography (PPG), and spectroscopic techniques, have demonstrated considerable potential due to their non-invasive nature, high sensitivity, and compatibility with continuous monitoring applications. Emerging nanotechnology-assisted sensing platforms further enhance detection capabilities through improved signal amplification, selectivity, and miniaturization. In parallel, advances in artificial intelligence and machine learning have enabled more accurate interpretation of complex sensing data, reducing the impact of physiological variability and environmental interference.

Despite these achievements, significant challenges remain. Sensor accuracy is still affected by individual physiological differences, motion artifacts, environmental conditions, and interference from biological constituents such as water, proteins, and electrolytes. Furthermore, many proposed sensing systems remain at the laboratory or prototype stage, with limited large-scale clinical validation and commercialization. Addressing these limitations will require interdisciplinary efforts involving sensor engineering, materials science, biomedical research, data analytics, and clinical medicine.

Overall, although no single technology has yet achieved the performance necessary to completely replace conventional glucose monitoring methods, ongoing advances in sensing materials, photonic devices, wearable systems, and machine learning algorithms continue to accelerate progress toward clinically viable solutions. With continued research, technological innovation, and rigorous clinical validation, non-invasive glucose sensing technologies have the potential to transform diabetes management and improve healthcare outcomes for millions of individuals worldwide.

References

Aghaei, Mojtaba, et al. “Evaluation of complete blood count parameters in patients with diabetes mellitus: a systematic review.” Health Science Reports 8.2 (2025): e70488.
Biswas, Saovasis Kumar, Abdullah al Noman, and SM Rakibul Islam. “Nonlinear coupling of whispering gallery mode silica microcavity for sensing application.” 2017 3rd International Conference on Electrical Information and Communication Technology (EICT). IEEE, 2017.
Sami, A., Javed, A., Ozsahin, D. U., Ozsahin, I., Muhammad, K., & Waheed, Y. (2025). Genetics of diabetes and its complications: a comprehensive review. Diabetology & Metabolic Syndrome, 17(1), 185.
Gong, J. Y., Sajjadi, S. F., Motala, A. A., Shaw, J. E., & Magliano, D. J. (2025). Variation in type 2 diabetes prevalence across different populations: the key drivers. Diabetologia, 68(11), 2327-2339.
Barchiesi, M. A., Calabrese, A., Costa, R., Di Pillo, F., D’Uffizi, A., Tiburzi, L., & Zahid, E. (2025). Continuous glucose monitoring in type 2 diabetes: a systematic review of barriers and opportunities for care improvement. International Journal for Quality in Health Care, 37(3), mzaf046.
Biswas, Shovasis Kumar, et al. “Design of an ultrahigh birefringence photonic crystal fiber with large nonlinearity using all circular air holes for a fiber-optic transmission system.” Photonics. Vol. 5. No. 3. MDPI, 2018.
Shi, J., Fernández-García, R., & Gil, I. (2025). Sensor technologies for non-invasive blood glucose monitoring. Sensors, 25(12), 3591.
Davies, M. J., Lim, S., Slater, T., Goldney, J., Philis-Tsimikas, A., Franco, D. R., ... & Lingvay, I. (2026). Type 2 diabetes mellitus. Nature Reviews Disease Primers, 12(1), 13.
Akash, Md Mahbubur Rahman, et al. “Cardiotoxicity prediction models in cancer patients using artificial intelligence and genomics.” International Journal of Drug Delivery Technology 16.23s (2026): 60-73.
M. R. Islam, P. Akther, A. Al Maimun, M. M. R. Akash, S. Munira, and M. S. Miah, “A review of PCF-based biosensors for cancer detection and biomedical applications,” 2026.
Haque, Emranul, et al. “Numerical investigation of PCF-based biosensor for pathologies detection in blood.” Sensing and Bio-Sensing Research (2026): 100961.
Alsultani, A. B., Kovács, K., Chase, J. G., & Benyo, B. (2025). Advances in invasive and non-invasive glucose monitoring: A review of microwave-based sensors. Sensors and Actuators Reports, 9, 100332.
Islam, SM Rakibul, et al. “Advances in Photonic Crystal Fiber Biosensors for Multi-Organ Cancer Diagnostics: A Systematic Review.” 2026 22nd IEEE International Colloquium on Signal Processing & Its Applications (CSPA). IEEE, 2026.
M. Khatun and M. R. Islam, “Design optimization of high-sensitivity PCF-SPR biosensor using machinelearning and explainable AI,” 2025/2026.
Wang, M., Zheng, J., Zhang, G., Lu, S., & Zhou, J. (2025). Wearable Electrochemical Glucose Sensors for Fluid Monitoring: Advances and Challenges in Non-Invasive and Minimally Invasive Technologies. Biosensors, 15(5), 309.
Al Maimun, Abdullah, et al. “A Numerical Simulation for the Diagnosis of Cancer Cells Using a Gold-Coated SPR-Boosted PCF Biosensor.” 2026 22nd IEEE International Colloquium on Signal Processing & Its Applications (CSPA). IEEE, 2026.
Haque, Emranul, et al. “Numerical Analysis of a PCF-SPR Sensor for Adulterant Detection in Milk.” 2025 Photonics Global Conference (PGC). IEEE, 2025.
S. Khani and M. Hayati, “Optical biosensors using plasmonic and photonic crystal band-gap structures for the detection of basal cell cancer,” Scientific Reports, vol. 12, no. 1, p. 5246, 2022.
Islam, SM Rakibul, et al. “Design of hexagonal photonic crystal fiber with ultra-high birefringent and large negative dispersion coefficient for the application of broadband fiber.” Int. J. Eng. Sci. Tech 2.1 (2017): 9-16.
Xavier, C., Eugenio, F. A., Malbert, C. H., Ollagnier, C., Brajon, S., & Gondret, F. (2026). A pilot study testing a continuous glucose monitoring sensor in lean growing pigs fed contrasting diets, to document nocturnal and diurnal glycemic excursions as well as their relationships. Veterinary and Animal Science, 100612.
Islam, Md Rubayet, et al. “Ultra high birefringent hexagonal photonic crystal fibers with ultra low confinement loss employing different sizes of elliptical air holes in the core.” Asia Pacific Journal of Engineering Science and Technology 3.4 (2017): 141-150.
JASIM, L. H., POLVONOV, B. P., & NAILEVICH, I. R. (2025). NONLINEAR PROPAGATION AND SCATTERING OF ELECTROMAGNETIC WAVES IN BIOLOGICAL TISSUES AT THZ FREQUENCIES. TPM–Testing, Psychometrics, Methodology in Applied Psychology, 32(S6 (2025): Posted 15 September), 114-122.
Faruk, Md Mostafa, et al. “Ultra-high Negative Dispersion Compensating Index Guiding Single Mode Octagonal Photonic Crystal Fiber: Design and Analysis.” 2019 Innovations in Power and Advanced Computing Technologies (i-PACT). Vol. 1. IEEE, 2019.
Biswas, Shovasis Kumar, et al. “Analytical model for coupling whisperring gallery mode spherical microresonator for sensing application.” 2017 IEEE International Conference on Telecommunications and Photonics (ICTP). IEEE, 2017.
Islam, S. R., Islam, M. R., Akther, P., & Al Maimun, A. (2026). Photonic Crystal Fiber-Based Sensor for Real-Time Power System Monitoring: Design and Applications.
Saiko, G., Sadrzadeh-Afsharazar, F., Burton, T., Prahl, S., & Douplik, A. (2025). Absorption, scattering, and refractive index of blood and its components: a review. Frontiers in Photonics, 6, 1636398.
Biswas, Shovasis Kumar, et al. “Highly Nonlinear Dispersion Compensating Octagonal Photonic Crystal Fiber: Design and Analysis.” 2019 International Conference on Electrical, Computer and Communication Engineering (ECCE). IEEE, 2019.
Dey, K., Santra, T. S., & Tseng, F. G. (2025). Advancements in Glucose Monitoring: From Traditional Methods to Wearable Sensors. Applied Sciences, 15(5), 2523.
M. S. Islam, A. H. M. I. Ferdous, A. Al Mamun, M. S. Anower, M. J. Hossen, and S. U. Ahmed, “Ultra-sensitive terahertz photonic crystal fiber sensor for detection of tuberculosis,” Sensing and Bio-Sensing Research, vol. 48, p. 100814, 2025.
M. S. Islam et al., “Urinary glucose detection with spiral shape hollow core photonic crystal fiber: towards improved diabetes management,” Sensing and Bio-Sensing Research, vol. 47, p. 100748, 2025.
H. M. I. Ferdous et al., “Octagonal PCF with square-core for surface-enhanced spectroscopic properties: a new frontier in terahertz chemical sensing,” Plasmonics, vol. 19, no. 3, pp. 1257–1268, 2024.
Hasan, W. M., Ahmed, H. M., Ahmed, A. M., Rezk, H. M., & Rabie, W. B. (2025). Exploring highly dispersive optical solitons and modulation instability in nonlinear Schrödinger equations with nonlocal self phase modulation and polarization dispersion. Scientific Reports, 15(1), 27070.
Akther, Priasa. “Photonic Crystal Fiber SPR Sensors for Power and Biomedical Applications: A Review.” (2026).
Ray, T. R., Choi, J., Bandodkar, A. J., Krishnan, S., Gutruf, P., Tian, L., ... & Rogers, J. A. (2019). Bio-integrated wearable systems: a comprehensive review. Chemical reviews, 119(8), 5461-5533.
Islam, Md Rubayet, et al. “Analysis of Dispersion and Nonlinear Property in Defected Core Square Photonic Crystal Fiber.” 2018 9th International Conference on Computing, Communication and Networking Technologies (ICCCNT). IEEE, 2018.
Biswas, Shovasis Kumar, et al. “Analytical model for coupling Whispering gallery mode silica microcavity for sensing application.” 2017 2nd International Conference on Electrical & Electronic Engineering (ICEEE). IEEE, 2017.
Pal, Y. Trabelsi et al., “Plasmonic SPR biosensor with Bayesian regression for non-invasive proteinbiomarker detection,” Scientific Reports, 2025 . [CrossRef]
M. T. Hoque et al., “U-grooved selectively coated and highly sensitive PCF-SPR sensor for broad-range analyte RI detection,” IEEE Access, vol. 11, pp. 74486–74499, 20.
Wu, H., Li, B., Lu, X., Qiao, Y., Zhou, Y., & Wang, X. (2026). Real-Time Anxiety Monitoring and Mitigation for eVTOL Passengers Based on In-Ear Wearable Sensors. Applied Sciences, 16(11), 5532.

Figure 2. Hierarchical Classification of Non-Invasive Glucose Monitoring Techniques.

Figure 3. Interaction and Penetration of Optical Wavelengths within Human Skin Tissues.

Figure 4. Working principle of Optical Coherence Tomography (OCT)-based glucose monitoring.

Figure 5. Schematic representations of optical polarimetry (OP)-based glucose sensing platforms: (a) ocular sensing and (b) wearable skin-based sensing.

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

© 2026 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

Copyright: This open access article is published under a Creative Commons CC BY 4.0 license, which permit the free download, distribution, and reuse, provided that the author and preprint are cited in any reuse.