Search

Bloodstain pattern analysis (BPA) is a subspecialty of forensic sciences, dealing with the analysis and interpretation of bloodstain patterns in crime scenes. The aim of BPA is uncovering new information about the actions that took place in a crime scene, potentially leading to a confirmation or refutation of a suspect's statement. A typical goal of BPA is to estimate the flight paths for a set of stains, followed by a directional analysis in order to estimate the area of origin for the stains. The traditional approach, referred to as stringing, consists of attaching a piece of string to each stain, and letting the string represent an approximation of the stain's flight path. Even though stringing has been used extensively, many (practical) downsides exist. We propose an automated and virtual approach, employing fiducial markers and digital images. By automatically reconstructing a single coordinate frame from several images, limited user input is required. Synthetic crime scenes were created and analysed in order to evaluate the approach. Results demonstrate the correct operation and practical advantages, suggesting that the proposed approach may become a valuable asset for practically analysing bloodstain spatter patterns. Accompanying software called HemoVision is currently provided as a demonstrator and will be further developed for practical use in forensic investigations., (Copyright © 2015 Elsevier Ireland Ltd. All rights reserved.)

In this paper, a specific method is presented to facilitate the semi-automatic segmentation of liver tumors and liver metastases in CT images. Accurate and reliable segmentation of tumors is essential for the follow-up of cancer treatment. The core of the algorithm is a level set method. The initialization is generated by a spiral-scanning technique based on dynamic programming. The level set evolves according to a speed image that is the result of a statistical pixel classification algorithm with supervised learning. This method is tested on CT images of the abdomen and compared with manual delineations of liver tumors. The described method outperformed the semi-automatic methods of the other participants of the "3D Liver Tumor Segmentation Challenge 2008". Evaluating the algorithm on the provided test data leads to an average overlap error of 32.6% and an average volume difference of 17.9%. The average, the RMS and the maximum surface distance are 2.0, 2.6 and 10.1 mm, respectively.

A generic supervised segmentation approach is presented. The object is described as a graph where the vertices correspond to landmarks points and the edges define the landmark relations. Instead of building one single global shape model, a priori shape information is represented as a concatenation of local shape models that consider only local dependencies between connected landmarks. The objective function is obtained from a maximum a posteriori criterion and is build up of localized energies of both shape and landmark intensity information. The optimization problem is discretized by searching candidates for each landmark using individual landmark intensity descriptors. The discrete optimization problem is then solved using mean field annealing or dynamic programming techniques. The algorithm is validated for hand bone segmentation from RX datasets and for 3D liver segmentation from contrast enhanced CT images.

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.

In this paper an algorithm for atlas-to-image non-rigid registration based on regional entropy minimization is presented. Tissue class probabilities in the atlas are registered with the intensities in the target image. The novel aspect of the paper consists in using tissue class probability maps that include the three main regions (for the brain, white matter, gray matter and csf) and a further partitioning thereof. For example, gray matter is further subdivided into basal ganglia (each of them defining its own class) and the rest (of gray matter). This guarantees a regional entropy minimization instead of just a global one. In other words, the local labels in the atlas will be adjusted in order to obtain the best explanation for the intensity distribution in the corresponding subregion of the target image.

We propose conditional mutual information (cMI) as a new similarity measure for nonrigid image registration. We start from a 3D joint histogram incorporating, besides the reference and floating intensity dimensions, also a spatial dimension expressing the location of the joint intensity pair in the reference image. cMI is calculated as the expectation value of the conditional mutual information between the reference and floating intensities given the spatial distribution. Validation experiments were performed comparing cMI and global MI on artificial CT/MR registrations and registrations complicated with a strong bias field; both a Parzen window and generalised partial volume kernel were used for histogram construction. In both experiments, cMI significantly outperforms global MI. Moreover, cMI is compared to global MI for the registration of three patient CT/MR datasets, using overlap and centroid distance as validation measure. The best results are obtained using cMI.

We propose two information theoretic similarity measures that allow to incorporate tissue class information in non-rigid image registration. The first measure assumes that tissue class probabilities have been assigned to each of the images to be registered by prior segmentation of both of them. One image is then non-rigidly deformed to match the other such that the fuzzy overlap of corresponding voxel object labels becomes similar to the ideal case whereby the tissue probability maps of both images are identical. Image similarity is assessed during registration by the divergence between the ideal and actual joint class probability distributions of both images. A second registration measure is proposed that applies in case a segmentation is available for only one of the images, for instance an atlas image that is to be matched to a study image to guide the segmentation thereof. Intensities in one image are matched to the fuzzy class labels in the other image by minimizing the conditional entropy of the intensities in the first image given the class labels in the second image. We derive analytic expressions for the gradient of each measure with respect to individual voxel displacements to derive a force field that drives the registration process, which is regularized by a viscous fluid model. The performance of the class-based measures is evaluated in the context of non-rigid inter-subject registration and atlas-based segmentation of MR brain images and compared with maximization of mutual information using only intensity information. Our results demonstrate that incorporation of class information in the registration measure significantly improves the overlap between corresponding tissue classes after non-rigid matching. The methods proposed here open new perspectives for integrating segmentation and registration in a single process, whereby the output of one is used to guide the other.

In this paper, we evaluate different schemes for constructing a mean shape anatomical atlas for atlas-based segmentation of MR brain images. Each atlas is constructed and validated using a database of 20 images for which detailed manual delineations of 49 different subcortical structures are available. Atlas construction and atlas based segmentation are performed by non-rigid intensity-based registration using a viscous fluid deformation model with parameters that were optimally tuned for this particular task. The segmentation performance of each atlas scheme is evaluated on the same database using a leave-one-out approach and measured by the volume overlap of corresponding regions in the ground-truth manual segmentation and the warped atlas label image.

We propose a statistical spline deformation model (SSDM) as a method to solve non-rigid image registration. Within this model, the deformation is expressed using a statistically trained B-spline deformation mesh. The model is trained by principal component analysis of a training set. This approach allows to reduce the number of degrees of freedom needed for non-rigid registration by only retaining the most significant modes of variation observed in the training set. User-defined transformation components, like affine modes, are merged with the principal components into a unified framework. Optimization proceeds along the transformation components rather then along the individual spline coefficients. The concept of SSDM's is applied to the temporal registration of thorax CR-images using pattern intensity as the registration measure. Our results show that, using 30 training pairs, a reduction of 33% is possible in the number of degrees of freedom without deterioration of the result. The same accuracy as without SSDM's is still achieved after a reduction up to 66% of the degrees of freedom.

Accurate brain tissue segmentation by intensity-based voxel classification of magnetic resonance (MR) images is complicated by partial volume (PV) voxels that contain a mixture of two or more tissue types. In this paper, we present a statistical framework for PV segmentation that encompasses and extends existing techniques. We start from a commonly used parametric statistical image model in which each voxel belongs to one single tissue type, and introduce an additional downsampling step that causes partial voluming along the borders between tissues. An expectation-maximization approach is used to simultaneously estimate the parameters of the resulting model and perform a PV classification. We present results on well-chosen simulated images and on real MR images of the brain, and demonstrate that the use of appropriate spatial prior knowledge not only improves the classifications, but is often indispensable for robust parameter estimation as well. We conclude that general robust PV segmentation of MR brain images requires statistical models that describe the spatial distribution of brain tissues more accurately than currently available models.