A team of Google researchers has used a deep-learningalgorithm to detect lung cancer accurately from computed scans.
The work demonstrates the potential for ArtificialIntelligence (AI) to increase both accuracy and consistency, which could helpaccelerate adoption of lung cancer screening worldwide.
Lung cancer is the deadliest of all cancers worldwide —more than breast, prostate, and colorectal cancers combined — and it’s thesixth most common cause of death globally, according to the World HealthOrganization.
“Using advances in 3D volumetric modeling alongsidedatasets from our partners (including Northwestern University), we’ve madeprogress in modeling lung cancer prediction as well as laying the groundworkfor future clinical testing,” Shravya Shetty, M.S. Technical Lead atGoogle explained in a blog post late Monday.
Google researchers created a model that can not onlygenerate the overall lung cancer malignancy prediction (viewed in 3D volume)but also identify subtle malignant tissue in the lungs (lung nodules).
In the research, Google AI leveraged 45,856 de-identifiedchest CT screening cases (some in which cancer was found).
“When using a single CT scan for diagnosis, our modelperformed on par or better than the six radiologists. We detected five per centmore cancer cases while reducing false-positive exams by more than 11 per centcompared to unassisted radiologists in our study,” said Google.
For an asymptomatic patient with no history of cancer, theAI system reviewed and detected potential lung cancer that had been previouslycalled normal.
These initial results are encouraging, but further studieswill assess the impact and utility in clinical practice, said Google.
The research was published in the journal Nature Medicine.