This paper presents a novel method for atlas-based segmentation of medical images. The method uses semi-supervised learning of a graph describing a manifold of anatomical variations of whole-body images, where unlabelled data are used to find a path with small deformations from the labelled atlas to the target image. The method is evaluated on 36 whole-body magnetic resonance images with manually segmented livers as ground truth. Significant improvement (p <; 0.001) was obtained compared to direct atlas-based registration.

Level set methods are a popular way to solve the image segmentation problem. The solution contour is found by solving an optimization problem where a cost functional is minimized. Gradient descent methods are often used to solve this optimization problem since they are very easy to implement and applicable to general nonconvex functionals. They are, however, sensitive to local minima and often display slow convergence. Traditionally, cost functionals have been modified to avoid these problems. In this paper, we instead propose using two modified gradient descent methods, one using a momentum term and one based on resilient propagation. These methods are commonly used in the machine learning community. In a series of 2-D/3-D-experiments using real and synthetic data with ground truth, the modifications are shown to reduce the sensitivity for local optima and to increase the convergence rate. The parameter sensitivity is also investigated. The proposed methods are very simple modifications of the basic method, and are directly compatible with any type of level set implementation. Downloadable reference code with examples is available online.

Computed Tomography Angiography (CTA) is commonly used in clinical routine for diagnosing vascular diseases. The procedure involves the injection of a contrast agent into the blood stream to increase the contrast between the blood vessels and the surrounding tissue in the image data. CTA is often visualized with Direct Volume Rendering (DVR) where the enhanced image contrast is important for the construction of Transfer Functions (TFs). For increased efficiency, clinical routine heavily relies on preset TFs to simplify the creation of such visualizations for a physician. In practice, however, TF presets often do not yield optimal images due to variations in mixture concentration of contrast agent in the blood stream. In this paper we propose an automatic, optimization-based method that shifts TF presets to account for general deviations and local variations of the intensity of contrast enhanced blood vessels. Some of the advantages of this method are the following. It computationally automates large parts of a process that is currently performed manually. It performs the TF shift locally and can thus optimize larger portions of the image than is possible with manual interaction. The method is based on a well known vesselness descriptor in the definition of the optimization criterion. The performance of the method is illustrated by clinically relevant CT angiography datasets displaying both improved structural overviews of vessel trees and improved adaption to local variations of contrast concentration.

Many applications in image analysis include the problem of linear structure detection, e.g. segmentation of blood vessels in medical images, roads in satellite images, etc. A simple and efficient solution is to apply linear filters tuned to the structures of interest and extract line and edge positions from the filter output. However, if the filter is not carefully designed, artifacts such as ringing can distort the results and hinder a robust detection. In this paper, we study the ringing effects using a common Gabor filter for linear structure detection, and suggest a method for generating non-ring filters in 2D and 3D. The benefits of the non-ring design are motivated by results on both synthetic and natural images.

Correlation is often used to measure the similarity between signals and is an important tool in signal and image processing. In some applications it is common that signals are corrupted by local bursts of noise. This adversely affects the performance of signal recognition algorithms. This paper presents a novel correlation estimator, which is robust to locally corrupted signals. The estimator is generalized to multivariate correlation analysis (general linear model, GLM, and canonical correlation analysis, CCA). Synthetic functional MRI data is used to demonstrate the estimator, and its robustness is shown to increase the performance of signal detection.

The segmentation and analysis of blood vessels has received much attention in the research community. The results aid numerous applications for diagnosis and treatment of vascular diseases. Here we use level set propagation with local phase information to capture the boundaries of vessels. The basic notion is that local phase, extracted using quadrature filters, allows us to distinguish between lines and edges in an image. Noting that vessels appear either as lines or edge pairs, we integrate multiple scales and capture information about vessels of varying width. The outcome is a ¿global¿ phase which can be used to drive a contour robustly towards the vessel edges. We show promising results in 2D and 3D. Comparison with a related method gives similar or even better results and at a computational cost several orders of magnitude less. Even with very sparse initializations, our method captures a large portion of the vessel tree.

This paper introduces a new method for automatic quantification of subcutaneous, visceral and non-visceral internal fat from MR-images acquired using the two point Dixon technique in the abdominal region. The method includes (1) a three dimensional phase unwrapping to provide water and fat images, (2) an image intensity inhomogeneity correction, and (3) a morphon based registration and segmentation of the tissue. This is followed by an integration of the corrected fat images within the different fat compartments that avoids the partial volume effects associated with traditional fat segmentation methods. The method was tested on 18 subjects before and after a period of fast-food hyper-alimentation showing high stability and performance in all analysis steps.

In analysis of fMRI data, it is common to average neighboring voxels in order to obtain robust estimates of the correlations between voxel time-series and the model of the signal expected to be present in activated regions. We have previously proposed a method where only voxels with similar correlation coefficients are averaged. In this paper we extend this idea, and present a novel method for analysis of fMRI data. In the proposed method, only voxels with similar correlation coefficients and similar time-series are averaged. The proposed method is compared to our previous method and to two well-known filtering strategies, and is shown to have superior ability to discriminate between active and inactive voxels

Automatic segmentation of anatomical structures is often performed using model-based non-rigid registration methods. These algorithms work well when the images do not contain any large deviations from the normal anatomy. We have previously used such a method to generate patient specific models of hip bones for surgery simulation. The method that was used, the morphon method, registers two-or three-dimensional images using a multi-resolution deformation scheme. A prototype image is iteratively registered to a target image using quadrature filter phase difference to estimate the local displacement. The morphon method has in this work been extended to deal with automatic segmentation of fractured bones. Two features have been added. First, the method is modified such that multiple prototypes (in this case two) can be used. Second, normalised convolution is utilized for the displacement estimation, to guide the registration of the second prototype, based on the result of the registration of the first one.

It has previously been shown that canonical correlation analysis (CCA) works well for detecting neural activity in fMRI data. This is due to the ability of CCA to perform simultaneous temporal modeling and adaptive spatial filtering of the data. In this paper, we demonstrate that our previously proposed method for CCA-based fMRI data analysis does not provide rotationally invariant detection of activated regions. We propose a modification of the previous method and show that it resolves the rotational invariance issue, thereby further improving the analysis method.

We present a method that finds edges between certain image features, e.g. gray-levels, and disregards edges between other features. The method uses a channel representation of the features and performs normalized convolution using the channel values as certainties. This means that areas with certain features can be disregarded by the edge filter. The method provides an important tool for finding tissue specific edges in medical images, as demonstrated by an MR-image example.

Presents a general strategy for automated generation of efficient representations in vision. The approach is highly task oriented and what constitutes the relevant information is defined by a set of examples. The examples are pairs of situations that are dependent through the chosen feature but are otherwise independent. Particularly important concepts in the work are mutual information and canonical correlation. How visual operators and representation can be generated from examples are presented for a number of features, e.g. local orientation, disparity and motion. Interesting similarities to biological vision functions are observed. The results clearly demonstrates the potential of combining advanced filtering techniques and learning strategies based on canonical correlation analysis.

This paper presents our general strategy for designing learning machines as well as a number of particular designs. The search for methods allowing a sufficient level of adaptivity are based on two main principles: 1) simple adaptive local models; and 2) adaptive model distribution. Particularly important concepts in our work is mutual information and canonical correlation. Examples are given on learning feature descriptors, modeling disparity, synthesis of a global 3-mode model and a setup for reinforcement learning of online video coder parameter control.