Detection of cerebral aneurysms using artificial intelligence
Slovak startup Deep Art Medicine is developing a tool for detecting cerebral aneurysms from magnetic resonance imaging. They are collaborating on the project with experts from UMC Utrecht, a hospital in Copenhagen, and Univerzitná nemocnica Martin. The goal is to alert physicians to suspicious findings more quickly and clearly. The project was presented by Tomáš Koščo and Denis Kutnár, who serves as CTO and a researcher in artificial intelligence. The team is working with the University Hospital in Martin on automatic detection of aneurysms from time-of-flight MRI. The solution is based on a neural network with an encoder–decoder architecture that generates a probability map of a finding from the image. Before training, data collection and annotation took place, followed by “bias field” correction, normalization, and selection of suitable hyperparameters.Who is behind it and what they are solving
Tomáš Koščo
