Toward physiologically motivated registration of diagnostic CT and PET/CT of lung volumes.

PURPOSE: Current clinical practice for lung cancer diagnosis and staging requires the acquisition of a diagnostic computed tomography (CT) as well as positron emission tomography (PET)/CT volumes from a hybrid scanner, where the CT is used for attenuation correction (AC-CT). The PET and AC-CT images are implicitly aligned, however, image registration between the diagnostic CT and PET volumes is needed to relate the anatomical correspondences. This is an important but difficult task due to the absence of a direct or functional relationship between the intensities. Alternatively, here we propose the diagnostic CT can be aligned with the PET image through an indirect registration process that uses the AC-CT. The resultant deformation field can then be used to align the PET image to the diagnostic CT. The registration of the diagnostic CT to AC-CT registration still presents two major challenges: (a) it is a multimodal registration problem since the diagnostic CT is acquired after the injection of a contrast agent, and (b) the type and amplitude of the deformations require a registration process that includes physically motivated properties to achieve an accurate and physiologically plausible alignment. METHODS: The authors propose a new framework based on fluid registration including three physiologically motivated properties: (i) sliding motion of the lungs against the pleura; (ii) preservation of rigid structures; (iii) preservation of topology. The sliding motion is modeled using direction dependent regularization that decouples the tangential and the normal components of the external force term. The rigid shape of the bones is preserved using a spatially varying filter for the deformations. Finally, the topology is maintained using the concept of log-unbiased deformations. To solve the multimodal problem, the authors use local cross correlation (LCC) as the similarity measure. RESULTS: The proposed framework is first evaluated on CT lung image pairs representing several phases of the respiratory cycle. The authors show that their proposed framework has a superior performance compared to the classic fluid registration, both in quantitative and qualitative terms. The authors then evaluate the framework using ten real patient scans, where the authors also demonstrate how their physiologically motivated registration framework can be successfully applied to the task of fusing diagnostic CT with the PET/CT image volumes. CONCLUSIONS: The proposed registration framework has better results for the fusion of diagnostic CT with PET images in comparison to the classic fluid registration framework.

Comparison and evaluation of methods for liver segmentation from CT datasets.

This paper presents a comparison study between 10 automatic and six interactive methods for liver segmentation from contrast-enhanced CT images. It is based on results from the "MICCAI 2007 Grand Challenge" workshop, where 16 teams evaluated their algorithms on a common database. A collection of 20 clinical images with reference segmentations was provided to train and tune algorithms in advance. Participants were also allowed to use additional proprietary training data for that purpose. All teams then had to apply their methods to 10 test datasets and submit the obtained results. Employed algorithms include statistical shape models, atlas registration, level-sets, graph-cuts and rule-based systems. All results were compared to reference segmentations five error measures that highlight different aspects of segmentation accuracy. All measures were combined according to a specific scoring system relating the obtained values to human expert variability. In general, interactive methods reached higher average scores than automatic approaches and featured a better consistency of segmentation quality. However, the best automatic methods (mainly based on statistical shape models with some additional free deformation) could compete well on the majority of test images. The study provides an insight in performance of different segmentation approaches under real-world conditions and highlights achievements and limitations of current image analysis techniques.