Applying tensor-based morphometry to parametric surfaces can improve MRI-based disease diagnosis

• Published in: NeuroImage (Orlando, Fla.) Vol. 74; pp. 209 - 230
• Main Authors: Wang, Yalin, Yuan, Lei, Shi, Jie, Greve, Alexander, Ye, Jieping, Toga, Arthur W., Reiss, Allan L., Thompson, Paul M.
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier Inc 01.07.2013; Elsevier; Elsevier Limited
• Subjects: Accuracy; Algorithms; Alzheimer's disease; Biological and medical sciences; Biomarkers; Brain; Brain - pathology; Brain Mapping - methods; Classification; Fundamental and applied biological sciences. Psychology; Humans; Image Interpretation, Computer-Assisted - methods; Imaging, Three-Dimensional - methods; Magnetic Resonance Imaging - methods; Medical imaging; Sparse learning; Stability selection; Statistics; Surface conformal parameterization; Surface registration; Tensor-based morphometry; Tomography; Vertebrates: nervous system and sense organs; Williams Syndrome - pathology; Surface conformal parameterization; Stability selection; Tensor-based morphometry; Sparse learning; Surface registration; Learning; Acquisition process; Morphometry; Nuclear magnetic resonance imaging
• ISSN: 1053-8119; 1095-9572; 1095-9572
• DOI: 10.1016/j.neuroimage.2013.02.011

Many methods have been proposed for computer-assisted diagnostic classification. Full tensor information and machine learning with 3D maps derived from brain images may help detect subtle differences or classify subjects into different groups. Here we develop a new approach to apply tensor-based morphometry to parametric surface models for diagnostic classification. We use this approach to identify cortical surface features for use in diagnostic classifiers. First, with holomorphic 1-forms, we compute an efficient and accurate conformal mapping from a multiply connected mesh to the so-called slit domain. Next, the surface parameterization approach provides a natural way to register anatomical surfaces across subjects using a constrained harmonic map. To analyze anatomical differences, we then analyze the full Riemannian surface metric tensors, which retain multivariate information on local surface geometry. As the number of voxels in a 3D image is large, sparse learning is a promising method to select a subset of imaging features and to improve classification accuracy. Focusing on vertices with greatest effect sizes, we train a diagnostic classifier using the surface features selected by an L1-norm based sparse learning method. Stability selection is applied to validate the selected feature sets. We tested the algorithm on MRI-derived cortical surfaces from 42 subjects with genetically confirmed Williams syndrome and 40 age-matched controls, multivariate statistics on the local tensors gave greater effect sizes for detecting group differences relative to other TBM-based statistics including analysis of the Jacobian determinant and the largest eigenvalue of the surface metric. Our method also gave reasonable classification results relative to the Jacobian determinant, the pair of eigenvalues of the Jacobian matrix and volume features. This analysis pipeline may boost the power of morphometry studies, and may assist with image-based classification. ► Cortical surface parameterization with slit map conformal mapping ► A sparse learning based method for surface feature selection and classification ► Surface mTBM achieved better group difference results than other surface statistics. ► Stability selection demonstrated consistent feature selection results. ► An automated and robust cortical surface registration and classification system