How efficient is artificial intelligence screening for breast cancer? 30 times faster than hospital old drivers!

Huffingtonpost

Breast cancer is a malignant tumor that occurs in the glandular epithelium of the breast. It has been one of the terrorist killers that threaten the health of women. Previously in the screening of the disease, most doctors used the mammogram test method, but this method usually requires breast biopsy to assist, which brings unnecessary pain to women.

However, through the efforts of researchers, the future AI will significantly reduce the discomfort caused by breast disease testing for women.

Forbes

Today, researchers from the Houston Methodist Hospital said that they have developed an AI software that is 30 times faster than ordinary doctors when parsing mammograms, and the accuracy rate is as high as 99%. They published the study in the Journal of Cancer. The results of the study showed that this AI software can intuitively translate the X-ray image into diagnostic information, so that doctors can quickly make judgments on the patient's condition so as not to delay the disease.

In order to test the strength of this software, the R&D team conducted a man-machine battle to directly arrange AI with 500 breast cancer patients' mammograms and pathological tissue section reports. At the same time, the team also arranged various related medical discoveries to try to confuse AI. However, this task can hardly live without the power of the AI, and within a few hours it successfully completed its mission. On the other side, the two old drivers of the breast cancer community took 50-70 hours to complete the diagnosis of 50 patients. From this we can imagine how scary the efficiency of this AI software is.

So why do developers think of doing this? The original cruel data shocked them. Data from the US Centers for Disease Control and Prevention (CDC) and the Cancer Society show that about 12.1 million people in the United States receive mammograms every year, but nearly half of them have â€œfalse positiveâ€ suspicions. This has caused a large number of women to undergo breast biopsy for peace of mind (about 1.6 million people per year), of which 20% of women are simply not sick.

In addition to reducing patient suffering and saving physician time, the software can also keep rising costs of cancer. The National Cancer Institute predicts that by 2020, the countryâ€™s money spent on cancer will reach an unprecedented $158 billion, which is an astronomical figure.

According to data from the research firm Grand View, although the correct rate of biopsy results is only 60%-80%, the market size of breast biopsy equipment will reach US$911 million by 2024. Researchers at Johns Hopkins University also said that the amount of money that is wasted on the simplest non-invasive breast cancer each year is as much as $35 million.

Researchers are also full of confidence in AI software. They hope that the software will help doctors better assess the patient's condition. They also hope to reduce the pain of patient biopsy.

Via forbes