Search

The non-destructive testing (NDT) of components is very important to the aerospace industry. Welds in these components may contain porosities and other defects. These reduce the fatigue life of components and may result in catastrophic accidents if they end up in the aircraft. Currently such welds are inspected by humans studying radiographs of the welds. We describe an automatic system for detecting defects in welds, with the aim of creating a triage system to reduce the workload on human inspectors. Given an X-ray image of the aerospace weld, the system locates the weld line, then analyses the region around the line to identify abnormalities. Our results show that the weld can be precisely extracted from X-ray images and the defect detection operation can identify 83% of defects with fewer than 3 false positives per image, and thus may be useful for prompting human inspectors to reduce their workload.

The detection of cartilage loss due to disease progression in Osteoarthritis remains a challenging problem. We have shown previously that the sensitivity of detection from 3D MR images can be improved significantly by focusing on regions of ‘at risk’ cartilage defined consistently across subjects and time-points. We define these regions in a frame of reference based on the bones, which requires that the bone surfaces are segmented in each image, and that anatomical correspondence is established between these surfaces. Previous results has shown that this can be achieved automatically using surface-based Active Appearance Models (AAMs) of the bones. In this paper we describe a method of refining the segmentations and correspondences by building a volumetric appearance model using the minimum message length principle. We present results from a study of 12 subjects which show that the new approach achieves a significant improvement in segmentation accuracy compared to the surface AAM approach, and reduce the variance in cartilage thickness measurements for key regions of interest. The study makes use of images of the same subjects obtained using different vendors' scanners, and also demonstrates the feasibility of multi-centre trials.

The objective of this work is to examine the feasibility of a method to register dynamic contrast enhanced computed X-ray tomography (DCE-CT) and dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) datasets in order to make possible the comparison of parametric maps generated from tracer kinetic modeling. First, the CT and MR dynamic sets were matched temporally using a cross-correlation maximization approach. The registration was then performed through an affine transformation followed by a non-linear registration using free-form deformations (FFDs) based on B-splines. This was determined from the CT-MR pair that maximized Normalized Mutual Information (NMI). Then the 'extended Kety' model was fitted to both CT and MR and Ktrans, Ve and vp parameters were obtained. The method was applied to 5 patients with bladder tumors. After registration, the overlap matching between CT and MR volume of interest (VOI) was on average 91%.

We address the problem of extracting information from groups of medical images of the same anatomy. We describe an algorithm which simultaneously segments and registers a set of such images, incrementally constructing a model of their structure and the correspondences across the set. The framework explicitly models the fraction of each tissue type, rather than the expected intensity in each voxel, to decouple the model from details of the imaging sequence and modality. When estimating the optimal deformation field, the current image is compared to a reconstructed image, generated from the model tissue fractions and the current estimate of intensity distributions for each tissue type in the current image (i.e. an estimate of how the model would appear given the imaging conditions for that image). We describe the algorithm in detail and present results of applying it to a set of MR images of the brain

We have described previously a method of automatically constructing statistical models of shape. The method treats model-building as an optimisation problem by re-parameterising each shape so as to minimise the description length of the training set. The approach requires an explicit parameterisation of each shape, which is straightforward in 2D, but non-trivial in 3D. It is necessary to provide some parameterisation of the training set to initialise the optimisation. An inappropriate initial parameterisation can cause the optimisation to converge at a slower rate or stop it from converging to a satisfactory solution. In this paper we describe a method of producing a consistent parameterisation for a given set of surfaces. The consistent parameterisations were used to initialise the model-building algorithm and produced results that were significantly better than alternative approaches.

Degeneration and loss of articular cartilage in osteoarthritis (OA) is difficult to measure, because changes are small and localised. We present a method that uses statistical shape models of the knee bones to define an anatomically consistent frame of reference across a population, providing sensitive measures of cartilage morphology in anatomically equivalent regions of interest. Bone and cartilage were manually segmented from magnetic resonance images (MRI) of volunteers' knees. Dense correspondences were defined across all subjects by constructing minimum description length (MDL) statistical shape models of the bones. Regions of interest were manually delineated on the mean bone shapes provided by the models, and propagated to each individual in an anatomically consistent manner, using the model-based correspondences. We show that this approach results in precise measurements that can be used to detect small localised changes in cartilage thickness. Results are reported for an OA study, in which significant focal loss of cartilage was detected over 6 months in a cohort of just 31 patients.

In this study we have developed techniques to compare local changes in signal intensity over time in fat suppressed 3D gradient echo MR images of articular cartilage in patients with osteoarthritis. We have applied these techniques to data taken over a periods of 1 and 3 years in two groups of patients with established OA of the knee. In both these studies, no significant change in total cartilage volume could be detected but we were able to observe some significant changes in signal intensity. We conclude that in a study of cartilage structure this technique can provide additional information without the overhead of extra scans.

We present a method for aggregating and comparing articular cartilage thickness in a population of patients using the underlying bone as the frame of reference. Femur, patella and tibia knee bone surfaces are constructed from segmentations of MR scans for 19 healthy female patients. Dense correspondences are defined on each example by building a minimum description length (MDL) statistical shape model of the bone shapes. Using repeat scans of the same subjects, these correspondences are shown to identify anatomically equivalent locations. Cartilage thickness is measured at corresponding points and aggregated for all patients to establish a normal range in healthy females. Linear discriminant analysis and permutation testing are used to visualise class differences and evaluate their significance. We illustrate this using a study of inter-segmentor differences.

In medical image analysis there are various examples where the detection of linear structures can provide important information. These include mammography, blood vessel detection and the extraction of trebecular structure from bone images. There are several generic techniques which can be used to obtain linear structure information at a pixel level. Several of these techniques are compared on the basis of their performance in detecting line strength and orientation. In addition the authors comment on the possibility of using the same techniques in a multi-scale approach to also obtain line scale information. The methods discussed include those based on simple orientation bins, a multidirectional line operator, directional second order Gaussian derivatives, directional morphology, curvilinear structures detection and directional Fourier space.

An important part of quality control in the fabrication of integrated circuit wafers involves making 3D measurements of the wafer's surface topography. We describe a method for obtaining a dense depth map of the of the surface from a single scanning electron microscope (SEM) image. The depth map gives a complete description of the surface topography, allowing the easy calculation of any measurement required in the quality control process. The technique presented uses a novel scale-space shape-from-shading algorithm. Results are shown for real SEM images and for a synthetic test image. We also describe how the method can be extended to use stereo images in order to achieve greater accuracy in the surface reconstruction. >

The 'viewpoint' approach, in which a system is described by several partial specifications, has been proposed as a way of making complex computing systems more understandable. The ISO's Open Distributing Processing (ODP) framework is an architecture for open distributed systems, involving five named viewpoints. This paper compares two partial specifications of a lending library-from the ODP's Enterprise and Information Viewpoints-and discusses the relation between them. Both specifications are written in Object-Z, an object-oriented variant of Z. Examining how such partial specifications might be unified raises broader issues of refinement and mutual consistency of partial specifications in Object-Z.

This paper presents advances in the construction and use of Active Appearance Models (AAMs) for image interpretation. AAMs are photo-realistic generative models of object appearance that can be used to rapidly locate deformable objects in images. We extend the AAM method to include coloured texture and present an enhanced search algorithm with the ability to locate partially occluded objects. Previously, AAMs have been limited by the need for good manual initialisation. In this paper, we describe a hierarchical search algorithm that overcomes this drawback. The extended AAM method provides a complete, unified scheme for model based image interpretation. We demonstrate the application of the scheme to the task of locating faces in images.

The absolute orientation problem involves finding the Euclidean transformation which minimises the sum of the squared errors between two pointsets. In the standard form of the problem a confidence may be attached to each of the errors via a set of positive scalar weights. In this paper we consider a generalisation of the standard problem in which the components of the error vectors are coupled via a set of weight matrices. We show how problems of this type arise and derive two distinct forms of the problem. We present a closed-form solution to the first form of the 3-D problem and iterative solutions to the second form of the 2-D problem and both forms of the 3-D problem.

We describe a system that performs fully automatic analysis of total hip replacement (THR) radiographs using a model-based approach. Previous methods for analysis of this type of radiograph have been manual or semi-automatic and require considerable investment of skilled operators time. Our system works unsupervised and is the first of its type. Image search is carried out using a statistically defined deformable template used as an active shape model (ASM). The ASM, which is trained to model prosthetic, femoral, and pelvic structures within the image, iteratively modifies its pose and shape to fit to the evidence in a given image. We present the results of systematic experiments which show that the system can robustly locate relevant structures in routine post-operative radiographs accurately enough for practical applications.