Tests of the algorithm were determined by MRI disease faster cardiologist-person
In the US, two former employees of the hedge Fund created a mathematical model that allows to accurately diagnose heart disease by drawing an MRI.
The algorithm was created as part of the competition, which was attended by almost a thousand teams. Winners – Tence Ki Lee and Liu have created your project based on the concept of convolutional neural network that includes a set of machine learning algorithms, Deep Learning, or “deep learning”.
In particular, convolutional neural network mimics the processing of visual data in mammals.
During early testing of the algorithm was determined by MRI disease quite effectively: better than the cardiologist-person, but much faster. The doctors, the new technology can be very valuable for industry visualization and scanning of internal organs.
“Some are working on projects like this for 15 years, I am amazed that the three-month contest has produced such results”, – said in an interview with the Financial Times, the head of Department of complex imaging of the vascular system of the National institutes of health Andrew Arai.
The winner of Tencia Ki Lee and Liu in the past worked as financiers. Some years ago he retired from the hedge Fund Crabel Capital Management, and Qi Liu previously worked in the Fund Two Sigma.
According to Retention, they used the software, are taken from public sources. The main part was downloaded from GitHub.
Created super-sensitive sensor for diagnosing cancer