AWS Startups Blog Uses Deep Learning to Analyze CT Scans and Save Lives

When someone suffers a stroke, every seconds counts and could mean the difference between being ok, permanent brain damage, or even death. With that taken into consideration, the current process for deciding how to treat the victim of a stroke just seems nonsensical. Typically, once a victim is rushed to the closest hospital, they are given a CT scan. A specialized doctor then needs to review the results of the scan, and that doctor is often located at a different hospital. The multi-step process of sending the scan, it being reviewed by the doctor, and the returning of their analysis, can take extremely long, with countless neurons dying in the patient’s brain every second. is on a mission to reduce that timeline, which sometimes takes up to 3 hours, down to just 3 minutes by leveraging specially trained deep learning models.

For both of the company’s cofounders, this problem and finding a solution is deeply personal. Chris Mansi, who serves as’s CEO, previously worked as a neurosurgeon, where we witnessed the inefficient process first hand. As for CTO David Golan, he saw it from the patient’s perspective after suffering a stroke himself as a postdoc at Stanford. Check out the above interview with Golan to learn more about the company and how they’re looking to save lives by leveraging AWS and deep learning.