Collaborative Computational Project in Synergistic Reconstruction for Biomedical Imaging

Synergistic Symposium 2019 Multimodal

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Meike Kinzel Joint reconstruction via ICBTV-regularization The combination of functional imaging and soft tissue morphological imaging, called Positron Emission Tomography-Magnetic Resonance Imaging (PET-MRI), is one of the very promising new tools in oncology, cardiology and neurology. Because of its incorporation of exquisite structural tissue characterization and extreme sensitivity of metabolism, a lot of medical researchers are enganged in understanding its diagnostic benefits. Using the example of this hybrid imaging technique in this talk we want to present an idea to improve individual reconstructions via the exploitation of joint features. Based on the assumption that the final images of both, the PET as well as the MRI technique, share a similar edge set and hence the sparsity pattern with respect to the gradient, we propose to use the Infimal Convolution of Bregman Distances of the TV functional (ICBTV) to enhance either the quality of the PET reconstruction or the speed of the MRI acquisition process.
Leon Bungert Blind image fusion using directional total variation We propose a simultaneous super-resolution and blind deconvolution approach that allows to fuse a low resolution functional image (e.g. hyperspectral, PET, etc.) with a highly resolved image. Its structural information is incorporated via the directional total variation regulariser. The resulting non-convex variational problem is solved with alternating minimization. Numerical results on hyperspectral and panchromatic images show that our model can sucessively fuse two different image modalities even when the structural side information is mis-aligned.
Jinyi Qi Multimodal multiscale PET image reconstruction
C-A. Collins-Fekete Synergistic proton/X-ray imaging and applications in medical physics In the last decade, proton computed tomography (pCT) has established itself as a standard to measure the relative stopping power for treatment planning in proton radiotherapy. The unique properties of proton imaging are related to its measurements of the proton scattering, energy loss and fluence loss through a medium, enabling the tomographic reconstruction of diverse and useful metric for dosimetry prediction. However, protons suffer a large amount of charged scattering as they cross the imaged object, effectively introducing a deviation in their path which results in a blurring of the final image. This effect limits the spatial resolution of pCT. Thus, synergistic reconstruction techniques have been the focus of recent research to try and combine X-rays CT high spatial resolution with pCT high contrast and accuracy. This talk will provide an overview of recent synergistic reconstruction techniques, including work on combining proton radiography and X-ray CT/CBCT for RSP prediction, synergistic tomographic reconstruction using a mix of pCT and CBCT projections, and a tomographic derivation of the medium mean excitation energy through the use of Dual-Energy X-ray CT and pCT reconstruction.