Resolution-adaptive networks for uniform performance
across heterogeneous medical image cohorts

In the setting of clinical imaging, differences in between vendors, hospitals and sequences can yield highly inhomogeneous imaging data. In MRI imaging in particular, voxel dimension, slice spacing and acquisition plane can vary dramatically. The usual strategy to deal with heterogeneity of resolution isharmonization by resampling to a common spatial resolution. This can lead to loss of fidelity arising from interpolation artifacts out-of-plane and downsampling in-plane. We propose a network architecture designed to be able to learn directly from spatially heterogeneous data, without resampling: a segmentation network based on a geometric learning framework that leverages a spherical harmonic, rather than voxel-grid, parameterization of convolutional kernels.

The lecture will be held in sitem-insel (Room O2.211), Freiburgstr. 3, Bern broadcast virtually via Zoom at

https://unibe-ch.zoom.us/j/61079804322

Please feel free to spread the word to anybody potentially interested.

For further info please contact: bernd.jung@insel.ch or jessica.bastiaansen@unibe.ch

Join the mailing list here: TIC Seminar