Machine learning could improve lung cancer screening

By Paul Biegler

Scientists have developed a blood test or a “liquid biopsy” that detects lung cancer at an early stage. This could save more than 11,000 lives in the United States alone.

The research conducted and published in by radio-oncologist Maximilian Diehn of the Stanford Cancer Institute in California natureuses machine learning to examine tiny amounts of DNA from the tumor in the bloodstream.

The team found that the amount of DNA revealed a lot of facts about the most common type of lung cancer, known as non-small cell lung cancer, in people who already had the disease, including their cell type, how advanced, and how aggressive it was would probably spread.

The researchers then trained a machine learning model to assess the likelihood that DNA variants in the blood would come from lung cancer. This method is referred to as “lung cancer probability in plasma” or “lung CLiP”.

When they put lung CLiP through its paces, it discovered lung cancer in 63% of patients in stage 1, the earliest phase when the cancer was lung-restricted and did not spread to lymph nodes or outside the breast.

This may be a critical advance as lung cancer is the leading cause of cancer death in the United States and more than 155,000 people die each year. Every fifth of these deaths is avoidable.

If the disease is recognized early, any treatment arm, including surgical removal, radiation therapy, and chemotherapy, will becomes more effective. However, recent efforts to look for early lung cancer have been a bleak failure.

In 2013, the U.S. Preventive Services Task Force recommended everyone high risk People are examined with a low-dose breast CT scan (LDCT). That is, if you are between 55 and 80 years old and smoke 30 or more packs of cigarettes a year – including people who have stopped in the past 15 years – you should get an LDCT every year.

However, one study found that fewer than four out of 100 candidates were scanned in 2015, which the authors attributed to doctors and patients who knew nothing about the test, as well as reimbursement problems.

LDCT also generates a large number of false alarm – Over 90% of the scans detect cancer that is actually not present.

The lung CLiP blood test could refine and speed up this screening process.

“One possible application of Lung-CLiP could be to serve as the first screening for some of the approximately 95% of high-risk patients in the United States who, although they are candidates for LDCT, are not screened,” the authors write.

People with a positive lung CLiP test could then be transferred more safely to LDCT. A “hybrid” approach, the authors say, could increase the number of lives saved annually in the United States from the current 600 to 12,000.

It is an approach that researchers believe could work for cancer beyond the lungs:

“[By] We believe that it may be possible to develop CLiP methods for a variety of malignancies by including molecular traits that are suitable for other cancers, ”they conclude.

Megan Thompson

Megan is an aspiring entrepreneur looking to add value to people's lives & give back to those in need. Follow the journey to discovering her purpose.

Write A Comment