Skip to main content

Multi-Scale Hippocampal Parcellation Improves Atlas-Based Segmentation Accuracy

Posted by on Friday, February 3, 2017 in News.

Andrew J. Plassard, Maureen McHugo, Stephan Heckers, Bennett A. Landman. “Multi-Scale Hippocampal Parcellation Improves Atlas-Based Segmentation Accuracy” In Proceedings of the SPIE Medical Imaging Conference. Orlando, Florida, February 2017.

Full Text:


The hippocampus is one of the most studied regions of the brain. Recent advances in MRI have produced high-contrast imaging of the hippocampus. In this work, we use a population of 195 atlases based on T1-weighted MR images with the left and right hippocampus delineated into the head and body. We initialize the multi-atlas segmentation to each lateralized hippocampus to both speed up and improve the accuracy of registration. The segmentation results in a Dice similarity coefficient over 0.9 for the full hippocampus. The head and body delineation resulted in a Dice coefficient over 0.87 for both structures.


Qualitative segmentation results.
Qualitative segmentation results.