The Gayfemboy botnet exploited a zero-day vulnerability in routers to infect more than 15,000 devices in various countries. It is a serious threat because it can carry out massive DDoS attacks with traffic volume of up to 100 GB.
38 
Artificial intelligence (AI) significantly improves the detection of breast cancer during screenings, according to the results of the first real-world study. The scientists found that thanks to the use of AI, one additional case of cancer was detected for every thousand women examined.
During the study, published in the journal *Nature Medicine*, scientists checked the results of almost 500,000 examinations of women in Germany. Half of these mammograms were analyzed with the help of artificial intelligence, while the other part was analyzed using the traditional method of double analysis by radiologists. The AI detected 6.7 cases of cancer per 1,000 women, one point higher than the traditional method’s rate of 5.7 cases. At the same time, the number of repeated calls for additional examinations remained the same, which indicates the absence of an increase in false positive results.
Professor Oleksandr Katalinich from the University of Lübeck, one of the authors of the study, noted that thanks to the use of AI, it was possible not only to improve the detection of the disease, but also to reduce the burden on radiologists, who in Germany annually analyze 24 million images and record more than 78 thousand new cases of breast cancer .
The PRAIM project, within which the study was conducted, covered the period from 2021 to 2023 and included data from more than 460,000 women from 12 screening centers in Germany. The scientists sought to confirm that assessments made with the help of AI are no less accurate than traditional methods. However, the results exceeded their expectations. This opens up prospects for further integration of technologies into medical systems.
The study proved that the use of artificial intelligence in breast cancer screening allows to detect more cases of the disease without increasing the number of false positive results and additional calls for examination. In the future, this can significantly improve the quality of medical diagnostics and optimize the work of radiologists.
38 
37 
56