We present a novel method for the automated extraction of blood vessel centerlines. There are two major contributions. First, in order to avoid the shortcuts to which minimal path methods are prone, we find optimal paths in a computed flow field. We solve for a steady state porous media flow inside a region of interest and trace centerlines as maximum flow paths. We explain how to estimate anisotropic orientation tensors which are used as permeability tensors in our flow field computation. Second, we introduce a convolutional neural network (CNN) classifier for removing extraneous paths in the detected centerlines. We apply our method to the extraction of coronary artery centerlines found in Computed Tomography Angiography (CTA). The robustness and stability of our method are enhanced by using a model-based detection of coronary specific territories and main branches to constrain the search space [ 15 ]. Validation against 20 comprehensively annotated datasets had a sensitivity and specificity at or above 90 %. Validation against 106 clinically annotated coronary arteries showed a sensitivity above 97 %. These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Tracking highly deforming structures in space and time arises in numerous applications in computer vision. Static Models are often referred to as linear combinations of a mean model and modes of variation learned from training examples. In Dynamic Modeling, the shape is represented as a function of shapes at previous time steps. In this paper, we introduce a novel technique that uses the spatial and the temporal information on the object deformation. We reformulate tracking as a high order time series prediction mechanism that adapts itself on-line to the newest results. Samples (toward dimensionality reduction) are represented in an orthogonal basis, and are introduced in an auto-regressive model that is determined through an optimization process in appropriate metric spaces. Toward capturing evolving deformations as well as cases that have not been part of the learning stage, a process that updates on-line both the orthogonal basis decomposition and the parameters of the autoregressive model is proposed. Experimental results with a nonstationary dynamic system prove adaptive AR models give better results than both stationary models and models learned over the whole sequence.

Recent advances in CT technology have made possible the acquisition of high resolution images of the heart in which important anatomical features such as the coronary vessels are visible. The state of the art in MR Imaging techniques, on the other hand, typically provides lower resolution but can facilitate the differentiation of viable (healthy) cardiac tissue from infarcted (injured) tissue. The registration of CT and MR images of the same heart permit the association of anatomy with function. As an example, given an infarcted region, the coronary vessels responsible for irrigating that region may be isolated and targeted for treatment (e.g., coronary bypass). We demonstrate a software package, PROTEUS, which is capable of registering CT and MR images. Since the manifestation of cardiac tissue in the two modalities differs so greatly we rely on features to perform the matching. Specifically, we employ a model-based approach that relies on segmented contours delineating the inner and outer borders of the left ventricle (LV). Given these contours, our technique is completely automatic.

We demonstrate a method for calculating the blood pool volume and myocardial mass of an in-vivo human Left-Ventricle (LV) non-invasively from multi-axis MR acquisitions. Segmented contours from both long and short-axis views are fused via a global model with parametric offsets which has a default configuration resembling a typical LV. From this model mass and volume may be computed. The results of the model fit are displayed for confirmation in a virtual reality environment which allows stereo viewing, multi-planar reformatting (MPR), and user interaction for model editing among other features.

We present a new algorithm for the automatic recovery of tag grid-line intersections in Tagged MR images of the left ventricle of the heart. Our method uses an active spring mesh to capture local properties of the motion and a global motion model to capture the global coherence of the motion. We recover the global component of the motion using robust estimation. Different motion models have been developed for short and long axis views of the heart. The algorithm has been tested on healthy and pathological data.

Shadows are a frequent occurrence, but they cannot be infallibly recognized until a scene's geometry and lighting are known. We present a number of cues which together strongly suggest the identification of a shadow and which can be examined with low cost. The techniques are: a color image segmentation method that recovers single material surfaces as single image regions irregardless of the surface partially in shadow, a method to recover the penumbra and umbra of shadow; a method for determining whether some object could be obstructing a light source. The last cue requires the examination of well understood shadows in the scene. Our observer is equipped with an extendable probe for casting its own shadows. Actively obtained shadows allow the observer to experimentally determine the location of the light sources in the scene. The system has been tested both indoors and out.>