SpineNet: automatically pinpointing classification evidence in spinal MRIs
We describe a method to automatically predict radiological scores in spinal Magnetic Resonance Images (MRIs). Furthermore, we also identify and localize the pathologies that are the reasons for these scores. We term these pathological regions the ``evidence hotspots'. Our contributions are two...
| Main Authors: | , , |
|---|---|
| Format: | Conference item |
| Published: |
Springer International Publishing AG
2016
|