AI Software Detects Lung Nodules on CT Scans in 20 Seconds

AI Software Detects Lung Nodules on CT Scans in 20 Seconds

A team of researchers from Russia have come up with an innovative AI software that reads CT scan images in around 20 seconds for diagnosing lung cancer.

According to the resources, the AI system which is named after the famous Russian science fictional writer, Doctor Alzimov can be installed on the computer which produces clear images of marked findings that helps in effortless interpretation.

The first CT images were derived in late months of 2018. Around 60 patients were detected and nodules were found on the lungs which were really small, sizing 2 millimetres. The diagnosis took place in an Oncological Center at St. Petersburg, Russia.

According to the news release, the researchers first structured an algorithm to discover the nodules with a starting size of 6 millimetres as the radiologists had kicked off with the treatment of this size tumour themselves. However, the system was analysed to be much smarter as it helped in detecting nodules of much lesser size than this.

The innovative software used chord system, dividing the CT scans to group together lung nodules. The system studied 1,000 CT scans. The group of researchers are now planning to finally shift the CT scans to supercomputer, so that the diagnostic checking time can be cut down by more 2 seconds.

Share this post