Machine learning software diagnoses breast cancer


Follow us onFollow Tech Explorist on Google News

Scientists from Houston Methodist Research Institute have developed a new machine learning software. This new machine learning software accurately diagnoses breast cancer risk 30 times faster than doctors.

This new machine learning software aids doctors in giving better treatment. It gives millions of records within a short time. Thus, it allows doctors to discover breast cancer risk more effectively by using patient mammograms. A mammogram is a breast X-ray that aims to spot any potentially cancerous cells before symptoms arise. Meanwhile, it has the potential to reduce unnecessary biopsies.

According to analysis, in many countries, women over 50 are advised to get a precautionary screening every two years. But as good as that system is, 50% results of tests in the US showed false positive results.

Generally, doctors use computer software to analyze mammogram images.

According to scientists, almost 20% of women follow unnecessary biopsies. So, this new machine learning software is designed to greatly reduce that number by making a more accurate diagnosis the first time.

Scientists tested the AI on 500 breast cancer patients’ mammogram results and pathology reports. The software gives diagnostic information only within few hours. It identified the breast cancer subtype each patient had. Then, scientists cross-checked the AI diagnosed with clinical results. Hence, they proved the software was 99 percent accurate. Doctors would take almost 500 hours to complete the same.

This machine learning software won’t be able to prevent all false positives or suspicious mammogram results. Sometimes there’s just not enough information available to make a diagnosis. But it should help doctors make a more accurate conclusion.

Stephen Wong, one of the researcher said, “We’re looking forward to seeing how this software could make breast cancer diagnosis as well as other cancers. It is faster and more accurate, and can save people’s time and stress of unnecessary further testing.”