Artificial intelligence detects breast cancer better than a professional doctor

The article is about fatal harm household chemicals we mentioned that breast cancer is one of the most common cancers among all women on the planet. According to the world health organization, annually in the world there are nearly 1 million cases and at the moment breast cancer is one of the main causes of death of women. In this regard, in many countries, women aged more than 50 years recommended every three years to pass the so-called mammography to detect cancerous tumors in the mammary glands. However, doctors often make mistakes and either put a fatal diagnosis perfectly healthy patients either do not detect the initial stages of the disease, which is extremely important for successful treatment. Fortunately, researchers from Google Health has developed an artificial intelligence that recognizes tumors with more precision than a professional doctor.

The process of mammography

About the benefits of artificial intelligence in the detection of breast cancer was described in a scientific publication ScienceAlert. We mentioned that the annual fatal diagnosis approximately 1 million women, however, today the situation is much worse. If you believe foreign sources, in 2018, an aggressive type of cancer was detected in 2 million patients. Because of this, the need for regular mammography significantly increased, because the sooner doctors will learn about developing the disease, the more higher is the likelihood of full recovery.

How to detect breast cancer

To detect developing breast cancer is possible by means of x-ray images made with the help of mammography. These images are studied by at least two radiologists, but even with a team of doctors make serious mistakes. However, according to scientists from Google Health, they have developed an artificial intelligence trained on the set of x-rays that can detect signs of cancer with virtually no errors.

The study of x-ray images obtained using the mammography unit

The technology has been tested by thousands of women from Britain and the United States. After artificial intelligence scanned x-rays, they were additionally examined by radiologists. As it turned out, the accuracy of the computer in this case is close to the professionalism of experienced doctors. It is noteworthy that for diagnosis artificial intelligence have not studied the medical history of patients, which could be prerequisites for the occurrence of breast cancer.

Eventually it became clear that artificial intelligence has reduced the production of incorrect diagnoses by 5.7% in USA and 1.2% in the UK. The technology also reduced the likelihood of missing the symptoms of the disease by 9.4% in the US and 2.7% in the UK. The head of Google Health in the UK Dominic king shared that these are excellent results considering that early diagnosis of cancer facilitates treatment.

The development of such artificial intelligence not only improves the accuracy of diagnosis but also is able to deliver radiographers from the laborious and highly responsible work. The researchers note that at the moment of experts in this matter, there are very few, and sometimes the results of a mammogram are examined for quite a while. It turns out that thanks to new technology solved two problems at once.