Based on artificial intelligence, the technology developed at Unicamp can be applied in other industrial sectors, in addition to health
Technology developed by researchers at Unicamp uses artificial intelligence to carry out diagnoses at a molecular level. A artificial intelligence mass spectrometry analysis technique was licensed, with the support of Innovates Unicamp, for Lacen (Central Public Health Laboratory) for exclusive use for Covid-19 tests. Detection is made from a patient's blood sample, in which a numerical signature is generated that shows changes caused by the coronavirus.
“When we insert the blood sample into the spectrometer, the blood and plasma are separated, and, subsequently, the spectrometer generates a numerical signature, around 14 and 15 thousand numbers, which are the molecules present in the plasma. These molecules are used to check whether there is any change, in this case, the presence of the virus”, explains Professor Anderson Rocha, one of the inventors of the technology.
Lacen is a network of national laboratories that respond to the National Health Surveillance Agency (Anvisa), therefore, the tests carried out by Lacen send the results to Anvisa in order to obtain regulatory registration for the product. The tests are being carried out by collecting samples from patients in hospitals in Santa Catarina. After collection, Lacen submits the samples to the spectrometer and compares them with the same results obtained by the PCR test.
Rocha also says that Lacen has already collected several samples and, to help validate the service, they are also collecting samples in São Paulo and comparing the results of the spectrometer and the PCR test. “We are helping Lacen with the collection and analysis of samples in São Paulo. In total, we sought a sampling of 300 tests performed by the spectrometer,” he reports.
The artificial intelligence algorithm used in the spectrometer, when trained to find patterns, automatically detects biomarkers with any change in the molecules. The cost of this equipment for Covid-19 tests is lower when compared to the PCR test, in addition to presenting results faster and being able to identify whether the patient is at greater risk of developing serious manifestations of the disease.
Read article in full published on the Unicamp Innovation Agency website.
