Researchers create a new AI System that Can Predict Lung Cancer Faster

Researchers create a new AI System that Can Predict Lung Cancer Faster
The Siliconreview
28 May, 2019

A team of researchers led by Dr. Mozziyar Etemadi, an Engineering professor at Northwestern University, America has come up with a new system which is claimed to detect lung cancer more effectively than doctors. The research work is published in the Journal of Natural Medicine.

The research team also included members from Google.

In the present day, CT is the most common tool which is used by the doctors to scan or detect lung cancer, but sometimes, the CT method fails or shows false detection.

According to Google’s website, powered by deep learning algorithm; the new system works on 3D technology- where it examines the CT scans. The technique works by learning how to recognize cancer signs from a large number of CT samples.

The National Cancer Institute helped the researchers to test the system on 6,716 cases with known diagnoses. As per the study reports, the overall success rate of the system was 94.4 percent. The deep learning system outperformed six radiologists when it examined a CT scan for diagnosis.

Mr. Etemadi stated that the use of 3D technology is a big achievement in the CT examination process.

The new system could identify very small signs of lung cancer that are difficult for doctors to detect. The deep learning technique can be very useful in identifying cancer cells in patients whose early scans do not reveal anything about cancer – even though Cancer might be present.