Large-data-size fluorescent molecule tomography reconstruction method

Abstract

The invention relates to a large-data-size fluorescent molecule tomography reconstruction method which comprises the following steps: acquiring a fluorescence picture and a white light picture of a to-be-imaged object by a full-angle free space FMT imaging system; extracting a borderline outline of the to-be-imaged object in each white light picture through an edge detection method, so as to obtain a projection outline picture; conducting back projection on protection outline pictures through a filter back reflection method in sequence to obtain a three-dimensional outline picture of the to-be-imaged object; solving the green function of a diffusion equation through a finite element method; dividing each fluorescence picture into a plurality of equal sub fluorescence picture; building a sub system equation corresponding to each sub fluorescence picture with the solved green function; compressing the corresponding sub system equation of each sub fluorescence picture divided from each fluorescence picture, then sequentially arranging the sub system equations to obtain a compression system equation; and solving the compression system equation to obtain the distribution of fluorescence markers in the to-be-imaged object. The large-data-size fluorescence molecule tomography reconstruction method is widely applied to the large-data-size fluorescent molecule tomography reconstruction.

Description

technical field [0001] The invention relates to a reconstruction method of fluorescence molecular tomography, in particular to a reconstruction method of fluorescence molecular tomography with large data volume based on data dimensionality reduction technology. Background technique [0002] Fluorescence Molecular Tomography (FMT) is an emerging optical tomography technology, which is a high-sensitivity, ionizing-radiation-free, and low-cost in vivo small animal imaging method. It has broad application prospects in diagnosis and other fields. Fluorescence molecular tomography technology uses fluorescent markers to mark specific molecules or cells in small animals, and when the labeled small animals are irradiated with excitation light of appropriate wavelength and intensity, the fluorescent markers in the small animals emit fluorescence. By detecting this fluorescent signal and using corresponding reconstruction methods, the distribution of fluorescent markers in small anima...

AI Technical Summary

Problems solved by technology

With the development of fluorescence molecular tomography system, charge-coupled device (CCD) is used to collect projection data, so that each projection can collect a large amount of projection data (for example, a 512*512 pixel CCD can collect 262144 projection data ), in order to obtain more effective information conveniently, the full-angle free-space imaging mode is widely used in fluorescence molecular tomography imaging systems. The collected projection data increases dramatically, and a large amount of projection data (approximately 10 7 ) for fluorescence molecular tomography reconstruction will effectively improve the resolution and stability of the reconstructed image, but this kind of large data volume fluorescence molecular tomography reconstruction consumes a lot of computer memory (about 10 2 GB) and calculation time, such a huge memory consumption far exceeds the memory capacity of ordinary PCs (personal computers) at present, and it is difficult to realize it on ordinary PCs. Even with the development of personal computers, insufficient memory is no longer a bottleneck, huge The calculation time will seriously restrict the rapid and real-time fluorescence molecular tomography process, so the reconstruction of large data volume fluorescence molecular tomography is almost impossible to complete with the existing methods

Images