Parametric spatial transformation models have been successfully applied to image registration tasks. In such models, the transformation of interest is parameterized by a fixed set of basis functions as for example B-splines. Each basis function is located on a fixed regular grid position among the image domain because the transformation of interest is not known in advance. As a consequence, not all basis functions will necessarily contribute to the final transformation which results in a non-compact representation of the transformation. We reformulate the pairwise registration problem as a recursive sequence of successive alignments. For each element in the sequence, a local deformation defined by its position, shape, and weight is computed by our recurrent registration neural network. The sum of all lo- cal deformations yield the final spatial alignment of both images. Formulating the registration problem in this way allows the network to detect non-aligned regions in the images and to learn how to locally refine the registration properly. In contrast to current non-sequence-based registration methods, our approach iteratively applies local spatial deformations to the images until the desired registration accuracy is achieved. We trained our network on 2D magnetic resonance images of the lung and compared our method to a standard parametric B-spline registration. The experiments show, that our method performs on par for the accuracy but yields a more compact representation of the transformation. Furthermore, we achieve a speedup of around 15 compared to the B-spline registration.

X-ray grating interferometry is a powerful emerging tool in biomedical imaging, providing access to three complementary image modalities. In addition to the conventional attenuation modality, interferometry provides a phase modality, which visualizes soft tissue structures, and a dark-field modality, which relates to the number and size of sub-resolution scattering objects. A particularly strong dark-field signal originates from the alveoli or air sacs in the lung. Dark-field lung radiographs in animal models have already shown increased sensitivity in diagnosing lung diseases, such as lung cancer or emphysema, compared to conventional X-ray chest radiography. However, to date, X-ray dark-field lung imaging has either averaged information over several breaths or has been captured during a breath hold. In this paper, we demonstrate the first time-resolved dark-field imaging of a breath cycle in a mechanically ventilated mouse, in vivo, which was obtained using a grating interferometer. We achieved a time resolution of 0.1 s, visualizing the changes in the dark-field, phase, and attenuation images during inhalation and exhalation. These measurements show that the dark-field signal depends on the air volume and, hence, the alveolar dimensions of the lung. Conducting this type of scan with animal disease models would help to locate the optimum breath point for single-image diagnostic dark-field imaging and could indicate if the changes in the dark-field signal during breath provide a diagnostically useful complementary measure.

Models of shape variations have become a central component for the automated analysis of images. An important class of shape models are point distribution models (PDMs). These models represent a class of shapes as a normal distribution of point variations, whose parameters are estimated from example shapes. Principal component analysis (PCA) is applied to obtain a low-dimensional representation of the shape variation in terms of the leading principal components. In this paper, we propose a generalization of PDMs, which we refer to as Gaussian Process Morphable Models (GPMMs). We model the shape variations with a Gaussian process, which we represent using the leading components of its Karhunen-Loève expansion. To compute the expansion, we make use of an approximation scheme based on the Nyström method. The resulting model can be seen as a continuous analog of a standard PDM. However, while for PDMs the shape variation is restricted to the linear span of the example data, with GPMMs we can define the shape variation using any Gaussian process. For example, we can build shape models that correspond to classical spline models and thus do not require any example data. Furthermore, Gaussian processes make it possible to combine different models. For example, a PDM can be extended with a spline model, to obtain a model that incorporates learned shape characteristics but is flexible enough to explain shapes that cannot be represented by the PDM. We introduce a simple algorithm for fitting a GPMM to a surface or image. This results in a non-rigid registration approach whose regularization properties are defined by a GPMM. We show how we can obtain different registration schemes, including methods for multi-scale or hybrid registration, by constructing an appropriate GPMM. As our approach strictly separates modeling from the fitting process, this is all achieved without changes to the fitting algorithm. To demonstrate the applicability and versatility of GPMMs, we perform a set of experiments in typical usage scenarios in medical image analysis and computer vision: The model-based segmentation of 3D forearm images and the building of a statistical model of the face. To complement the paper, we have made all our methods available as open source.

We present a novel image segmentation method based on statistical shape model fitting. Instead of fitting the model to raw intensity values we consider object probabilities. The abstraction from the plain intensity images to probability maps makes the segmentation more robust against misleading texture inside the object or surrounding background. The target object probability is predicted based on random forest regression trained with neighborhood dependent features of sample images. In contrast to similar approaches, both, the object boundary as well as the whole object and background region are considered for segmentation. We apply our approach to a 3D cone beam computed tomography image dataset of the jaw region where we segment the wisdom tooth shape. Compared to a boundary-and a region-based method we obtain superior segmentation performance.