Haemoglobin Assesment by AI

Bosch has developed a portable haemoglobin monitor for screening anaemia cases non-invasively. By non-invasive, we mean there is no need for a blood test.

Basically, a finger scanner determines the value. It uses multi-wavelength spectometry on the surface of the skin. An optical sensor measures PPG or photoplethysmogram signals. It is an optical test that detects changes in blood in peripheral circulation. It takes 30 seconds to detect low haemoglobin concentrations.

Here Machine Learning (ML) is used. The device is embedded with an algorithm which monitors the wavelength of the light and uses 27 different characteristics to determine and classify haemoglobine value.

The algorithm has been trained with more than 10000 anaemia data points. The data has been collected clinically, and there is corresponding ground truth data. This forms the basis for ML algorithm.

The more validated data sets are fed into the ML algorithm which keeps on learning continuously. It bypasses lab analysis. There is no risk of infection from contaminated needles.

This device is battery-operated and does not need subsequent calibration. It is easy to use in the interiors. It is likely to be available in India in mid 2021.

print

Leave a Reply

Your email address will not be published. Required fields are marked *