A noninvasive mid-infrared spectroscopy system for accurate and pain-free glucose concentration monitoring

Background

Diabetes management is a critical area of healthcare, with over 340 million people worldwide requiring regular monitoring of blood glucose levels to prevent acute complications and long-term organ damage. The standard approach involves invasive finger-prick blood tests, which are often painful and inconvenient, leading to poor patient compliance and suboptimal disease control. As a result, there has been significant interest in developing noninvasive glucose monitoring technologies that can provide accurate, real-time measurements without the need for blood samples. Optical sensing methods, particularly those based on spectroscopy, have emerged as promising candidates due to their potential for pain-free, continuous monitoring.

Despite extensive research, current noninvasive approaches—especially those utilizing near-infrared (near-IR) spectroscopy—face significant limitations. Near-IR techniques struggle with the challenge of overlapping absorption features from various skin components such as water, proteins, and hemoglobin, making it difficult to isolate the glucose signal. This spectral congestion necessitates complex multivariate analyses, which often yield unreliable or inconsistent glucose readings. Additionally, near-IR light does not penetrate deeply enough to reach the dermal interstitial fluid, where glucose concentrations closely mirror those in blood. These technical hurdles, combined with issues like motion artifacts and variability across different skin types, have hindered the widespread adoption of noninvasive glucose sensors based on current optical methods.

Technology

This technology is a noninvasive glucose monitoring system that utilizes mid-infrared (mid-IR) spectroscopy powered by Quantum Cascade (QC) lasers to detect glucose levels through the skin. The system emits high peak power, pulsed mid-IR light in the 8.4–9.9 µm wavelength range, which penetrates the skin to reach the dermal interstitial fluid (ISF) layer—an area where glucose concentrations closely mirror those in the blood. Backscattered mid-IR light, which carries information about glucose absorption features, is collected using hollow core mid-IR fibers and analyzed with Partial Least Squares Regression (PLSR) to quantitatively predict glucose concentrations. The setup includes a fiber-optic probe, a cooled mercury cadmium telluride (MCT) detector, and mechanical stabilization to minimize noise from physiological movements, enabling pain-free, real-time glucose monitoring without the need for blood samples or finger pricks.

What differentiates this technology is its strategic use of mid-IR wavelengths, where glucose has strong, distinct absorption features that are not significantly overlapped by other skin constituents—unlike near-IR methods, which suffer from spectral interference and unreliable readings. The high peak power of QC lasers allows for sufficient tissue penetration to access the ISF, while the pulsed operation minimizes thermal damage. The use of hollow core fibers for both delivery and collection of light enhances signal stability and reduces artifacts from patient movement. Advanced data analysis with PLSR further refines glucose prediction accuracy, and experimental validation has demonstrated clinically relevant performance in both in vitro and in vivo settings. This integrated approach addresses longstanding challenges in noninvasive glucose sensing, offering a pathway to reliable, user-friendly, and continuous glucose monitoring for diabetic patients.

Benefits

• Noninvasive glucose monitoring eliminates painful finger pricks, improving patient comfort and compliance.
• Mid-infrared spectroscopy with Quantum Cascade lasers provides high specificity by targeting glucose’s distinct absorption features, reducing interference from other skin components.
• High peak power pulsed mid-IR light enables sufficient penetration (~1 mm) into the dermal interstitial fluid for accurate glucose detection.
• Use of hollow core mid-IR fibers for light delivery and collection enhances signal stability and reduces noise from physiological motion.
• Partial Least Squares Regression (PLSR) analysis enables reliable quantitative prediction of glucose concentration from spectral data.
• Mechanical stabilization techniques reduce spectral noise from involuntary movements, improving in vivo measurement accuracy.
• Technology supports potential development of a home-use, continuous glucose monitor, enhancing diabetes management and quality of life.

Applications

• Noninvasive home glucose monitoring
• Continuous glucose monitoring wearables
• Hospital bedside glucose assessment
• Point-of-care diabetes screening

Additional info

This noninvasive glucose monitor utilizes pulsed mid-infrared QC lasers (8.4–9.9 µm) to penetrate skin and detect distinct glucose absorption features in dermal interstitial fluid. Backscattered light is collected via hollow core fibers, and Partial Least Squares Regression quantifies glucose concentration, providing a pain-free alternative to blood sampling.