Blood sugar levels have always been an important diagnostic in the detection of several health related conditions. This is especially true in the case of hypoglycemia diagnosis. But given the unnecessary hassle surrounding traditional blood sugar tests, it is a no brainer that any patient or even a doctor would prefer a simpler route if it was available. So for this reason, researchers from the University of Warwick have come up with exactly what they need. Let us now take a look at their groundbreaking new development.
Current methods for hypoglycemia detection involve the usage of Continuous Glucose Monitors (CGM) that can measure glucose in intestinal fluid with the help of an invasive sensor that has a small needle attached to it. Following this, depending on the diagnosis, the sensors send alarms and data to a seperate device.
Although effective, this method is a little tedious as the device may need calibration twice a day by performing blood glucose tests which would in turn require invasive finger pricking. But now, thanks to the latest findings of the researcher from the University of Warwick, a method has been devised that uses the power of AI to detect hypoglycemic events from raw ECG signals acquired through non invasive wearable sensors.
By combining a smart AI algorithm with the data from the ECG, medical professionals can observe outputs where in a green line would represent normal glucose levels, a red line for low glucose levels and a horizontal line for 4mmol/L glucose value. The grey area that would surround the continuous line would reflect the measurement error bar. Moreover there is another reading in the form of vertical bars that represents the relative importance of each ECG wave which can help determine whether a heartbeat is hypo or not.
Such readings are only possible thanks to the adaptive abilities of the AI. By factoring in subjective differences, the AI can make personalised diagnosis which was not possible in the past. Combining this with the fact that this method is comparable in accuracy to current methods and that ECGs can even be detected while sleeping or any other state for that matter, this new technique could quite easily become the norm in times to come and help people avoid the hassle of pricking their skin over and over again just to get a blood sugar reading.