Implantable Glucose Sensor — Biomedical Instrumentation & Clinical Sensing

Abstract — Implantable Glucose Sensor

nutrition intake, physical activities in everyday life and glucose control. CGMS available in market are of two types based on their working principle. Needle type systems with few weeks lifespan (e.g., enzyme-based Freestyle Libre ) and implant type system (e.g., fluorescence-based Senseonics ) with few months of lifespan are commercially available. An alternate to both working methods, herein, we propose electromagnetic-based sensor that can be subcutaneously implanted and capable of tracking minute changes in dielectric permittivity owing to changes in blood glucose level (BGL). Proof-of-concept of proposed electromagnetic-based implant sensor has been validated in intravenous glucose tolerance test (IVGTT) conducted on swine and beagle in a controlled environment. Sensor interface modules, mobile applications, and glucose mapping algorithms are also developed for continuous measurement in a freely moving beagle during oral glucose tolerance test (OGTT). The results of the short-term (1 h, IVGTT) and long-term (52 h, OGTT) test are summarized in this work. A close trend is observed between sensor frequency and BGL during GTT experiments on both animal species.

Sensor Interfacing & Analog Front-End Conditioning

A growing research interest in healthcare systems has been observed in the last few decades, primarily focusing on the development of various point-of-care medical devices that can be of significant help to patients with financial or remote location constraints. Data base driven algorithms, multimodal information processing, and trend analysis are the key parts of bioinformatics that can help managing individual health, and issue predictive alerts to prevent fatal situations 1 , 2 . Metabolic diseases require the continuous management of eating habits, dietary plans, physical exercise, exposome and other general behaviors including sleep. The number of people with diabetes worldwide is 460 million as of 2019, and it is steadily increasing every year, with approximately 11.3% of deaths worldwide is due to the diabetes-related complications 3 . Diabetes mellitus increases further incidences of health complications exponentially 4 , 5 . The risk in diabetes can be minimized by monitoring and adapting habits to control blood glucose level (BGL). Commonly used method to measure BGL is using enzyme-based glucose sensor relying on electrochemical reaction 6 , 7 , known as amperometric sensor that produces measurable current proportional to BGL. This can be implemented in two ways either measuring directly from blood or from interstitial fluid (ISF). Former method is also known as self-blood glucose monitoring (SBGM) relies on finger-pricking method where a small blood is drawn from fingertip and put on glucose measuring strip. Although it measures glucose level with considerable accuracy, it is painful, causes skin irritation to the patients requiring frequent measurements. This method does not give any insight of BGL changes over time. Later method measures glucose level from the ISF through a needle inserted into the subcutaneous fat and connected to a measuring device placed and attached on skin. The device has main measurement unit and other control, communication units. This method is of one-time use, have short self-life (few days), expensive, and painful due to the insertion and presence of a needle for in-situ measurements 8 , 9 , 10 . It provides a BGL trend over continuous time, however the sensor life is few days. Extending the enzyme activity using electrochemical coating, sensor life could be increased from few days; however, limited to approximately two weeks of use 11 , 12 . Various methods alternate to enzyme-based blood glucose detection have been extensively reviewed to improve lifespan of sensor, accuracy of CGMS in literature 13 , 14 . Optical methods that use light-emitting diode (LEDs) of different wavelengths 15 , 16 , 17 are also extensively studied for blood glucose detection. Usually, the sensor is a photo sensitive detector that can detect variations in optical intensity owing to variations in blood glucose level. Near-Infrared optical frequencies show glucose dependent intensity variations due to spectral absorption when illuminated, and the reflectance changes with the blood glucose variations. However, the strong absorption and weak intensity of the reflected signal affects the accuracy. To overcome this, improved optical methods using fluorescent materials with glucose selectivity have also been investigated 18 , 19 . The method uses a glucose-binding molecule that causes a change in fluorescent activity depending on the glucose level in blood. However, this method is limited by the degradation of the fluorescent material itself over time. As a result, the fluorescence reflectance diminishes gradually, and as time passes, it is barely detectable 20 , 21 .

Bio-Signal Waveform Analysis & Simulation Findings

The development of electromagnetic based glucose sensors has been studied in various lab experiments to non-invasive measurement of blood glucose level 22 . From a recent study it is understood that mechanism of glucose metabolism is not straightforward rather a sophisticated cascade of biochemical reactions. However, the dielectric response of water is directly affected by the glucose and a relevant marker for indirect measurement of glucose level in vivo 23 . The fundamental working principle of EM-based glucose sensors is to sense glucose dependent dielectric permittivity changes. The glucose dependent dielectric permittivity of blood has been characterized over a wide frequency range 24 . It has been observed that the dielectric permittivity decreases with increase in glucose level 25 . In general, glucose level dependent permittivity changes are reflected as a change in the resonance frequency of these EM sensors. Few of the relevant published works using electromagnetic based glucose sensor are in vitro measurement technique 26 , 27 and wearable type measuring in vivo from outside body 28 . EM-based permittivity sensing biosensors have already proven to be effective for the detection of tumors and malignant cells in the body. The dielectric properties of various organs and tissues in living bodies have been previously characterized over a wide spectrum 29 , 30 . The resonance frequency of proposed EM-based implant sensor depends on permittivity of the surrounding environment. The sensor frequency changes inversely with changes in dielectric permittivity of the material in which sensor is embedded. The dielectric permittivity of blood changes as the glucose level changes. This glucose dependent permittivity changes are reflected in the sensor resonance frequency. Using a regression model, sensor frequency can be mapped to BGL. The EM-based sensors for BG measurement have already been attempted, and encouraging results have been reported in several research 31 , 32 .

Scope & Clinical Applications

• Real-time multi-parameter vital signs telemetry (ECG, Heart Rate, SpO2, Temperature, Respiration)
• Low-noise Analog Front-End (AFE) conditioning with active Driven Right Leg (DRL) 50 Hz cancellation
• Digital signal processing in MATLAB: Pan-Tompkins QRS detection, FIR/IIR notch, and Continuous Wavelet Transforms
• Finite element tissue bio-electromagnetics and orthopedic mechanical simulations in COMSOL & ANSYS
• Complete Arduino firmware .INO sketches, circuit schematics, and university dissertation reports