Know Labs debuts a portable, noninvasive glucose monitor prototype

Know Labs has made significant strides in advancing its noninvasive glucose monitoring technology, known as Body-Radio Frequency Identification (Bio-RFID). By successfully demonstrating the technology’s ability to accurately measure blood sugar levels without invasive procedures, the company has developed the initial prototype of the hardware that will accommodate this innovative technology.

During the unveiling of the prototype, Steve Kent, the Chief Product Officer of Know Labs, lauded the device, referred to as Generation 1, as the “ultimate research and development tool.” Generation 1 is designed to be compact and portable, allowing users to easily carry it in their pocket or hold it in their hands while conducting on-demand glucose tests. Kent highlighted several impressive features of the device, such as its robust onboard computing capabilities, built-in machine learning functionality, extended battery life, and flexible configurability to support diverse development initiatives.

The development of this portable device proved to be a complex and time-consuming undertaking, as CEO Ron Erickson acknowledged. Overcoming numerous engineering challenges and undergoing hundreds of iterations, Know Labs and its partners devoted significant resources to bringing the prototype to fruition. Erickson emphasized that the completion of Generation 1 signifies a remarkable engineering achievement not only for the company but also for the field of medical diagnostics. 

Previous attempts at noninvasive glucose measurement have encountered limited success or remain years away from fruition. The groundbreaking Bio-RFID sensor, leading the way in a new branch of scientific advancement, brings Know Labs closer to its objective of enhancing the quality of life for individuals with diabetes.

With the availability of the Generation 1 prototype, Know Labs has outlined its plans to continue validating the Bio-RFID technology throughout the year. The company will concentrate on refining the physical device with the ultimate goal of seeking clearance from the FDA, thus becoming the first noninvasive glucose monitor authorized by the regulatory agency.

The Bio-RFID sensor functions by emitting radio waves through the skin to capture molecular signatures in the blood. These molecular patterns are subsequently analyzed by Know Labs’ machine learning algorithms to calculate the user’s blood sugar levels. The company is actively working on enhancing the accuracy of these calculations. Previous studies demonstrated that the sensor achieved a mean absolute relative difference (MARD) of approximately 20%, which indicates the variation between the sensor’s readings and a reference glucose measurement.

Improved Accuracy of the Glucose Monitor through LightGBM Model Implementation

In a recent study conducted in May, Know Labs introduced a new machine-learning model called LightGBM, which significantly reduced the discrepancy between the sensor’s readings and those of the Dexcom G6 continuous glucose monitor. As a result, the overall MARD improved to just under 13%. It is worth noting that a glucose sensor with a MARD below 10% is generally considered highly accurate. Dexcom’s recently FDA-cleared G7 CGM, for instance, boasts a MARD slightly above 8% in both children and adults, while Abbott achieved a 7.9% difference in a study of their own 14-day FreeStyle Libre 3 sensor.

Know Labs remains committed to further enhancing the accuracy of its noninvasive glucose monitoring technology, thereby drawing closer to its goal of providing a reliable and convenient solution for individuals managing diabetes.