The measurement principle of implantable sensors

The implantable sensor is a small cylinder, about the size of half a matchstick, which the doctor places under the skin, usually in the upper arm. The procedure is quick and is carried out under local anaesthetic with lidocaine. Once it is in place beneath the skin, the sensor is surrounded by interstitial fluid, and it measures the concentration of glucose in this fluid, with no direct connection to the blood vessels [1].

The outer casing of the sensor carries a thin layer of porous polymer. Glucose from the interstitial fluid passes freely in and out of this layer, so the amount of glucose in the outer coating closely follows the amount in the surrounding tissue. Inside the cylinder sit a small light source, light detectors and a temperature sensor. Every few minutes the device wakes for a fraction of a second, switches on the light, reads the response of the polymer layer, passes it on and then falls dormant again [2].

The implantable sensor uses an optical technology based on fluorescence. It differs from the enzymatic, electrical technology of ordinary (transcutaneous) sensors. Anchored within the polymer coating are indicator molecules, built around boronic acid groups. These groups bind reversibly and selectively to glucose molecules and then let them go again, without changing them [3].

When glucose binds, the optical behaviour of the indicator molecule changes. A small light source inside the sensor sends a beam into the polymer layer, and the indicator molecules respond by emitting light of their own, but in a different colour (wavelength), a phenomenon known as fluorescence. The more indicator molecules have glucose bound to them, the stronger the light they emit. A detector inside the sensor measures this intensity and converts it into a glucose concentration. Substances such as paracetamol or vitamin C do not affect the result, because no enzymatic reaction takes place here [4].

As the glucose concentration in the interstitial fluid rises, more and more glucose molecules enter the polymer layer and bind to the indicator molecules. The result is a stronger fluorescent light at every measurement. When the glucose concentration falls, glucose comes off the indicator molecules and leaves the layer, and the fluorescent light weakens. Binding and release happen quickly and continuously, so the light signal follows the changes in the tissue [3] [5].

The raw light signal alone is not enough, however. At the same time, the sensor also measures the local temperature, because temperature affects both the light sent towards the indicator molecules and the fluorescence they give back [2]. A reference channel checks the true intensity of the stimulating light, and the software takes into account the time elapsed since implantation and the slow degradation of the indicator molecules exposed to light. All of this together yields a corrected value for the local glucose concentration [1] [5].

The implanted sensor has no battery and does not continuously broadcast a radio signal of its own. The component worn on the skin, sitting directly over the implant, contains a coil that generates a magnetic field (NFC) every few minutes. This field passes through the skin and powers the sensor for less than a second. The rest of the time, the sensor lies inactive [1] [2].

The information travels through the same magnetic field. The sensor alters the field slightly, and the component on the skin reads these changes, obtaining in this way the intensity of the raw light signal together with the measured temperature. From this signal, the external component calculates the corrected glucose concentration, stores it locally and sends it via Bluetooth to the app on your phone [3] [6]. The magnetic field works only over a very short distance, which means that without the on-skin component correctly in place, the sensor cannot be activated and cannot be communicated with.

No. The indicator molecules in the implantable sensor bind glucose and then let it go, without breaking it down and without turning it into anything else. This is called affinity sensing, that is, measurement by reversible attraction, and the amount of glucose in the tissue is left unchanged. What the sensor reads is essentially a balance between how many glucose molecules are bound, at any given moment, to the indicator molecules [7].

Conventional sensors, worn on the skin, work differently. They rely on an enzyme, glucose oxidase, which breaks glucose down at the surface of an electrode and produces an electric current proportional to the glucose concentration. The amount consumed is minute and does not change your blood glucose, but the reaction depends on the presence of oxygen and gradually exhausts the enzyme. The main advantage of affinity sensing is its long-term stability, which is what currently allows the implantable sensor to run for a year [8].

Glucose detection itself happens entirely inside the implanted sensor, since that is where the polymer layer with the indicator molecules, the light source and the detectors are. The component worn on the skin never comes into contact with the interstitial fluid and does not detect glucose [3]. Even so, the system cannot work without it.

The external component powers the sensor, triggers each measurement, receives the raw signal, turns it into a glucose concentration through calculations and corrections, stores the readings, alerts you by vibration when needed and sends the data to your phone [6] [9]. If you take it off, the sensor stays in place and undamaged, but no readings are recorded until you put it back, correctly positioned over the implant. The external component of this continuous glucose monitoring system is called the transmitter.