Publications
Text queries can be conducted by Author, Title, or Keyword.
Image Analysis of Brain Physiology
Source: IEEEE Computer Graphics amd Applications 1985 Dec;5(12):20-25.
Author: Toga AW, Arnicar TL.
Abstract:
By observing changes in brain activity in relationship to changes in behavior, neuroscientists can further the understanding of the function of the brain. The data used in this research is in the form of autoradiograms, made on brain slices. Digital image processing is the only practical way to analyze the large number of slices required. However, if image processing technologies are to significantly contribute to the study of brain function, they must be able to accurately and automatically align sections taken from the brain; establish quantitative relationships among different pictorial representations of the brain; and provide a 3D visualization of quantitative brain activity in the whole brain. The technique described here uses fiduciary marks to align brain slices; thresholding and look-up-table techniques to correlate images obtained with different technologies; and new methods to genereate a 3D image. The development and refinement of these methods will greatly aid in the study of brain structure and function. METHODS: The procedure for generating the fiduciary marks to be used in the alignment algorithms simply requries the placement of 2 columns alongside the brain before it is cut. 2 tubes are filled with radioactive and stainable material. A guide provides the mechanism to ensure that the fiduciary marks are vertical and parallel. The whole brain and the 2 tubes are then surrounded with cold embedding matrix and quick-frozen. Brain sections are cut 20u thick on a freezing microtome throughout the entire brain. Each section contains the brain tissue and the slices of radioactive tubing. The sections are then processed for autoradiography and histology. Because each brain section now contains fiduciary marks for image registration, the alignment algorithms can be greatly simplified and designed with minimal user interaction. CONCLUSION: Analyzing images of brain physiology poses three problems which we addressed in this article: placing successive brain sections in register, correlating images obtained using different methods, and displaying this data in a meaningful 3D model. This work is significant because of the need to understand basic structure-function relationships throughout the brain. Prior to the application of computer and digital imaging technologies, the study of the brain was reductionistic: The neuroscientist was able to examine only isolated parts of a whole working system. This was because of the vast quantity of data, and the inability to understand completely the relationships among the structural components. Similarly, it was difficult to correlate the various activities of those components. Because computer graphics and imaging technologies now provide the tools to accomplish these goals, the brain can be studied holistically, and a more comprehensive understanding of the mechanisms responsible for behavior is possible.
Source: IEEEE Computer Graphics amd Applications 1985 Dec;5(12):20-25.
Author: Toga AW, Arnicar TL.
Abstract:
By observing changes in brain activity in relationship to changes in behavior, neuroscientists can further the understanding of the function of the brain. The data used in this research is in the form of autoradiograms, made on brain slices. Digital image processing is the only practical way to analyze the large number of slices required. However, if image processing technologies are to significantly contribute to the study of brain function, they must be able to accurately and automatically align sections taken from the brain; establish quantitative relationships among different pictorial representations of the brain; and provide a 3D visualization of quantitative brain activity in the whole brain. The technique described here uses fiduciary marks to align brain slices; thresholding and look-up-table techniques to correlate images obtained with different technologies; and new methods to genereate a 3D image. The development and refinement of these methods will greatly aid in the study of brain structure and function. METHODS: The procedure for generating the fiduciary marks to be used in the alignment algorithms simply requries the placement of 2 columns alongside the brain before it is cut. 2 tubes are filled with radioactive and stainable material. A guide provides the mechanism to ensure that the fiduciary marks are vertical and parallel. The whole brain and the 2 tubes are then surrounded with cold embedding matrix and quick-frozen. Brain sections are cut 20u thick on a freezing microtome throughout the entire brain. Each section contains the brain tissue and the slices of radioactive tubing. The sections are then processed for autoradiography and histology. Because each brain section now contains fiduciary marks for image registration, the alignment algorithms can be greatly simplified and designed with minimal user interaction. CONCLUSION: Analyzing images of brain physiology poses three problems which we addressed in this article: placing successive brain sections in register, correlating images obtained using different methods, and displaying this data in a meaningful 3D model. This work is significant because of the need to understand basic structure-function relationships throughout the brain. Prior to the application of computer and digital imaging technologies, the study of the brain was reductionistic: The neuroscientist was able to examine only isolated parts of a whole working system. This was because of the vast quantity of data, and the inability to understand completely the relationships among the structural components. Similarly, it was difficult to correlate the various activities of those components. Because computer graphics and imaging technologies now provide the tools to accomplish these goals, the brain can be studied holistically, and a more comprehensive understanding of the mechanisms responsible for behavior is possible.