verbatim): Identification and segmentation of the constituent parts of a human brain in three-dimensional (3-D) magnetic resonance imaging (MRI) studies is a starting point for much of the current research into human brain structure and function. Accurate segmentation and registration allows clinicians and researchers to track and measure changes in the brain over time. Manual brain morphometry is a laborious and time consuming task. The voluminous amount of data and the specialized knowledge required to manually analyze a human brain image preclude clinical studies from rapidly moving forward. In order to address this problem, we propose the creation of a commercial software package that would register and normalize a subject's brain image to a digital atlas of neuroanatomy as well as rigidly align two brain images. Anatomic labels identified on the atlas can then be automatically mapped to segment the corresponding structures in the registered subject image. Normalizing to an atlas allows measurements of brain structure and function from a group of subjects to be compared with respect to the reference space established by the atlas. Rigid alignment of two brain images allows the detection of changes in longitudinal studies. Our software will also provide tools for region-of-interest analysis. We believe that the ability to automatically make structural and functional measurements and to detect changes in the brain is a major advantage of our approach.
There is no commercial software available that allows 3-D MR brain images to be automatically registered to an atlas, along with performing atlas-derived measurements. Our software would fill a niche in the conduct of brain imaging studies in the pharmaceutical industry and address the needs of academic researchers. Part of our commercialization plan will be to provide a brain mapping service for customers who require standardized protocols or who cannot afford to access the necessary resources in terms of personnel, software, hardware, and training. These customers would securely transmit their MRI brain data to MathSoft over the Internet for processing and receive the output electronically. This technology can be extended to the segmentation of other parts of the body. Another commercially attractive direction for future application of this software is screening and characterization for animal imaging studies.
| Gee, James; Ding, Lijun; Xie, Zhiyong et al. (2003) Alzheimer's disease and frontotemporal dementia exhibit distinct atrophy-behavior correlates: a computer-assisted imaging study. Acad Radiol 10:1392-401 |
