Publications
Text queries can be conducted by Author, Title, or Keyword.
Visualization and Warping of Multimodality Brain Imaging
Source: 1994;(6):171-180.
Author: Toga AW.
Abstract:
Visualizing biomedical data is often a prerequisite to its understanding; to see it, is to know it. Visualization enables us to extract meaningful information from complex data sets. Although the brain is the most complex of organs, its multidimensional composition lends itself to a variety of computerized visualization techniques concerned with representation, manipulation and display. If we can visualize a structure, we can identify it. By employing the techniques of image processing, image synthesis and computer graphics, we combine the utility of image and number enabling us to statistically measure the visual representation. Further, such statistical analyses can be extended to compare and correlate a given view of brain with data from other modalities and other subjects. Comparing images across modalities and subjects requires positional and shape transformations to make them occupy the same coordinate space. For example, accurate interpretation of PET or other views of functional anatomy can be improved by correlations with standardized templates. MRI can be used to identify the anatomy on an individual basis and stereotactic atlases can be used as a common reference coordinate system. In addition, the atlas templates can aid in the segmentation of anatomic structures. The development of more generalized representations requires the ability to compare brains from different subjects and depends on the goodness of fit between datasets. However, no single representation, whether it be an average or an atlas, can prove accurate even within a homogeneous population of subjects. Thus it is necessary to warp one brain to conform to another and to quantitate the degree of deformation necessary to make them coincident. This chapter discusses the use of visualization techniques combined with geometric transformation methods to elucidate the relationship between modalities, and between subjects. CONCLUSIONS: Visualizing reconstructed data, warping one dataset to another, and multimodality mappings all have the ability to accentuate or hide the variability that is included within the datasets. Possible sources of variability include error introduced during acquisition, sampling, segmentation, or measurement; normal individual differences; and pathological deviations. The degree to which visualization and warping alters the dataset is affected by the sources of variability, the assumptions of the algorithms, and the product of their execution.
Source: 1994;(6):171-180.
Author: Toga AW.
Abstract:
Visualizing biomedical data is often a prerequisite to its understanding; to see it, is to know it. Visualization enables us to extract meaningful information from complex data sets. Although the brain is the most complex of organs, its multidimensional composition lends itself to a variety of computerized visualization techniques concerned with representation, manipulation and display. If we can visualize a structure, we can identify it. By employing the techniques of image processing, image synthesis and computer graphics, we combine the utility of image and number enabling us to statistically measure the visual representation. Further, such statistical analyses can be extended to compare and correlate a given view of brain with data from other modalities and other subjects. Comparing images across modalities and subjects requires positional and shape transformations to make them occupy the same coordinate space. For example, accurate interpretation of PET or other views of functional anatomy can be improved by correlations with standardized templates. MRI can be used to identify the anatomy on an individual basis and stereotactic atlases can be used as a common reference coordinate system. In addition, the atlas templates can aid in the segmentation of anatomic structures. The development of more generalized representations requires the ability to compare brains from different subjects and depends on the goodness of fit between datasets. However, no single representation, whether it be an average or an atlas, can prove accurate even within a homogeneous population of subjects. Thus it is necessary to warp one brain to conform to another and to quantitate the degree of deformation necessary to make them coincident. This chapter discusses the use of visualization techniques combined with geometric transformation methods to elucidate the relationship between modalities, and between subjects. CONCLUSIONS: Visualizing reconstructed data, warping one dataset to another, and multimodality mappings all have the ability to accentuate or hide the variability that is included within the datasets. Possible sources of variability include error introduced during acquisition, sampling, segmentation, or measurement; normal individual differences; and pathological deviations. The degree to which visualization and warping alters the dataset is affected by the sources of variability, the assumptions of the algorithms, and the product of their execution.