Researchers have morphed computer algorithms used by astronomers to pick out hard-to-distinguish planets and galaxies into a system that can help medical workers spot hard to find breast cancer cells.
Researchers from the Cancer Research UK Cambridge Institute and the Department of Oncology and the Institute of Astronomy at the University of Cambridge in the UK say the idea behind the astronomy/medical collaboration is to help health care professionals speed breast cancer diagnosis by accurately automating a process that today is manual and requires pathologists to peer through microscopes to spot subtle differences in staining of tumor samples.
[RELATED: First bionic eye gets FDA blessing]
Such a process was not unlike what the astronomers would do in trying to pick out distinct objects in the night sky using a telescope.
To test the algorithms, the researchers used them to assess levels of known biomarkers for aggressive cancer in samples from more than 2,000 breast cancer patients, according to a University of Cambridge statement.
"Researchers then compared the accuracy of manually scoring these results, by observing the staining of the tumor samples down the microscope, versus relying on a computer to do this automatically. This showed that the new automated system was at least as accurate as the manual one, while at the same time being many times faster," the researchers stated.
"The results have been even better than we'd hoped, with our new automated approach performing with accuracy comparable to the time-consuming task of scoring images manually, after only relatively minor adjustments to the formula. We're now planning a larger international study involving samples from more than 20,000 breast cancer patients to further refine our strategy," said lead author of the paper on the on the research Dr. Raza Ali, a pathology fellow from Cancer Research UK's Cambridge Institute at the University of Cambridge. The study was published in the British Journal of Cancer.
The researchers went on to say before these can be applied in a clinic, their usefulness needs to be verified in hundreds or sometimes thousands of tumor samples. The automated approach means users could can analyze up to 4,000 images a day, helping streamline the process of translating these discoveries into the clinic.
Check out these other hot stories: