Called AnamNet, the software tool can ‘read’ chest CT scans of potential Covid-19 patients and estimate the damage caused to the lungs by scanning for specific abnormal features. It can provide automated assistance to doctors and therefore help in faster diagnosis and better management of Covid-19.
AnamNet was developed by researchers from the departments of computational and data science (CDS) and instrumentation and applied physics in collaboration with researchers from Oslo University Hospital and University of Agder in Norway. The study has been published in IEEE Transactions on Neural Networks and Learning Systems journal.
The software employs deep learning and other image processing techniques. Researchers trained Anam-Net to look for abnormalities and classify areas of the lung scan as either infected or not. The tool can accurately judge the severity of the disease by comparing an infected area with a healthy area.
Naveen Paluru, first author and PhD student in the lab of Phaneendra Yalavarthy, associate professor, CDS, said: “It basically extracts features from the chest CT images and projects them onto a non-linear space [a mathematical representation], and then recreates the [segmented] image from this representation.”
The software is lightweight with a small memory footprint. The team has developed an app called CovSeg that can be run on a mobile phone and potentially be used by healthcare professionals. Paluru says that this feature is missing from currently available state-of-the-art technologies that require specialised hardware.
The software tool is available free to the public.