AI Disease Detection

Brain Tumours
Pneumonia

Inspiration

Every year, about 1 in 20 people die every year as a result of misdiagnosis, in North America alone. Or to put that into a different perspective, that's over 300,000 deaths that could have been prevented in 2019. The truth is doctors and medical staff, in general, are constantly overworked. With the pandemic, we saw a huge shortage of staff, and those who were working had to put in a ton of extra hours. Working such long hours for a job that requires extremely high attention. It really isn't a surprise that we have so many misdiagnoses every year. Performing at such a high level is not only near impossible but mistakes in scenarios like these mean life and death for patients. Well, the question is, how do we actually remove a huge weight of this responsibility off their hands?

What it does

We designed an AI algorithm to actually take patient ultrasounds or any scan in general and use machine learning to train the module to detect some of the more common issues we see in hospitals. Specifically, use a piece of technology known as Convolutional Neural Networks.