Extraction of the Plane of Minimal Cross-Sectional Area of the Corpus Callosum using Template-Driven Segmentation Neda Changizi 1 , Ghassan Hamarneh 1 , Omer Ishaq 1,2 , Aaron Ward 1,3 , and Roger Tam 4 1 Medical Image Analysis Lab, Simon Fraser University, Canada 2 Department of Computer Sciences, Air University, Pakistan 3 Robarts Research Institute, The University of Western Ontario, Canada 4 MS/MRI Research Group, University of British Columbia, Canada Abstract. Changes in corpus callosum (CC) size are typically quanti- fied in clinical studies by measuring the CC cross-sectional area on a midsagittal plane. We propose an alternative measurement plane based on the role of the CC as a bottleneck structure in determining the rate of interhemispheric neural transmission. We designate this plane as the Minimum Corpus Callosum Area Plane (MCCAP), which captures the cross section of the CC that best represents an upper bound on inter- hemispheric transmission. Our MCCAP extraction method uses a nested optimization framework, segmenting the CC as it appears on each candi- date plane, using registration-based segmentation. We demonstrate the robust convergence and high accuracy of our method for magnetic res- onance images and present preliminary clinical results showing higher sensitivity to disease-induced atrophy. 1 INTRODUCTION The corpus callosum (CC) is an anatomic structure that acts as a communication bridge connecting the two brain hemispheres [1]. Certain neurological diseases are known to affect the shape and size of the CC. In particular, there have been numerous studies correlating CC measurements to multiple sclerosis [2], schizophrenia [3], autism [4], and many other mental and physical ailments. The accurate measurement of CC area changes is dependent on the repeatable iden- tification of the cross-sectional plane of interest in all studied images. In previous studies, changes in CC size have been quantified by measuring its sagittal cross- sectional area on a midsagittal plane (MSP). MSP identification approaches can be classified as either symmetry- or feature- based. Symmetry-based approaches assume bilaterally symmetric hemispheres, with the MSP chosen to maximize this symmetry. Published approaches sug- gest different symmetry criteria; e.g., intensity ratios [5], cross-correlation [6], or edge-based [7]. In feature-based approaches, the MSP is defined as the plane best matching the cerebral interhemispheric fissure. The Hough transform [8],