Catálogo de publicaciones - libros

This paper presents a new approach to regional myocardial contractility analysis based on inter-landmark motion (ILM) vectors and multivariate outlier detection. The proposed spatio-temporal representation is used to describe the coupled changes occurring at pairs of regions of the left ventricle, thus enabling the detection of geometrical and dynamic inconsistencies. Multivariate tolerance regions are derived from training samples to describe the variability within the normal population using the ILM vectors. For new left ventricular datasets, outlier detection enables the localization of extreme ILM observations and the corresponding myocardial abnormalities. The framework is validated on a relatively large sample of 50 subjects and the results show promise in localization and visualization of regional left ventricular dysfunctions.

This paper presents a novel orthopedics surgery training system with both the components for modeling as well as simulating the deformation and visualization in an efficient way. By employing techniques such as optimization, segmentation and center line extraction, the modeling of deformable model can be completed with minimal manual involvement. The novel trainer can simulate rigid body, soft tissue and blood with state-of-the-art techniques, so that convincing deformation and realistic bleeding can be achieved. More important, newly released Physics Processing Unit (PPU) is adopted to tackle the high requirement for physics related computations. Experiment shows that the acceleration gain from PPU is significant for maintaining interactive frame rate under a complex surgical environments of orthopedics surgery.

Accurately describing interactions between medical devices and anatomical structures, or between anatomical structures themselves, is an essential step towards the adoption of computer-based medical simulation as an alternative to traditional training methods. However, while substantial work has been done in the area of real-time soft tissue modeling, little has been done to study the problem of contacts occurring during tissue manipulation. In this paper we introduce a new method for correctly handling complex contacts between various combination of rigid and deformable objects. Our approach verifies Signorini’s law by combining Lagrange multipliers and the status method to solve unilateral constraints. Our method handles both concave and convex surfaces by using a displacement subdivision strategy, and the proposed algorithm allows interactive computation times even in very constrained situations. We demonstrate the efficiency of our approach in the context of interventional radiology, with the navigation of catheters and guidewires in tortuous vessels and with the deployment of coils to treat aneurysms.

A fundamental prerequisite of hysteroscopy is the proper distension of the uterine cavity with a fluid, also known as hydrometra. For a virtual reality based simulation of hysteroscopy, the uterus deformation process due to different pressure settings has to be modeled. In previous work we have introduced a hybrid method, which relies on precomputed deformation states to derive the hydrometra changes during runtime. However, new offline computations were necessary for every newly introduced organ mesh. This is not viable if a new surgical scene is to be generated for every training session. Therefore, we include the deformation states during hydrometra into our previously developed statistical shape model for undeformed organ instances. This allows deriving the hydrometra steps together with new undeformed uterus meshes. These can then be used during the interactive simulation for predicting uterus deformation without time-intensive precomputation steps.

With the increasing versatility of CMR, further understanding of intrinsic contractility of the myocardium can be achieved by performing subject-specific modeling by integrating structural and functional information available. The recent introduction of the virtual tagging framework allows for visualization of the localized deformation of the myocardium based on phase contrast myocardial velocity mapping. The purpose of this study is to examine the use of a non-linear, Kernel-Partial Least Squares Regression (K-PLSR) predictive motion modeling scheme for the virtual tagging framework. The method allows for the derivation of a compact non-linear deformation model such that the entire deformation field can be predicted by a limited number of control points. When applied to virtual tagging, the technique can be used to predictively guide the mesh refinement based on the motion of the coarse grid, thus greatly reducing the search space and increasing the convergence speed of the algorithm. The effectiveness and numerical accuracy of the proposed technique are assessed with both numerically simulated data sets and phase contrast CMR velocity mapping from a group of 7 subjects. The technique presented has a distinct advantage over the conventional mesh refinement scheme and brings CMR myocardial contractility analysis closer to routine clinical practice.

We present a model of solid tumor growth which can account for several stages of tumorigenesis, from the early avascular phase to the angiogenesis driven proliferation. The model combines several previously identified components in a consistent framework, including neoplastic tissue growth, blood and oxygen transport, and angiogenic sprouting. First experiments with the framework and comparisons with observations made on solid tumors in vivo illustrate the plausibility of the approach. Explanations of several experimental observations are naturally provided by the model. To the best of our knowledge this is the first report of a model coupling tumor growth and angiogenesis.

The importance of accurate early diagnostics of autism that severely affects personal behavior and communication skills cannot be overstated. Neuropathological studies have revealed an abnormal anatomy of the cerebral white matter (CWM) in autistic brains. We explore a possibility of distinguishing between autistic and normal brains by a quantitative shape analysis of CWM gyrifications on 3D proton density MRI (PD-MRI) images. Our approach consists of () segmentation of the CWM on a 3D brain image using a deformable 3D boundary; () extraction of gyrifications from the segmented CWM, and () shape analysis to quantify thickness of the extracted gyrifications and classify autistic and normal subjects. The boundary evolution is controlled by two probabilistic models of visual appearance of 3D CWM: the learned prior and the current appearance model. Initial experimental results suggest that the proposed 3D texture analysis is a promising supplement to the current techniques for diagnosing autism.

With the ability to study brain anatomy in vivo using magnetic resonance imaging, studies on regional brain atrophy suggest possible improvements for differential diagnosis of movement disorders with parkinsonian symptoms. In this study, we investigate effects of different parkinsonian syndromes on the cortical gray matter thickness and the geometric shape of the cerebral cortex. The study consists of a total of 24 patients with a diagnosis of probable progressive supranuclear palsy (PSP), multiple systems atrophy (MSA) or idiopathic Parkinson’s disease (IPD). We examine dense estimates of cortical gray matter thickness, sulcal depth, and measures of the curvature in a surface-based cortical morphometry analysis framework. Group difference results indicate higher cortical atrophy rate in the frontal lobe in PSP patients when compared to either MSA or IPD. These findings are indicative of the potential use of routine MRI and cortical morphometry in performing differential diagnosis in PSP, MSA and IPD.

This paper is the first report on the analysis of ultrasound RF echo time series acquired using high frequency ultrasound. We show that variations in the intensity of one sample of RF echo over time is correlated with tissue microstructure. To form the RF time series, a high frequency probe and a tissue sample were fixed in position and RF signals backscattered from the tissue were continuously recorded. The fractal dimension of RF time series was used as a feature for tissue classification. Feature values acquired from different areas of one tissue type were statistically similar. For animal tissues with different cellular microstructure, we successfully used the fractal dimension of RF time series to distinguish segments as small as 20 microns with accuracies as high as 98%. The results of this study demonstrate that the analysis of RF time series is a promising approach for distinguishing tissue types with different cellular microstructure.

Nuclear medicine imaging modalities assist commonly in surgical guidance given their functional nature. However, when used in the operating room they present limitations. Pre-operative tomographic 3D imaging can only serve as a vague guidance intra-operatively, due to movement, deformation and changes in anatomy since the time of imaging, while standard intra-operative nuclear measurements are limited to 1D or (in some cases) 2D images with no depth information. To resolve this problem we propose the synchronized acquisition of position, orientation and readings of gamma probes intra-operatively to reconstruct a 3D activity volume. In contrast to conventional emission tomography, here, in a first proof-of-concept, the reconstruction succeeds without requiring symmetry in the positions and angles of acquisition, which allows greater flexibility. We present our results in phantom experiments for sentinel node lymph node localization. The results indicate that 3D intra-operative nuclear images can be generated in such a setup up to an accuracy equivalent to conventional SPECT systems. This technology has the potential to advance standard procedures towards intra-operative 3D nuclear imaging and offers a novel approach for robust and precise localization of functional information to facilitate less invasive, image-guided surgery.