AI tool analyzes CT scans to spot prostate cancer in seconds
Continuous advances in artificial intelligence promise to shake up medical care in all kinds of exciting ways, with the ability to rapidly scan medical images and spot signs of disease far more efficiently than humans can. Scientists in Australia have now adapted this technology for the early detection of prostate cancer, with their software outperforming trained radiologists to detect cancerous growths in seconds.
For many medical ailments, an early diagnosis can greatly improve the treatments available and therefore the chances of overcoming them. Improvements in machine learning and computing power have led to highly capable forms of artificial intelligence that could be invaluable in this regard. We've seen AI tools that can improve an ECG's ability to reveal heart dysfunction, more accurately predict survival rates of ovarian cancer and just this week, calculate diabetes risk by measuring fat around the heart.
The latest example of this comes from researchers at Melbourne's RMIT University and St Vincent's Hospital, who started with CT scans of asymptomatic patients both with and without prostate cancer. The scientists note that, generally speaking, CT scans are useful for detecting ailments like bone and joint problems, but it is difficult for radiologists to use them to detect prostate cancers.
Using the CT scans, the AI software was trained to search for irregularities that could be indicative of the disease. The tool improved with each scan, refining its abilities and adapting to analyze scans from different machines, eventually spotting the smallest features of the disease. In time, it was able to outperform radiologists and detect cancerous growths in seconds, even before patients exhibited any symptoms.
“We’ve trained our software to see what the human eye can’t, with the aim of spotting prostate cancer through incidental detection,” says study author Dr Ruwan Tennakoon, from RMIT. “It’s like training a sniffer dog – we can teach the AI to see things that we can't with our own eyes, in the same way a dog can smell things human noses can’t.”
The scientists say the technology could be adapted to a variety of diagnostic equipment, such as MRI machines. The hope is that it can be used as a type of integrated screening tool for CT scans, which involve high doses of radiation and therefore aren't suitable for regular cancer screening. But if a patient is having a CT scan for some other reason, the AI tool could be used to screen them for cancer at the same time.
“Australia doesn’t have a screening program for prostate cancer but armed with this technology, we hope to catch cases early in patients who are scanned for other reasons,” says Dr Mark Page, Head of CT in Diagnostic Imaging at St Vincent’s Hospital Melbourne. “For example, emergency patients who have CT scans could be simultaneously screened for prostate cancer. If we can detect it earlier and refer them to specialist care faster, this could make a significant difference to their prognosis.”
The research was published in the journal Scientific Reports.