The 3D CT reconstruction algorithm to directly reconstruct multi-characteristic based on EMD

In the traditional three-dimensional CT reconstruction algorithms, high-frequency characteristics are difficult to obtain, and the noise is difficult to filter. In addition to, the reconstruction purpose is of diversification, and the three-dimensional filter to extract feature is more difficult to design and implement. Based on the characteristics of three-dimensional reconstruction algorithm and the RADON transform, a 3D CT reconstruction method to directly reconstruct multi-characteristic based on EMD was proposed and derived from the theoretical possibility and feasibility. EMD algorithm could firstly be used to decompose projection image before the reconstruction, and we could combine different IMF components to achieve a variety of feature, then use three-dimensional reconstruction algorithm to directly reconstruct the characteristics. The experimental results showed that the algorithm can better highlight a variety of useful information, simplify the filter design, and is of a certain practicability.

► Deduce the three-dimension 3D CT reconstruction algorithm to directly reconstruct multi-characteristic.
► Import EMD to the new algorithm and prove the feasibility.
► Obtain the required various features by combining different IMF components and meet to multi-characteristic reconstruction.
► Provide convinced experiments.