Nvidia wants to revolutionize medical scans with its new $ 1 organ recognition AI

X-rays, tomography and magnetic resonance imaging are today one of the key tools to detect various injuries and diseases, however, their use is still far from being accessible to everyone. Today Nvidia seeks to make an important leap in this segment, which incidentally is its first intervention in the medical field.

Nvidia is partnering with GE Healthcare to provide its Quadro GP100 GPUs and deep learning and artificial intelligence platform for medical patient scanning. Such technology is currently present in various autonomous car projects and is now to be released in hospitals.

The real change is the price

GE Healthcare is reporting that over the next few days they will be updating more than 500,000 of their medical devices to incorporate Nvidia's AI platform, ranging from scanners, MRIs, CTs and more. With this, it will be possible to have better clinical results in the detection of liver, kidney and even brain injuries, since the speed of the information will allow an almost immediate diagnosis.

Another important point is that each of these scans will cost approximately one dollar per image, that is, a radical change since today such a scan is between 150 and 250 dollars. This seeks to have a greater reach and thus be able to reach low-resource areas.

However, it is not all good news, since the use of artificial intelligence in these devices has the drawback that each hospital will generate more than 50,000 TB of data every day, so they will need a robust and fast storage platform, so some Hospitals are opting for cloud solutions. The curious thing about this is that of all the data generated, only 3% is used in diagnostics, since the rest are in-depth details captured by the devices and processed by the AI, which could serve to carry out more specific analyzes .

Along with Nvidia, GE Healthcare is also leaning on Intel, which will be responsible for providing Xeon servers that will help image processing and computing power, so that radiologists will be able to have the information quickly and efficiently.

More information | Nvidia

Share none:  Mobile Science Our-Selection 

Interesting Articles

add