Automatic Rodent Brain Extraction Based on a Deformable Surface Model
Open AccessWhile pre-clinical MR imaging studies on small animals, such as rats or guinea pigs, are on the rise, fully automatic image processing techniques to analyze the neuroimaging data are still lacking. The extraction of the brain from a neuroanatomic MRI image, also called skull stripping, is a critical step for image analysis because further image processing tasks, including brain registration and segmentation, rely on the extracted brain images. In this dissertation, we present an automatic rat brain extraction method, Rat Brain Deformable model method (RBD), which adapts the popular human Brain Extraction Tool (BET) by incorporating information on brain geometry and MR image characteristics of the rat brain. The robustness of the method was demonstrated on T2-weighted MR images of 64 rats and compared with other brain extraction methods (BET, PCNN, PCNN-3D). Results demonstrate that RBD reliably extracts the rat brains with high accuracy (>92% volume overlap) and is robust against signal inhomogeneity in the images. I then extended this work to develop the Guinea pig Brain Deformable surface model (GBD) for guinea pig brain extraction. GBD is an extension of the RBD method with adjusted parameters to suit the brain geometry difference between guinea pig brain and the rat brain. GBD also includes a template-based registration process to take initialization parameters from a pre-defined template, with improved robustness against improper animal placement in the MR scanner during image acquisition. GBD provided an overall brain extraction accuracy of above 90% volume overlap with the manually defined brain contour, which served as the ground truth, on a set of 20 guinea pig brains. Future work will include extending the method further to other rodent types, such as mice.
- All rights reserved
Notice to Authors
If you are the author of this work and you have any questions about the information on this page, please use the Contact form to get in touch with us.