Alexandre Bousse

An anatomically driven anisotropic diffusion filtering method for 3D SPECT reconstruction

In this study, we aim to reconstruct single-photon emission computed tomography images using anatomical information from magnetic resonance imaging as a priori knowledge about the activity distribution. The trade-off between anatomical and emission data is one of the main concerns for such studies. In this work, we propose an anatomically driven anisotropic diffusion filter (ADADF) as a penalized maximum likelihood expectation maximization optimization framework. The ADADF method has improved edge-preserving denoising characteristics compared to other smoothing penalty terms based on quadratic and non-quadratic functions. The proposed method has an important ability to retain information which is absent in the anatomy. To make our approach more stable to the noise-edge classification problem, robust statistics have been employed. Comparison of the ADADF method is performed with a successful anatomically driven technique, namely, the Bowsher prior (BP). Quantitative assessment using simulated and clinical neuroreceptor volumetric data show the advantage of the ADADF over the BP. For the modelled data, the overall image resolution, the contrast, the signal-to-noise ratio and the ability to preserve important features in the data are all improved by using the proposed method. For clinical data, the contrast in the region of interest is significantly improved using the ADADF compared to the BP, while successfully eliminating noise.

GPU accelerated rotation-based emission tomography reconstruction

Stochastic methods based on Maximum Likelihood Estimation (MLE) provide accurate tomographic reconstruction for emission imaging. Moreover methods based on MLE allow to include an accurate physical model of the imaging setup in the reconstruction process, thus enabling quantitative reconstruction of radio-tracer activity distribution. It has been shown that inclusion of a spatially dependent PSF that models dependence of the CDR with distance from the detector, improves the quality of reconstruction in terms of noise and bias. The computational complexity associated with stochastic methods has limited adoption of such algorithms for clinical use and inclusion of the PSF further increases the computational cost. This work proposes an accelerated implementation of a reconstruction algorithm specifically designed to take advantage of the architecture of a General Purpose Graphics Processing Unit (GPGPU).

Markov random field and Gaussian mixture for segmented MRI-based partial volume correction in PET.

In this paper we propose a segmented magnetic resonance imaging (MRI) prior-based maximum penalized likelihood deconvolution technique for positron emission tomography (PET) images. The model assumes the existence of activity classes that behave like a hidden Markov random field (MRF) driven by the segmented MRI. We utilize a mean field approximation to compute the likelihood of the MRF. We tested our method on both simulated and clinical data (brain PET) and compared our results with PET images corrected with the re-blurred Van Cittert (VC) algorithm, the simplified Guven (SG) algorithm and the region-based voxel-wise (RBV) technique. We demonstrated our algorithm outperforms the VC algorithm and outperforms SG and RBV corrections when the segmented MRI is inconsistent (e.g. mis-segmentation, lesions, etc) with the PET image.

A Bayesian MAP-EM Algorithm for PET Image Reconstruction Using Wavelet Transform

In this paper, we present a PET reconstruction method using the wavelet-based maximum a posteriori (MAP) expectation-maximization (EM) algorithm. The proposed method, namely WV-MAP-EM, shows several advantages over conventional methods. It provides an adaptive way for hyperparameter determination. Since the wavelet transform allows the use of fast algorithms, WV-MAP-EM also does not increase the order of computational complexity. The spatial noise behavior (bias/variance and resolution) of the proposed MAP estimator is analyzed. Quantitative comparisons to MAP methods with Markov random field (MRF) prior models point out that our alternative method, wavelet-base method, offers competitive performance in PET image reconstruction.

Edge preserving bowsher prior with nonlocal weighting for 3D spect reconstruction

In this study, we aim at reconstructing single photon emission computed tomography (SPECT) images using a Bayesian framework to incorporate anatomical information from magnetic resonance (MR) as a priori knowledge about the activity distribution. This is achieved using an anatomically-driven Bowsher prior (BP). Standard BP has the potential to obtain similar results as other state-of-the-art prior embedding techniques, although without expensive calculation demands. However, due to the suboptimal selection of similar neighbours on anatomy, BP is not able to preserve edges. In this paper, we propose a modification of the BP, which can maintain discontinuities while still allowing locally smoothing effect on the same tissue type. Furthermore, a non-local algorithm for the weights estimation is included into the modified BP in order to improve the robustness of the method to non-correlated information between activity and anatomy. Quantitative assessment of our new algorithm is performed using co-registered simulated 3D SPECT/MR data, and compared with a regular BP technique. Preliminary results demonstrate considerable improvement of the edge preserving characteristics and hot lesion detection on the overall quality of the reconstructed images.

Maximum-Likelihood Joint Image Reconstruction/Motion Estimation in Attenuation-Corrected Respiratory Gated PET/CT Using a Single Attenuation Map

This work provides an insight into positron emission tomography (PET) joint image reconstruction/motion estimation (JRM) by maximization of the likelihood, where the probabilistic model accounts for warped attenuation. Our analysis shows that maximum-likelihood (ML) JRM returns the same reconstructed gates for any attenuation map ( μ-map) that is a deformation of a given μ-map, regardless of its alignment with the PET gates. We derived a joint optimization algorithm accordingly, and applied it to simulated and patient gated PET data. We first evaluated the proposed algorithm on simulations of respiratory gated PET/CT data based on the XCAT phantom. Our results show that independently of which μ-map is used as input to JRM: (i) the warped μ-maps correspond to the gated μ-maps, (ii) JRM outperforms the traditional post-registration reconstruction and consolidation (PRRC) for hot lesion quantification and (iii) reconstructed gated PET images are similar to those obtained with gated μ-maps. This suggests that a breath-held μ-map can be used. We then applied JRM on patient data with a μ-map derived from a breath-held high resolution CT (HRCT), and compared the results with PRRC, where each reconstructed PET image was obtained with a corresponding cine-CT gated μ-map. Results show that JRM with breath-held HRCT achieves similar reconstruction to that using PRRC with cine-CT. This suggests a practical low-dose solution for implementation of motion-corrected respiratory gated PET/CT.

Weighted MRI-Based bowsher priors for SPECT brain image reconstruction

This paper presents a generalization of the Bowsher prior for SPECT reconstruction using anatomical prior. Instead of considering a binary selection of the neighbors of each reconstructed voxel based on the anatomical prior values, each neighbors are taken into account with a suitable weight. We tried three different weights. We showed that, for brain SPECT reconstruction using MRI, in the case of perfect registration, the weighted Bowsher prior give similar results to the regular Bowsher prior. In the case of misregistration, the weighted Bowsher prior outperforms the Bowsher prior.

PETPVC: a toolbox for performing partial volume correction techniques in positron emission tomography

Positron emission tomography (PET) images are degraded by a phenomenon known as the partial volume effect (PVE). Approaches have been developed to reduce PVEs, typically through the utilisation of structural information provided by other imaging modalities such as MRI or CT. These methods, known as partial volume correction (PVC) techniques, reduce PVEs by compensating for the effects of the scanner resolution, thereby improving the quantitative accuracy. The PETPVC toolbox described in this paper comprises a suite of methods, both classic and more recent approaches, for the purposes of applying PVC to PET data. Eight core PVC techniques are available. These core methods can be combined to create a total of 22 different PVC techniques. Simulated brain PET data are used to demonstrate the utility of toolbox in idealised conditions, the effects of applying PVC with mismatched point-spread function (PSF) estimates and the potential of novel hybrid PVC methods to improve the quantification of lesions. All anatomy-based PVC techniques achieve complete recovery of the PET signal in cortical grey matter (GM) when performed in idealised conditions. Applying deconvolution-based approaches results in incomplete recovery due to premature termination of the iterative process. PVC techniques are sensitive to PSF mismatch, causing a bias of up to 16.7% in GM recovery when over-estimating the PSF by 3 mm. The recovery of both GM and a simulated lesion was improved by combining two PVC techniques together. The PETPVC toolbox has been written in C++, supports Windows, Mac and Linux operating systems, is open-source and publicly available.

Maximum-likelihood joint image reconstruction and motion estimation with misaligned attenuation in TOF-PET/CT

This work is an extension of our recent work on joint activity reconstruction/motion estimation (JRM) from positron emission tomography (PET) data. We performed JRM by maximization of the penalized log-likelihood in which the probabilistic model assumes that the same motion field affects both the activity distribution and the attenuation map. Our previous results showed that JRM can successfully reconstruct the activity distribution when the attenuation map is misaligned with the PET data, but converges slowly due to the significant cross-talk in the likelihood. In this paper, we utilize time-of-flight PET for JRM and demonstrate that the convergence speed is significantly improved compared to JRM with conventional PET data.

Joint parametric reconstruction and motion correction framework for dynamic PET data.

In this paper we propose a novel algorithm for jointly performing data based motion correction and direct parametric reconstruction of dynamic PET data. We derive a closed form update for the penalised likelihood maximisation which greatly enhances the algorithms computational efficiency for practical use. Our algorithm achieves sub-voxel motion correction residual with noisy data in the simulation-based validation and reduces the bias of the direct estimation of the kinetic parameter of interest. A preliminary evaluation on clinical brain data using [18F]Choline shows improved contrast for regions of high activity. The proposed method is based on a data-driven kinetic modelling method and is directly applicable to reversible and irreversible PET tracers, covering a range of clinical applications.