To date, MR scanners operate vastly independent from each other and acquire all MR data required for each image individually. However, it becomes ever more clear that in order to extract the required information from the least amount of acquired data, MR scanners have to work in a network and at the same time also as a network of various types of sensors. This because on the one hand the interpretation or reconstruction of data can be greatly boosted by statistical information learned from vastly greater data sets as it is impressively demonstrated by very recent and topical machine learning work. On the other hand, the statistical power of the available data is greatly leveraged by additional information describing the parameter and the context of the acquisition. Although all implications of these theoretical relations cannot be conclusively discussed on the background of the very fast moving technologies in this sector, the statements from above can be extrapolated by seeing the most recent development in the general trend of data reductionism.

Discussion points of the webinar:

How can MRI systems respond to these needs?

How do recent developments alter requirements of MR platforms?

What are the requirements set by data-driven approaches?

by David Brunner, Dr. sc.

CSO and Cofounder at Skope