Researchers at Johns Hopkins Kimmel Cancer Center in the US have developed Delfi (DNA evaluation of fragments for early interception), an artificial intelligence (AI)-based blood testing technology that can detect lung cancers.

According to a new study conducted in nearly 800 patients with and without cancer, the DELFI test technology was able to identify more than 90% of lung cancer cases.

The DELFI technology uses machine learning to analyse millions of cell-free DNA fragments in blood samples for abnormal patterns in various genomic regions, including DNA size and amount

It detects unique patterns in the DNA fragmentation shed from cancer cells circulating in the bloodstream.

When tested using blood samples from 796 people in Denmark, the Netherlands and the US, DELFI was able to precisely differentiate between patients with and without lung cancer.

When combined with analysis of clinical risk factors, a protein biomarker and computed tomography imaging, the new test technology helped to spot 94% of cancer cases across stages and subtypes.

These cases included 96% of patients with advanced-stage III/IV cancers and 91% of those with earlier or less invasive stage I/II cancers.

Johns Hopkins Kimmel Cancer Center postdoctoral fellow Dimitrios Mathios said: “It is clear that there is an urgent, unmet clinical need for development of alternative, noninvasive approaches to improve cancer screening for high-risk individuals and, ultimately, the general population.

“We believe that a blood test, or ‘liquid biopsy,’ for lung cancer could be a good way to enhance screening efforts because it would be easy to do, broadly accessible and cost-effective.”

Currently, the DELFI-L101 clinical trial is assessing a test using the DELFI technology in 1,700 subjects, who include healthy individuals, lung cancer patients and those with other cancers in the US.

This trial is sponsored by Delfi Diagnostics, a spin-out of the Johns Hopkins University, and is intended to further evaluate DELFI in other cancer types.