Prof. Yonina Eldar Uses AI Tools to Diagnose COVID-19



In this video, Prof. Yonina Eldar (Department of Computer Science and Applied Mathematics; Head, Biomedical Engineering & Signal Processing Center) discusses her development of novel methods for diagnosing COVID.

A characteristic of COVID infection are “ground glass” spots in the lungs – however, the spots could also indicate, say, the flu. Prof. Eldar’s team of doctors, scientists, data analysts, and other experts are devising AI-based algorithms that can assess lung images and determine whether the spots are due to COVID.

So far, she says, they have about a 90% accurate rate of diagnosis; the widely used PCR tests are about 70% accurate.

Her team is also refining the use of ultrasounds for diagnosis – a tool that will help not only with COVID, but other lung diseases.