Radiologists It is not easy: many of them have to analyze numerous MRI scans (magnetic resonance imaging) of patients every day. In a single tomography more than 2,000 photographs are taken, and in search of tumors examine doctors previously each recording manually. A sometimes tedious task in now to help Artificial Intelligence: AI fuse, a startup based in Hamburg, has developed a system that can detect cancer on MRI scans and classified – and thus helps in the diagnosis.
The idea for the artificial intelligence that makes diagnosis more precise and reliable, the founders came in app development in the medical field. “In the course of time, increasingly clear that artificial intelligence in medicine would gain in importance,” says Maximilian Waschka, one of the founders of fuse AI. “Hence the idea, physicians use to assist in the diagnosis of machine learning specifically using neural networks emerged.”
Machine Learning from the cloud
Here’s how it works: The MRI scans are sent over an encrypted connection to the certified data centers Telekom to Saxony Anhalt. There, the artificial intelligence of fuse AI analyzes each image – much faster than a human could. If they find something, it estimates beyond even if it could be a benign or a malignant tumor. This information is marked in the image and encrypts the doctor sent back into practice. This helps radiologists to recognize distinctive tissue changes faster and more reliable.
The processing power for intelligent algorithm that analyzes the MRI scans, the northern German company from the open TelekomCloud relates. A decisive advantage for the North German entrepreneur is the flexibility: “Our demand fluctuates. Only for image analysis, we need very high capacity, “says Dirk Schäfer, an expert in machine learning and co-founder of Fuse-AI. “Such reproach IT resources permanently needed for such analyzes would be very inefficient.”
Next step: skin cancer detection from the cloud
The principle fuse AI want to transfer in the near future to further diagnostic processes. For example, the skin cancer detection. To this end, the company has developed a so-called dermatoscope together with dermatologists that scans the skin. And the images that arise here are analyzed in the same manner of artificial intelligence in the cloud.
“We are helping to using Open TelekomCloud to recognize many common diseases more reliable and faster,” says Sabrina Reimers Kipping, a PhD biochemists and co-founder of Fuse-AI. “We not only reduce the effort and cost, but also increase thanks to artificial intelligence also the chances of recovery for patients.”
Fuse-AI: e-health from the cloud
Additional information about Fuse-AI and big data analytics solutions for medical diagnosis from the open TelekomCloud can be found at T-Systems in the Internet.