List-mode wavelet-based multiresolution image reconstruction for Compton imaging

Iterative image reconstruction of data measured by a Compton scattering camera has to overcome various difficulties, e.g. a large amount of data, noise arising from both low counts recorded and imaging response, etc. Several approaches were proposed trying to attain an efficient Compton scattered data reconstruction. A forward step is introduction of list-mode data acquisition, which represents an efficient way of measurements allowing to preserve full precision of events as well as to gain computation efficiency. Iterative approaches include e.g. ML based algorithms, Bayesian methods, etc. Image estimation by ML criterion brings noise amplification when an iterative algorithm is applied, so a regularization step is needed. The proposed solution is a regularization technique using wavelet-based thresholding of the OSEM correction factors, called WOSEM. Methods of wavelet thresholding are efficient for edge-preserving image de-noising. The threshold used depends on the wavelet decomposition level. The method is illustrated by computer experiments.