Facebook and the team at New York University School of Medicine are developing research that uses the potential of artificial intelligence to improve the dynamics of an MRI scan.
They want to revolutionize the mechanics of MRI (magnetic resonance imaging) and promise to speed up the examination up to 10 times. Currently, this type of study can take from 15 minutes to an hour, making it a slow and complicated process for some patients.
The reason for this is that they have to collect a lot of data and information in order to create images in the form of 3D blocks. The Facebook and NYU projects, on the other hand, want to accelerate the process of image creation and follow a different dynamic.
Simply put, their proposal is to collect less data, reduce scanning time and use the AI to close gaps:
The key is to train artificial neural networks to recognize the underlying structure of the images to complete the skipped views in the accelerated scan.
To understand this point, Facebook illustrates it with the process we use to process sensory information…. we get an incomplete picture of what we experience, and then the brain fills in these blind spots.
This research began in 2015 and on the basis of the tests carried out, the results are promising. However, it is still in its infancy and it will take much longer to see whether the project will be successful.