The Labeling of Cortical Sulci using Multidimensional Scaling

Meena Mani,Anuj Srivastava,Christian Barillot

doi:10.54294/qo3app

Abstract

The task of classifying or labeling cortical sulci is made difficult by the fact that individual sulci may not have unique distinguishing features and usually need to be identified by a multivariate feature set that takes the relative spatial arrangement into account. In this paper, classical multidimensional scaling (MDS), which gives a geometric interpretation to input dissimilarity data, is used to classify 180 sulci drawn from the ten major classes of sulci. Using a leave-one-out validation strategy, we acheive a success rate of 100% in the best case and 78% in the worst case. For these more difficult cases, we propose a second stage of classification using shape based features. One of these features is the geodesic distance between sulcal curves obtained from a new open curve representation in a geometric framework. With MDS, we oﬀer a simple and intuitive approach to a challenging problem. Not only can we easily separate left and right brain sulci, but we also narrow the classification problem from, in this case, a 10-class to a 2-class problem. More generally, we can identify a region-of-interest (ROI) within which one can carry out further classification.

Highlights

The large variability present in the human brain cortex makes it one of the most challenging of brain anatomies to segment, label, classify or model
We used a database of 18 T1-MR 3D SPGR brain images of healthy subjects, matched for sex, handedness, and age (35 ± 10 years)
The central, postcentral, superior frontal, sylvian fissure and superior temporal, were extracted from each hemisphere using the method described in [2, 9]. This gave us 18 × 10 = 180 sulci for classification. These primary sulci were chosen because they can consistently be identified in all normal individuals

Summary

Introduction

The large variability present in the human brain cortex makes it one of the most challenging of brain anatomies to segment, label, classify or model. The underlying architectonic and functional organization of the brain influences to some degree the size, shape and placement of the folding patterns [1]. The length, depth, the pure shape (i.e. after removing scaling, rotation and translation), the spatial arrangement and orientation of some of these individual sulci are characteristic features shared across a population. These features can form the basis for classification, labeling and the study of selective differentiation of tissue that results from neurodegenrative pathologies, the action of genes or even as a result of cognitive activity. This study fills an important need and can lead to such wide-ranging applications as landmark or guidance-based neurosurgical procedures or the differential diagnosis of degenerative brain disease

Objectives

Results

Conclusion