Elizabeth Bullitt

A brain tumor segmentation framework based on outlier detection

This paper describes a framework for automatic brain tumor segmentation from MR images. The detection of edema is done simultaneously with tumor segmentation, as the knowledge of the extent of edema is important for diagnosis, planning, and treatment. Whereas many other tumor segmentation methods rely on the intensity enhancement produced by the gadolinium contrast agent in the T1-weighted image, the method proposed here does not require contrast enhanced image channels. The only required input for the segmentation procedure is the T2 MR image channel, but it can make use of any additional non-enhanced image channels for improved tissue segmentation. The segmentation framework is composed of three stages. First, we detect abnormal regions using a registered brain atlas as a model for healthy brains. We then make use of the robust estimates of the location and dispersion of the normal brain tissue intensity clusters to determine the intensity properties of the different tissue types. In the second stage, we determine from the T2 image intensities whether edema appears together with tumor in the abnormal regions. Finally, we apply geometric and spatial constraints to the detected tumor and edema regions. The segmentation procedure has been applied to three real datasets, representing different tumor shapes, locations, sizes, image intensities, and enhancement.

Phase II Trial of Lapatinib for Brain Metastases in Patients With Human Epidermal Growth Factor Receptor 2–Positive Breast Cancer

PURPOSE One third of women with advanced human epidermal growth factor receptor 2 (HER-2)-positive breast cancer develop brain metastases; a subset progress in the CNS despite standard approaches. Medical therapies for refractory brain metastases are neither well-studied nor established. We evaluated the safety and efficacy of lapatinib, an oral inhibitor of epidermal growth factor receptor (EGFR) and HER-2, in patients with HER-2-positive brain metastases. PATIENTS AND METHODS Patients had HER-2-positive breast cancer, progressive brain metastases, prior trastuzumab treatment, and at least one measurable metastatic brain lesion. Patients received lapatinib 750 mg orally twice a day. Tumor response was assessed by magnetic resonance imaging every 8 weeks. The primary end point was objective response (complete response [CR] plus partial response [PR]) in the CNS by Response Evaluation Criteria in Solid Tumors (RECIST). Secondary end points included objective response in non-CNS sites, time to progression, overall survival, and toxicity. RESULTS Thirty-nine patients were enrolled. All patients had developed brain metastases while receiving trastuzumab; 37 had progressed after prior radiation. One patient achieved a PR in the brain by RECIST (objective response rate 2.6%, 95% conditional CI, 0.21% to 26%). Seven patients (18%) were progression free in both CNS and non-CNS sites at 16 weeks. Exploratory analyses identified additional patients with some degree of volumetric reduction in brain tumor burden. The most common adverse events (AEs) were diarrhea (grade 3, 21%) and fatigue (grade 3, 15%). CONCLUSION The study did not meet the predefined criteria for antitumor activity in highly refractory patients with HER-2-positive brain metastases. Because of the volumetric changes observed in our exploratory analysis, further studies are underway utilizing volumetric changes as a primary end point.

Induction of c-fos-like protein within the lumbar spinal cord and thalamus of the rat following peripheral stimulation

Noxious stimulation induces c-fos-like protein within neurons of the spinal cord and thalamus in patterns that suggest c-fos induction may serve as a marker for activity within nociresponsive pathways. Similar patterns of immunoreactivity were not seen following gentle mechanical stimulation or in control animals. Within the thalamus, noxious stimulation induces immunoreactivity not only in traditionally expected locations, but also within the paraventricular, submedial and reuniens nuclei.

Level-set evolution with region competition: automatic 3-D segmentation of brain tumors

We develop a new method for automatic segmentation of anatomical structures from volumetric medical images. Driving application is tumor segmentation from 3-D MRIs, which is known to be a very challenging problem due to the variability of tumor geometry and intensity patterns. Level-set snakes offer significant advantages over conventional statistical classification and mathematical morphology, however snakes with constant propagation need careful initialization and can leak through weak or missing boundary parts. Our region competition method overcomes these problems by modulating the propagation term with a signed local statistical force, leading to a stable solution. A pre- vs. post-contrast difference image is used to calculate probabilities for background and tumor regions, with a mixture-modelling fit of the histogram. Preliminary results on five cases with significant shape and intensity variability demonstrate that the new method might become a powerful and efficient tool for the clinic. Validity is demonstrated by comparison with manual expert segmentation.

Automatic Brain Tumor Segmentation by Subject Specific Modification of Atlas Priors

RATIONALE AND OBJECTIVES Manual segmentation of brain tumors from magnetic resonance images is a challenging and time-consuming task. An automated system has been developed for brain tumor segmentation that will provide objective, reproducible segmentations that are close to the manual results. Additionally, the method segments white matter, grey matter, cerebrospinal fluid, and edema. The segmentation of pathology and healthy structures is crucial for surgical planning and intervention. MATERIALS AND METHODS The method performs the segmentation of a registered set of magnetic resonance images using an expectation-maximization scheme. The segmentation is guided by a spatial probabilistic atlas that contains expert prior knowledge about brain structures. This atlas is modified with the subject-specific brain tumor prior that is computed based on contrast enhancement. RESULTS Five cases with different types of tumors are selected for evaluation. The results obtained from the automatic segmentation program are compared with results from manual and semi-automated methods. The automated method yields results that have surface distances at roughly 1-4 mm compared with the manual results. CONCLUSION The automated method can be applied to different types of tumors. Although its performance is below that of the semi-automated method, it has the advantage of requiring no user supervision.

The effect of stimulus duration on noxious-stimulus induced c-fos expression in the rodent spinal cord

C-fos is a proto-oncogene that is expressed within some neurons following depolarization. The protein product, fos, has been proposed as an anatomical marker for neuronal activity following noxious peripheral stimulation. However, the literature on noxious-stimulus induced fos expression contains several puzzling observations on the time course and laminar distribution of neuronal labeling within the spinal cord. This study has analyzed the effect of stimulus duration on the expression of fos-like immunoreactivity (FLI) within the spinal cord of anesthetized rats. In order to examine the time course of fos expression following brief periods of stimulation, we required a type of stimulus that was intense enough to activate nociceptors but that did not produce tissue damage. We have therefore employed pulsed, high intensity electrical stimulation, with stimulus durations ranging from 3 s to 24 h. The results indicate that stimulus duration has a profound effect upon the number of labeled cells, the intensity of neuronal labeling, the laminar pattern of FLI, and the time course of fos expression. Brief stimulation periods induce relatively few and relatively lightly labeled neurons, located predominantly within the most superficial laminae of the dorsal horn. Maximal immunoreactivity appears approximately 2 h after stimulation has ceased, and disappears within hours. Continuous stimulation produces many more labeled cells, darker labeling, and FLI within both dorsal and ventral laminar regions. Maximal FLI is seen after approximately 4.5 h of continuous stimulation, with reduction in the number of labeled cells thereafter. These data indicate that the results of any study employing c-fos as a marker for neuronal activity may be affected by the duration of the exciting stimulus.

Intensity ridge and widths for tubular object segmentation and description

Introduces a technique for the automated description of tubular objects in 3D medical images. The goal of automated 3D object description is to extract a representation which consistently details the location, size, and structure of objects in 3D images using minimal user interaction. Such a representation provides a means by which objects can be classified, quantifiably evaluated, and registered. It also serves as a region of interest specification for visualization processes. The technique presented in this paper is suited for generating representations of 3D objects with nearly circular cross sections which have, possibly as a result of a global operation (e.g., blurring), intensity extrema near their centers. Such tubular objects commonly occur within human anatomy (e.g., vessels and selected bones). The medial axis of each of these objects is well approximated by its intensity ridge. The scales of the local maxima in medialness at all points along the ridge can be mapped to local width estimates. Together these measures capture the location, size, and structure of tubular objects. This paper covers the mathematical basis, the implementation issues, and the application of this technique to the extraction of vessels from 3D magnetic resonance angiographic images and bones from 3D X-ray computed tomographic images.

Registration and Analysis of Vascular Images

We have developed a method for rigidly aligning images of tubes. This paper presents an evaluation of the consistency of that method for three-dimensional images of human vasculature. Vascular images may contain alignment ambiguities, poorly corresponding vascular networks, and non-rigid deformations, yet the Monte Carlo experiments presented in this paper show that our method registers vascular images with sub-voxel consistency in a matter of seconds. Furthermore, we show that the methods insensitivity to non-rigid deformations enables the localization, quantification, and visualization of those deformations.Our method aligns a source image with a target image by registering a model of the tubes in the source image directly with the target image. Time can be spent to extract an accurate model of the tubes in the source image. Multiple target images can then be registered with that model without additional extractions.Our registration method builds upon the principles of our tubular object segmentation work that combines dynamic-scale central ridge traversal with radius estimation. In particular, our registration methods consistency stems from incorporating multi-scale ridge and radius measures into the model-image match metric. Additionally, the methods speed is due in part to the use of coarse-to-fine optimization strategies that are enabled by measures made during model extraction and by the parameters inherent to the model-image match metric.

Population Shape Regression From Random Design Data

Regression analysis is a powerful tool for the study of changes in a dependent variable as a function of an independent regressor variable, and in particular it is applicable to the study of anatomical growth and shape change. When the underlying process can be modeled by parameters in a Euclidean space, classical regression techniques are applicable and have been studied extensively. However, recent work suggests that attempts to describe anatomical shapes using flat Euclidean spaces undermines our ability to represent natural biological variability. In this paper we develop a method for regression analysis of general, manifold-valued data. Specifically, we extend Nadaraya-Watson kernel regression by recasting the regression problem in terms of Frechet expectation. Although this method is quite general, our driving problem is the study anatomical shape change as a function of age from random design image data. We demonstrate our method by analyzing shape change in the brain from a random design dataset of MR images of 89 healthy adults ranging in age from 22 to 79 years. To study the small scale changes in anatomy, we use the infinite dimensional manifold of diffeomorphic transformations, with an associated metric. We regress a representative anatomical shape, as a function of age, from this population.

Multi-Modal Image Set Registration and Atlas Formation

In this paper, we present a Bayesian framework for both generating inter-subject large deformation transformations between two multi-modal image sets of the brain and for forming multi-class brain atlases. In this framework, the estimated transformations are generated using maximal information about the underlying neuroanatomy present in each of the different modalities. This modality independent registration framework is achieved by jointly estimating the posterior probabilities associated with the multi-modal image sets and the high-dimensional registration transformations mapping these posteriors. To maximally use the information present in all the modalities for registration, Kullback-Leibler divergence between the estimated posteriors is minimized. Registration results for image sets composed of multi-modal MR images of healthy adult human brains are presented. Atlas formation results are presented for a population of five infant human brains.