Developing AI Models In Scale

Short Talk (20 Minutes)

Stroke is the number one long livity disabler in the world. Over one million people in the United States alone are suffering from stroke every year. Unfortunately we all know someone who had a stroke (Tzvika peak, Arik Sharon, Daenerys Targaryen).
An ischemic stroke is a blockage of vessels in the brain, which causes a limited oxygen transfer. WIth the lack of oxygen, the brian starts to decay and die. With every minute that goes by, 2 millions of neurons die. With every minute that goes by, another week of the patient’s life is gone. Treatment needs to be accurate and fast.
Viz enables better patient care by bringing the right doctor with the right patient at the right time. We use deep learning based AI models to identify stroke. When a patient is undergoing a CT scan, we get the CT imaging data, process it and notify the treating doctor when needed.
Deep learning is all about teaching the computer to identify patterns data. We provider positive and negative signals regarding the accuracy of an algorithm results to the computer, having it learning bit by bit our desired result. We show it CT scans of normal people and of those having strokes, teaching it to distinguish between the two.
In order to make our algorithm more accurate, we constantly need to feed it with new data. The more data the algorithm see, the better it will handle new data.
An AI models diagnosys takes us around 7 minutes to process on a strong CPU and 2.5 minutes over a GPU. A CT scan is around 2GB pe
~/event sponsors
Platinum Sponsors
Gold Sponsors
Silver Sponsors
Food & Swag Sponsors
Community Partners
Event Organizer
Learn more about each of our Event Sponsors.