"Statistical Models for Medical Image Analysis"
Ph.D. Defense, May 2000.
This talk presents a novel medical image segmentation method that
incorporates prior statistical models of intensity, local curvature, and
global shape to direct a deformable surface toward a likely segmentation.
| |  | |
"Statistical Models in Medical Image Analysis"
High level overview of dissertation. (38M PowerPoint file)
This talk presents an overview of various approaches of using
statistical models for problems in medical image analysis, such as
medical image segmentation and registration.
| |  | |
"Statistical Shape Influence in Geodesic Active Contours"
Computer Vision and Pattern Recognition, June 2000. Awarded best
student paper
This talk introduces a representation for deformable shapes and
defines a probability distribution over the variances of a set of
training shapes. The segmentation process embeds an initial curve as
the zero level set of a higher dimensional surface, and evolves the
surface such that the zero level set converges on the boundary of the
object to be segmented.
| | | |
"Level Set Based Segmentation with Intensity and Curvature Priors"
Mathematical Methods in Biomedical Image Analysis, June 2000.
This talk presents a method for segmentation of anatomical structures
that incorporates prior information about the intensity and curvature
profile of the structure from a training set of images and boundaries.
| | | |
"Methods of Conformational Search for Protein-Ligand Docking"
MIT Area Exam Presentation, February 2000.
This talk presents a review of four papers addressing the problem of
flexible-ligand docking. Four stages of the docking problem are
conformational search, query matching, energy minimization, and
conformational clustering. Various approaches to each of these steps
are discussed and compared.
| | | |
"Human Tracking From Video Using Linear Combinations of Motion"
This talk hilights a method of fitting a 3D person-model to a video
sequence of a person moving around. Our method expresses human
motions as a linear combination of training motions, from which we can
extract 3D information about the position of the person's arms and
legs.
| | | |
"High-Resolution Anatomical Model Generation Using Multiple Internal Scans"
Information Processing in Medical Imaging, June 1999
This talk describes a method of combining two orthogonal medical
images to generate a 3D surface model of anatomical structures with
higher resolution than models created from either of the scans alone.
The two scans are first registered to each other and then a net of
linked surface nodes is initialized for each of the scans. The nodes
from the two nets are then merged and relaxed, subject to constraints
set by the resolution of each scan. This generates a smooth surface
representation which stays faithful to the original binary data.
| | | |
