Phase II Research Projects
Research Theme - Biomedical Imaging
Investigator: Dr. Chandra Kambhamettu, UD Computer & Information Sciences
Mentor: Karl V. Steiner, Ph.D., DBI, UD Electrical & Computer Engineering
Research Title: 3-D Image Analysis Algorithms for Automatic Computation of Ground Glass Opacity of Lung Tumors
Abstract: This proposal aims at the study, design and development of image analysis and visualization tools for assistance in the radiologic analysis of high-resolution computed tomographic images in order to automatically detect, extract, time-align lung nodules, and classify them with ground-glass opacity (GGO) ratios. Ground-glass refers to the HRCT appearance of a hazy opacity that does not obscure the associated pulmonary vessels [4]. The current research and software are able to detect and measure lung nodules using image information, however not capable of comparing measurements across different times after alignment, or estimating the GGO overage of nodules. The most recent works in this area [1][5] use manual analysis for GGO coverage. None of the works offer to understand the evolution of the tumor between CT data taken at different time instances (typically stretched apart by 3 months).
This motivates our proposed work, to develop a system for objective assessment of GGO regions using novel, state-of-the-art image analysis methods and provide time-alignment tools to analyze CT data taken at different times. This allows the radiologist to see not only the evolution of a tumor, but also the measurement changes including GGO coverage changes and the changes in orientation, maximum areas of growth, etc. We aim to extract the rigid/nonrigid changes of the tumor between different times. We also propose to develop Desktop and Virtual Reality visualization tools to show the evolution of tumors along with respective GGO area of coverage, represented in different colors based on extracted opacity information. Development of proposed tools is focused to assist in selection of candidates for curative limited resection. This work is a collaboration between the UD-Video/Image Modeling and Synthesis (VIMS) Lab., Christiana Care, and the DBI Bioinformatics Center.
Investigator: Dr. Fengshan Liu, DSU Applied Mathematics and Theoretical Physics
Research Title: Analysis of 3-D Breast Surface Deformation during Mammographic Compression
Abstract: The aims of this proposal are: 1) to develop a breast compression analysis system including a physical framework for breast compression analysis and a software package for computing 3D breast surface images and compression force distribution, using optical digitization images; 2) to evaluate the compression analysis system including the accuracy of computing 3D surfaces and the accuracy of computing breast deformations and compression forces, using a simple physical phantom; and 3) to perform a pilot clinical study including acquiring optical digitization breast images from 10-20 volunteers and performing statistical analysis of breast deformations and compression forces. We will also train our students by involving them in the proposed research.
