Dirk Loeckx

Evaluation of Registration Methods on Thoracic CT: The EMPIRE10 Challenge

EMPIRE10 (Evaluation of Methods for Pulmonary Image REgistration 2010) is a public platform for fair and meaningful comparison of registration algorithms which are applied to a database of intra patient thoracic CT image pairs. Evaluation of nonrigid registration techniques is a nontrivial task. This is compounded by the fact that researchers typically test only on their own data, which varies widely. For this reason, reliable assessment and comparison of different registration algorithms has been virtually impossible in the past. In this work we present the results of the launch phase of EMPIRE10, which comprised the comprehensive evaluation and comparison of 20 individual algorithms from leading academic and industrial research groups. All algorithms are applied to the same set of 30 thoracic CT pairs. Algorithm settings and parameters are chosen by researchers expert in the con figuration of their own method and the evaluation is independent, using the same criteria for all participants. All results are published on the EMPIRE10 website (http://empire10.isi.uu.nl). The challenge remains ongoing and open to new participants. Full results from 24 algorithms have been published at the time of writing. This paper details the organization of the challenge, the data and evaluation methods and the outcome of the initial launch with 20 algorithms. The gain in knowledge and future work are discussed.

Nonrigid Image Registration Using Conditional Mutual Information

Maximization of mutual information (MMI) is a popular similarity measure for medical image registration. Although its accuracy and robustness has been demonstrated for rigid body image registration, extending MMI to nonrigid image registration is not trivial and an active field of research. We propose conditional mutual information (cMI) as a new similarity measure for nonrigid image registration. cMI starts from a 3-D joint histogram incorporating, besides the intensity dimensions, also a spatial dimension expressing the location of the joint intensity pair. cMI is calculated as the expected value of the cMI between the image intensities given the spatial distribution. The cMI measure was incorporated in a tensor-product B-spline nonrigid registration method, using either a Parzen window or generalized partial volume kernel for histogram construction. cMI was compared to the classical global mutual information (gMI) approach in theoretical, phantom, and clinical settings. We show that cMI significantly outperforms gMI for all applications.

Three-Dimensional Cardiac Strain Estimation Using Spatio–Temporal Elastic Registration of Ultrasound Images: A Feasibility Study

Current ultrasound methods for measuring myocardial strain are often limited to measurements in one or two dimensions. Cardiac motion and deformation however are truly 3-D. With the introduction of matrix transducer technology, 3-D ultrasound imaging of the heart has become feasible but suffers from low temporal and spatial resolution, making 3-D strain estimation challenging. In this paper, it is shown that automatic intensity-based spatio-temporal elastic registration of currently available 3-D volumetric ultrasound data sets can be used to measure the full 3-D strain tensor. The method was validated using simulated 3-D ultrasound data sets of the left ventricle (LV). Three types of data sets were simulated: a normal and symmetric LV with different heart rates, a more realistic asymmetric normal LV and an infarcted LV. The absolute error in the estimated displacement was between 0.47 plusmn0.23 and 1.00 plusmn0.59 mm, depending on heart rate and amount of background noise. The absolute error on the estimated strain was 9%-21% for the radial strain and 1%-4% for the longitudinal and circumferential strains. No large differences were found between the different types of data sets. The shape of the strain curves was estimated properly and the position of the infarcts could be identified correctly. Preliminary results on clinical data taken in vivo from three healthy volunteers and one patient with an apical aneurism confirmed these findings in a qualitative manner as the strain curves obtained with the proposed method have an amplitude and shape similar to what could be expected.

Relationship between fluorodeoxyglucose uptake in the large vessels and late aortic diameter in giant cell arteritis

OBJECTIVE GCA carries an increased risk of developing thoracic aortic aneurysms. Previous work with fluorodeoxyglucose (FDG)-PET has shown that the aorta is frequently involved in this type of vasculitis. We wanted to investigate whether there is a correlation between the extent of vascular FDG uptake during the acute phase of GCA and the aortic diameter at late follow-up. METHODS All patients with biopsy-proven GCA who ever underwent an FDG-PET scan in our centre were asked to undergo a CT scan of the aorta. The diameter of the aorta was measured at six different levels (ascending aorta, aortic arch, descending aorta, abdominal suprarenal, juxtarenal and infrarenal aorta) and the volumes of the thoracic and of the abdominal aorta were calculated. RESULTS Forty-six patients agreed to participate (32 females, 14 males). A mean of 46.7 +/- 29.9 months elapsed between diagnosis and CT scan. All aortic dimensions were significantly smaller in women than in men, except for the diameter of the ascending aorta. Patients who had an increased FDG uptake in the aorta at diagnosis of GCA, had a significantly larger diameter of the ascending aorta (P = 0.025) and descending aorta (P = 0.044) and a significantly larger volume of the thoracic aorta (P = 0.029). In multivariate analysis, FDG uptake at the thoracic aorta was associated with late volume of the thoracic aorta (P = 0.039). CONCLUSION GCA-patients with increased FDG uptake in the aorta may be more prone to develop thoracic aortic dilatation than GCA patients without this sign of aortic involvement.

Minimal Shape and Intensity Cost Path Segmentation

A new generic model-based segmentation algorithm is presented, which can be trained from examples akin to the active shape model (ASM) approach in order to acquire knowledge about the shape to be segmented and about the gray-level appearance of the object in the image. Whereas ASM alternates between shape and intensity information during search, the proposed approach optimizes for shape and intensity characteristics simultaneously. Local gray-level appearance information at the landmark points extracted from feature images is used to automatically detect a number of plausible candidate locations for each landmark. The shape information is described by multiple landmark-specific statistical models that capture local dependencies between adjacent landmarks on the shape. The shape and intensity models are combined in a single cost function that is optimized noniteratively using dynamic programming, without the need for initialization. The algorithm was validated for segmentation of anatomical structures in chest and hand radiographs. In each experiment, the presented method had a significant higher performance when compared to the ASM schemes. As the method is highly effective, optimally suited for pathological cases and easy to implement, it is highly useful for many medical image segmentation tasks.

Nonrigid Image Registration Using Free-Form Deformations with a Local Rigidity Constraint

Voxel-based nonrigid image registration can be formulated as an optimisation problem whose goal is to minimise a cost function, consisting of a first term that characterises the similarity between both images and a second term that regularises the transformation and/or penalties improbable or impossible deformations. Within this paper, we extend previous works on nonrigid image registration by the introduction of a new penalty term that expresses the local rigidity of the deformation. A necessary and sufficient condition for the transformation to be locally rigid at a particular location is that its Jacobian matrix J T at this location is orthogonal, satisfying the orthogonality condition \(J_{\rm T} J_{\rm T}^{T} = {\bf 1}\). So we define the penalty term as the weighted integral of the Frobenius norm of \(J_{\rm T} J_{\rm T}^{T} - {\bf 1}\) integrated over the overlap of the images to be registered. We fit the implementation of the penalty term in a multidimensional, continuous and differentiable B-spline deformation framework and analytically determine the derivative of the similarity criterion and the penalty term with respect to the deformation parameters. We show results of the impact of the proposed rigidity constraint on artificial and clinical images demonstrating local shape preservation with the proposed constraint.

Biological image-guided radiotherapy in rectal cancer: Is there a role for FMISO or FLT, next to FDG?

Purpose. The purpose of this study is to investigate the use of PET/CT with fluorodeoxyglucose (FDG), fluorothymidine (FLT) and fluoromisonidazole (FMISO) for radiotherapy (RT) target definition and evolution in rectal cancer. Materials and methods. PET/CT was performed before and during preoperative chemoradiotherapy (CRT) in 15 patients with resectable rectal cancer. PET signals were delineated and CT images on the different time points were non-rigidly registered. Mismatch analyses were carried out to quantify the overlap between FDG and FLT or FMISO tumour volumes (TV) and between PET TVs over time. Results. Ninety sequential PET/CT images were analyzed. The mean FDG, FLT and FMISO-PET TVs showed a tendency to shrink during preoperative CRT. On each time point, the mean FDG-PET TV was significantly larger than the FMISO-PET TV but not significantly larger than the mean FLT-PET TV. There was a mean 65% mismatch between the FMISO and FDG TVs obtained before and during CRT. FLT TVs corresponded better with the FDG TVs (25% mismatch before and 56% during CRT). During CRT, on average 61% of the mean FDG TV (7 cc) overlapped with the baseline mean TV (15.5 cc) (n=15). For FLT, the TV overlap was 49% (n=5) and for FMISO only 20% of the TV during CRT remained inside the contour at baseline (n=10). Conclusion. FDG, FLT and FMISO-PET reflect different functional characteristics that change during CRT in rectal cancer. FLT and FDG show good spatial correspondence, while FMISO seems less reliable due to the non-specific FMISO uptake in normoxic tissue and tracer diffusion through the bowel wall. FDG and FLT-PET/CT imaging seem most appropriate to integrate in preoperative RT for rectal cancer.

Temporal subtraction of thorax CR images using a statistical deformation model

We propose a voxel-based nonrigid registration algorithm for temporal subtraction of two-dimensional thorax X-ray computed radiography images of the same subject. The deformation field is represented by a B-spline with a limited number of degrees of freedom, that allows global rib alignment to minimize subtraction artifacts within the lung field without obliterating interval changes of clinically relevant soft-tissue abnormalities. The spline parameters are constrained by a statistical deformation model that is learned from a training set of manually aligned image pairs using principal component analysis. Optimization proceeds along the transformation components rather then along the individual spline coefficients, using pattern intensity of the subtraction image within the automatically segmented lung field region as the criterion to be minimized and applying a simulated annealing strategy for global optimization in the presence of multiple local optima. The impact of different transformation models with varying number of deformation modes is evaluated on a training set of 26 images using a leave-one-out strategy and compared to the manual registration result in terms of criterion value and deformation error. Registration quality is assessed on a second set of validation images by a human expert rating each subtraction image on screen. In 85% of the cases, the registration is subjectively rated to be adequate for clinical use.

Changes in left atrial anatomy due to respiration: impact on three-dimensional image integration during atrial fibrillation ablation.

Introduction: Patient respiration influences the accuracy of image integration approaches used during atrial fibrillation (AF) ablation procedures. We assessed both absolute and relative changes in left atrial (LA) and pulmonary venous (PV) anatomy due to respiration and their implications for 3D image integration.

Biological image-guided radiotherapy in rectal cancer: challenges and pitfalls.

PURPOSE To investigate the feasibility of integrating multiple imaging modalities for image-guided radiotherapy in rectal cancer. PATIENTS AND METHODS Magnetic resonance imaging (MRI) and fluorodeoxyglucose positron emission tomography/computed tomography (FDG-PET/CT) were performed before, during, and after preoperative chemoradiotherapy (CRT) in patients with resectable rectal cancer. The FDG-PET signals were segmented with an adaptive threshold-based and a gradient-based method. Magnetic resonance tumor volumes (TVs) were manually delineated. A nonrigid registration algorithm was applied to register the images, and mismatch analyses were carried out between MR and FDG-PET TVs and between TVs over time. Tumor volumes delineated on the images after CRT were compared with the pathologic TV. RESULTS Forty-five FDG-PET/CT and 45 MR images were analyzed from 15 patients. The mean MRI and FDG-PET TVs showed a tendency to shrink during and after CRT. In general, MRI showed larger TVs than FDG-PET. There was an approximately 50% mismatch between the FDG-PET TV and the MRI TV at baseline and during CRT. Sixty-one percent of the FDG-PET TV and 76% of the MRI TV obtained after 10 fractions of CRT remained inside the corresponding baseline TV. On MRI, residual tumor was still suspected in all 6 patients with a pathologic complete response, whereas FDG-PET showed a metabolic complete response in 3 of them. The FDG-PET TVs delineated with the gradient-based method matched closest with pathologic findings. CONCLUSIONS Integration of MRI and FDG-PET into radiotherapy seems feasible. Gradient-based segmentation is recommended for FDG-PET. Spatial variance between MRI and FDG-PET TVs should be taken into account for target definition.