Modul Scientific and Technical Center (STC) has developed software for detection of viral pneumonia, including coronavirus. It is capable of diagnosing the disease with 93% accuracy, the Made in Russia project reported.

The software can detect lung pathology in a few minutes after analysis of X-ray images. It creates three-dimensional visualization of images, determines the volume, area and extent of lung affection. A deep neural network is used to detect lung lesions, which was trained and tested on the medical biophysical center's databases. Here, work was also carried out on the markup of CT images.

According to Alexey Mokhnatkin, Deputy Director for Development of the STC, the new software can be easily integrated into the medical IT infrastructure. In the future, the company intends to significantly increase the functionality of the software. This will make it possible to use it in other medical areas besides pulmonology.

Currently, the software is part of the neural network hardware and software system (NNPS) demonstrator, which is a scalable computing and communication platform. It is designed for the creation and implementation of medical artificial intelligence products.

The STC solution consists of interconnected automated workstations (AWS), where Russian microprocessors are used. At the same time, the workstations are connected to the data center through a local network.

Made in Russia // Made in Russia

Author: Ksenia Gustova