Publications
Text queries can be conducted by Author, Title, or Keyword.
The animation of 3D structure-function relationships
Source: Drug News & Perspectives 1990;3(4):197-201.
Author: Toga AW.
Abstract:
Measuring the site and effect of pharmacologically active agents often involves autoradiographic experiments. These experiments generate data that identify the magnitude and location of radioactivity. The results are images on x-ray film. In one form of the experiment, a radioactive analog of some metabolic substrate is administred to the subject, where it becomes trapped in the metabolic pathway proportional to the rate at which it was taken up. Serial sections of the brain can then be place against x-ray film to expose it and provide images of functional anatomy. In another autoradiographic method, serial sections of brain tissue are dipped in radioactive ligands for binding. The exposed film can reflect the location and density of receptors. Naturally, there are many derivatives of these experimental protocols. What they all have in common is visual data. The analysis, synthesis and communication of visual data has become a science in its own right. Neuroimaging, as it is known, is a field devoted to the computations necessary to visualize the complex shapes and forms of brain structure and function derived from autoradiographic (and other) experiments. The science of neuroimgaing, in its quest for more realism, more quantitation and more information about brain structure and function, has succeeded beyond the investigator's capacity to comprehend and assimilate this deluge. Therefore, we have had to employ sophisticated presentation schemes, such as animation, to enhance our appreciation of the data. The science of neuroimgaing and the technique of animation will be discussed here, with particular attention to visualizing brain structure/function relationships. For the sake of brevity, discussions will be limited to data collected following the serial sectioning of brain tissue. COLLECTING THE DATA: In order to examine areas of the brain that lie deep to the cortical surface, the brain must be sectioned into a series of 2D samples. These slices can be analyzed and when properly stacked up provide the data necessary for 3D reconstruction. Although much of the data from the original 3D volume may be lost during the sampling, if enough is retained, it can be used to synthesize a spatially accurate model of the whole brain. RECONSTRUCTING THE ANATOMY: Once the 3D surface has been defined, it can be rendered to form a picture. The algorithms must take into account parameters such as lighting, viewer direction, the anatomic structures themselves and their attributes. Rendering the anatomy gives us not only a quantitative assessment of its shape, size,and location, but also a visual appreciation of it. RECONSTRUCTING THE PHYSIOLOGY: The 3 dimensions of space defining the anatomy can be augmented with the 4th dimension of physiologic measurement. In other words, the spatial model can provide structure for quantitating function. The autoradiographic data provide a 3D array of densitometric information that we an use during reconstruction. One method is to select all those regions that exceed a certain density and plot them as a cloud of points in the anatomic model. Another method is to use the cloud as a framework for reconstructing a model describing the shape and location of the cloud. STRUCTURE & FUNCTION MAPPING: The reconstructions of anatomy and physiology are performed independently even though they may be based upon the same source data. Since our goal is the display of both simultaneously, we must display one without obscuring the other. In a very basic sense, animation uses three dimensions, two spatial and one time. Motion control is the main ingredient of animation and can be defined as the specification of position and orientation of objects in time. These are computed using temporal controls in conjunction with spatial controls. Movements are made within a coordinate system and consist of translations, rotations, and scalings. Animations of neuroanatomy and most neurophysiologic measurements usually are performed by moving the camera location rather than the models themselves.
Source: Drug News & Perspectives 1990;3(4):197-201.
Author: Toga AW.
Abstract:
Measuring the site and effect of pharmacologically active agents often involves autoradiographic experiments. These experiments generate data that identify the magnitude and location of radioactivity. The results are images on x-ray film. In one form of the experiment, a radioactive analog of some metabolic substrate is administred to the subject, where it becomes trapped in the metabolic pathway proportional to the rate at which it was taken up. Serial sections of the brain can then be place against x-ray film to expose it and provide images of functional anatomy. In another autoradiographic method, serial sections of brain tissue are dipped in radioactive ligands for binding. The exposed film can reflect the location and density of receptors. Naturally, there are many derivatives of these experimental protocols. What they all have in common is visual data. The analysis, synthesis and communication of visual data has become a science in its own right. Neuroimaging, as it is known, is a field devoted to the computations necessary to visualize the complex shapes and forms of brain structure and function derived from autoradiographic (and other) experiments. The science of neuroimgaing, in its quest for more realism, more quantitation and more information about brain structure and function, has succeeded beyond the investigator's capacity to comprehend and assimilate this deluge. Therefore, we have had to employ sophisticated presentation schemes, such as animation, to enhance our appreciation of the data. The science of neuroimgaing and the technique of animation will be discussed here, with particular attention to visualizing brain structure/function relationships. For the sake of brevity, discussions will be limited to data collected following the serial sectioning of brain tissue. COLLECTING THE DATA: In order to examine areas of the brain that lie deep to the cortical surface, the brain must be sectioned into a series of 2D samples. These slices can be analyzed and when properly stacked up provide the data necessary for 3D reconstruction. Although much of the data from the original 3D volume may be lost during the sampling, if enough is retained, it can be used to synthesize a spatially accurate model of the whole brain. RECONSTRUCTING THE ANATOMY: Once the 3D surface has been defined, it can be rendered to form a picture. The algorithms must take into account parameters such as lighting, viewer direction, the anatomic structures themselves and their attributes. Rendering the anatomy gives us not only a quantitative assessment of its shape, size,and location, but also a visual appreciation of it. RECONSTRUCTING THE PHYSIOLOGY: The 3 dimensions of space defining the anatomy can be augmented with the 4th dimension of physiologic measurement. In other words, the spatial model can provide structure for quantitating function. The autoradiographic data provide a 3D array of densitometric information that we an use during reconstruction. One method is to select all those regions that exceed a certain density and plot them as a cloud of points in the anatomic model. Another method is to use the cloud as a framework for reconstructing a model describing the shape and location of the cloud. STRUCTURE & FUNCTION MAPPING: The reconstructions of anatomy and physiology are performed independently even though they may be based upon the same source data. Since our goal is the display of both simultaneously, we must display one without obscuring the other. In a very basic sense, animation uses three dimensions, two spatial and one time. Motion control is the main ingredient of animation and can be defined as the specification of position and orientation of objects in time. These are computed using temporal controls in conjunction with spatial controls. Movements are made within a coordinate system and consist of translations, rotations, and scalings. Animations of neuroanatomy and most neurophysiologic measurements usually are performed by moving the camera location rather than the models themselves.