Reconstruction of Under-sampled Dynamic Images by Modeling the Motion of Object Elements
Mir R, Prieto C, Irarrazaval P, Guarini M, Batchelor P, G. Hill D
Pontificia Universidad Catolica de Chile
We propose a new method to reconstruct under-sampled in vivo dynamic images. This technique recovers the lost data from the motion estimation of each object element (obel, think of a piece of tissue) in a continuous space. The results show that the method can effectively reconstruct sequences of 2D cardiac images with an under-sampling factor of 4. Furthermore, the method obtains the vectors of movement of each obel and does not require the motion to be confined to a part of the field of view or to a portion of the temporal frequency.