The objective of this research is the wide accessibility of 3D deconvolution (computational deblurring and visual clarification of 3D image data) for the neuroscientific community and other life-science communities. Long-term aims are to provide the most reliable, robust and quantitatively accurate system for 3D visualization and morphometry, and to provide neuroscientists and other life-scientists with easy-to-use tools for studying the structure and function of normal and pathologic tissue. The focus of this project is on neuroscientific applications. The commercial objective is a profitable software product, which, owing to innovations, well outperforms competitive products in reliability, robustness, quantitative accuracy, speed and ease of usage.
The specific aims of this project are: (1) To develop and test variations of the algorithm for a broader range of widely used confocal microscope types, including the slit-scan and array-detector geomeries; (2) To improve (and test) the robustness of the algorithm under a number of common adverse conditions; (3) To further validate the correctness of the image reconstructions by using fabricated test objects of known geometry; (4) To demonstrate the system on two of the most suitable computer platforms for wide usage. These are the Silicon Graphics unix-based platform and the IBM-PC compatible platform; (5) To beta test prototypes of the product at several potential customer sites. Innovations introduced by us in this research include: (1), a blind deconvolution approach, which obviates the need to measure the point spread function, (2), a maximum likelihood optimization approach, which makes the methodology robust against photodetector noise and other adverse conditions, and (3), a unified underlying mathematical model, which makes the algorithm easily adaptable among confocal geometries.
We estimate on the order of 800 biological pinhole confocal microscopes and 150 biological slit scan confocal microscopes to exist worldwide. We expect these numbers to increase to 8,000 confocal microscopes over 10 years. We expect that 1/2 of these potentIal sites will purchase deconvolution software. We should secure at least 1/3rd of this potentIal market and thereby (conservatively) Bell at,least 100 units per year over 10 years. Numerous nonblologlcal spin-off markets (e.g., pharmeceuticals, chemicals) exist as well.
| Holmes, T J; O'Connor, N J (2000) Blind deconvolution of 3D transmitted light brightfield micrographs. J Microsc 200:114-27 |
