Biomedical subjects

Brian Avants

Publications and source records attributed to Brian Avants.

5 recordsLinked to original sources

PubMed2007

Validation of plaster endocast morphology through 3D CT image analysis.

A crucial component of research on brain evolution has been the comparison of fossil endocranial surfaces with modern human and primate endocrania. The latter have generally been obtained by creating endocasts out of rubber latex shells filled with plaster. The extent to which the method of production introduces errors in endocast replicas is unknown. We demonstrate a powerful method of comparing complex shapes in 3-dimensions (3D) that is broadly applicable to a wide range of paleoanthropological questions. Pairs of virtual endocasts (VEs) created from high-resolution CT scans of corresponding latex/plaster endocasts and their associated crania were rigidly registered (aligned) in 3D space for two Homo sapiens and two Pan troglodytes specimens. Distances between each cranial VE and its corresponding latex/plaster VE were then mapped on a voxel-by-voxel basis. The results show that between 79.7% and 91.0% of the voxels in the four latex/plaster VEs are within 2 mm of their corresponding cranial VEs surfaces. The average error is relatively small, and variation in the pattern of error across the surfaces appears to be generally random overall. However, inferior areas around the cranial base and the temporal poles were somewhat overestimated in both human and chimpanzee specimens, and the area overlaying Broca's area in humans was somewhat underestimated. This study gives an idea of the size of possible error inherent in latex/plaster endocasts, indicating the level of confidence we can have with studies relying on comparisons between them and, e.g., hominid fossil endocasts.

Animals↗

PubMed2004

Geodesic estimation for large deformation anatomical shape averaging and interpolation.

The goal of this research is to promote variational methods for anatomical averaging that operate within the space of the underlying image registration problem. This approach is effective when using the large deformation viscous framework, where linear averaging is not valid, or in the elastic case. The theory behind this novel atlas building algorithm is similar to the traditional pairwise registration problem, but with single image forces replaced by average forces. These group forces drive an average transport ordinary differential equation allowing one to estimate the geodesic that moves an image toward the mean shape configuration. This model gives large deformation atlases that are optimal with respect to the shape manifold as defined by the data and the image registration assumptions. We use the techniques in the large deformation context here, but they also pertain to small deformation atlas construction. Furthermore, a natural, inherently inverse consistent image registration is gained for free, as is a tool for constant arc length geodesic shape interpolation. The geodesic atlas creation algorithm is quantitatively compared to the Euclidean anatomical average to elucidate the need for optimized atlases. The procedures generate improved average representations of highly variable anatomy from distinct populations.

Algorithms↗

PubMed2003

The shape operator for differential analysis of images.

This work provides a new technique for surface oriented volumetric image analysis. The method makes no assumptions about topology, instead constructing a local neighborhood from image information, such as a segmentation or edge map, to define a surface patch. Neighborhood constructions using extrinsic and intrinsic distances are given. This representation allows one to estimate differential properties directly from the image's Gauss map. We develop a novel technique for this purpose which estimates the shape operator and yields both principal directions and curvatures. Only first derivatives need be estimated, making the method numerically stable. We show the use of these measures for multi-scale classification of image structure by the mean and Gaussian curvatures. Finally, we propose to register image volumes by surface curvature. This is particularly useful when geometry is the only variable. To illustrate this, we register binary segmented data by surface curvature, both rigidly and non-rigidly. A novel variant of Demons registration, extensible for use with differentiable similarity metrics, is also applied for deformable curvature-driven registration of medical images.

Algorithms↗

PubMed2003

Characterization of sexual dimorphism in the human corpus callosum.

Despite decades of research, there is still no agreement over the presence of gender-based morphologic differences in the human corpus callosum. We approached the problem using a highly precise computational technique for shape comparison. Starting with a prospectively acquired sample of cranial MRIs of healthy volunteers (age ranges 18-84), the variations of individual callosa are quantified with respect to a reference callosum shape in the form of Jacobian determinant maps derived from the geometric transformations that map the reference callosum into anatomic alignment with the subject callosa. Voxelwise t tests performed over the determinant values demonstrated that females had a larger splenium than males (P < 0.001 uncorrected for multiple comparisons) while males possessed a larger genu (P < 0.001). In addition, pointwise Pearson plots using age as a correlate showed a different pattern of age-related changes in male and female callosa, with female splenia tending to expand more with age, while the male genu tended to contract. Our results demonstrate significant morphologic differences in the corpus callosum between genders and a possible sex difference in the neuro-developmental cycle.

Adolescent↗

PubMedDate not supplied

The correlation of cognitive decline with frontotemporal dementia induced annualized gray matter loss using diffeomorphic morphometry.

This study uses large deformation medical image registration to analyze, in a disease-specific normalized space, the annual rate of gray matter atrophy caused by frontotemporal dementia (FTD) and its correlation with cognitive decline. The analysis consists of three parts. First, a labeled structural MRI atlas is deformed into the shape of an average FTD brain. Second, annualized FTD-related atrophy of gray matter structures is estimated for each patient in the database. Third, the group-wise annualized atrophy rate caused by FTD is correlated, for each gray matter voxel, with declining performance on cognitive tests. This study gives insight into the relationship between FTD-related progressive cortical atrophy and loss in cognitive function.

Aged↗