Why this is important is because radiologists tend to go through many images of brain scans on a daily basis. Through these images, the radiologists have to look for tiny abnormalities that could point towards brain hemorrhages that could be potentially fatal if not caught in time or detected.
However, radiologists are still human at the end of the day, and sometimes they might miss things. This is why the algorithm was developed, where it would improve on the detection and make the overall process more efficient. When it detects a potential abnormality, it will then alert radiologists who can then examine these scans more closely.
According to one of the researchers, Esther Yuh, “We wanted something that was practical, and for this technology to be useful clinically, the accuracy level needs to be close to perfect. The performance bar is high for this application, due to the potential consequences of a missed abnormality, and people won’t tolerate less than human performance or accuracy.”