Automated analysis of investigational near-infrared fluorescence lymphatic imaging in humans.

ALFIA (Automated Lymphatic Function Imaging Analysis), an algorithm providing quantitative analysis of investigational near-infrared fluorescence lymphatic images, is described. Images from nine human subjects were analyzed for apparent lymphatic propagation velocities and propulsion periods using manual analysis and ALFIA. While lymphatic propulsion was more easily detected using ALFIA than with manual analysis, statistical analyses indicate no significant difference in the apparent lymphatic velocities although ALFIA tended to calculate longer propulsion periods. With the base ALFIA algorithms validated, further automation can now proceed to provide a clinically relevant analytic tool for quantitatively assessing lymphatic function in humans.

Cross-modality assessment and planning for pulmonary trunk treatment using CT and MRI imaging.

Congenital heart defect is the primary cause of death in newborns, due to typically complex malformation of the cardiac system. The pulmonary valve and trunk are often affected and require complex clinical management and in most cases surgical or interventional treatment. While minimal invasive methods are emerging, non-invasive imaging-based assessment tools become crucial components in the clinical setting. For advanced evaluation and therapy planning purposes, cardiac Computed Tomography (CT) and cardiac Magnetic Resonance Imaging (cMRI) are important non-invasive investigation techniques with complementary properties. Although, characterized by high temporal resolution, cMRI does not cover the full motion of the pulmonary trunk. The sparse cMRI data acquired in this context include only one 3D scan of the heart in the end-diastolic phase and two 2D planes (long and short axes) over the whole cardiac cycle. In this paper we present a cross-modality framework for the evaluation of the pulmonary trunk, which combines the advantages of both, cardiac CT and cMRI. A patient-specific model is estimated from both modalities using hierarchical learning-based techniques. The pulmonary trunk model is exploited within a novel dynamic regression-based reconstruction to infer the incomplete cMRI temporal information. Extensive experiments performed on 72 cardiac CT and 74 cMRI sequences demonstrated the average speed of 110 seconds and accuracy of 1.4mm for the proposed approach. To the best of our knowledge this is the first dynamic model of the pulmonary trunk and right ventricle outflow track estimated from sparse 4D cMRI data.

Discriminative distance functions and the patient neighborhood graph for clinical decision support.

There are two essential reasons for the slow progress in the acceptance of clinical similarity search-based decision support systems (DSSs); the especial complexity of biomedical data making it difficult to define a meaningful and effective distance function and the lack of transparency and explanation ability in many existing DSSs. In this chapter, we address these two problems by introducing a novel technique for visualizing patient similarity with neighborhood graphs and by considering two techniques for learning discriminative distance functions. We present an experimental study and discuss our implementation of similarity visualization within a clinical DSS.

Shape regression machine and efficient segmentation of left ventricle endocardium from 2D B-mode echocardiogram.

We present a machine learning approach called shape regression machine (SRM) for efficient segmentation of an anatomic structure that exhibits a deformable shape in a medical image, e.g., left ventricle endocardial wall in an echocardiogram. The SRM achieves efficient segmentation via statistical learning of the interrelations among shape, appearance, and anatomy, which are exemplified by an annotated database. The SRM is a two-stage approach. In the first stage that estimates a rigid shape to solve an automatic initialization problem, it derives a regression solution to object detection that needs just one scan in principle and a sparse set of scans in practice, avoiding the exhaustive scanning required by the state-of-the-art classification-based detection approach while yielding comparable detection accuracy. In the second stage that estimates the nonrigid shape, it again learns a nonlinear regressor to directly associate nonrigid shape with image appearance. The underpinning of both stages is a novel image-based boosting ridge regression (IBRR) method that enables multivariate, nonlinear modeling and accommodates fast evaluation. We demonstrate the efficiency and effectiveness of the SRM using experiments on segmenting the left ventricle endocardium from a B-mode echocardiogram of apical four chamber view. The proposed algorithm is able to automatically detect and accurately segment the LV endocardial border in about 120ms.

Robust height estimation of moving objects from uncalibrated videos.

This paper presents an approach for video metrology. From videos acquired by an uncalibrated stationary camera, we first recover the vanishing line and the vertical point of the scene based upon tracking moving objects that primarily lie on a ground plane. Using geometric properties of moving objects, a probabilistic model is constructed for simultaneously grouping trajectories and estimating vanishing points. Then we apply a single view mensuration algorithm to each of the frames to obtain height measurements. We finally fuse the multiframe measurements using the least median of squares (LMedS) as a robust cost function and the Robbins-Monro stochastic approximation (RMSA) technique. This method enables less human supervision, more flexibility and improved robustness. From the uncertainty analysis, we conclude that the method with auto-calibration is robust in practice. Results are shown based upon realistic tracking data from a variety of scenes.

Neighborhood graph and learning discriminative distance functions for clinical decision support.

There are two essential reasons for the slow progress in the acceptance of clinical case retrieval and similarity search-based decision support systems; the especial complexity of clinical data making it difficult to define a meaningful and effective distance function on them and the lack of transparency and explanation ability in many existing clinical case retrieval decision support systems. In this paper, we try to address these two problems by introducing a novel technique for visualizing inter-patient similarity based on a node-link representation with neighborhood graphs and by considering two techniques for learning discriminative distance function that help to combine the power of strong "black box" learners with the transparency of case retrieval and nearest neighbor classification.

Appearance modeling using a geometric transform.

A general transform, called the geometric transform (GeT), that models the appearance inside a closed contour is proposed. The proposed GeT is a functional of an image intensity function and a region indicator function derived from a closed contour. It can be designed to combine the shape and appearance information at different resolutions and to generate models invariant to deformation, articulation, or occlusion. By choosing appropriate functionals and region indicator functions, the GeT unifies Radon transform, trace transform, and a class of image warpings. By varying the region indicator and the types of features used for appearance modeling, five novel types of GeTs are introduced and applied to fingerprinting the appearance inside a contour. They include the GeTs based on a level set, shape matching, feature curves, and the GeT invariant to occlusion, and a multiresolution GeT (MRGeT). Applications of GeT to pedestrian identity recognition, human body part segmentation, and image synthesis are illustrated. The proposed approach produces promising results when applied to fingerprinting the appearance of a human and body parts despite the presence of nonrigid deformations and articulated motion.

Shape regression machine.

We present a machine learning approach called shape regression machine (SRM) to segmenting in real time an anatomic structure that manifests a deformable shape in a medical image. Traditional shape segmentation methods rely on various assumptions. For instance, the deformable model assumes that edge defines the shape; the Mumford-Shah variational method assumes that the regions inside/outside the (closed) contour are homogenous in intensity; and the active appearance model assumes that shape/appearance variations are linear. In addition, they all need a good initialization. In contrast, SRM poses no such restrictions. It is a two-stage approach that leverages (a) the underlying medical context that defines the anatomic structure and (b) an annotated database that exemplifies the shape and appearance variations of the anatomy. In the first stage, it solves the initialization problem as object detection and derives a regression solution that needs just one scan in principle. In the second stage, it learns a nonlinear regressor that predicts the nonrigid shape from image appearance. We also propose a boosting regression approach that supports real time segmentation. We demonstrate the effectiveness of SRM using experiments on segmenting the left ventricle endocardium from an echocardiogram of an apical four chamber view.

Appearance characterization of linear Lambertian objects, generalized photometric stereo, and illumination-invariant face recognition.

Traditional photometric stereo algorithms employ a Lambertian reflectance model with a varying albedo field and involve the appearance of only one object. In this paper, we generalize photometric stereo algorithms to handle all appearances of all objects in a class, in particular the human face class, by making use of the linear Lambertian property. A linear Lambertian object is one which is linearly spanned by a set of basis objects and has a Lambertian surface. The linear property leads to a rank constraint and, consequently, a factorization of an observation matrix that consists of exemplar images of different objects (e.g., faces of different subjects) under different, unknown illuminations. Integrability and symmetry constraints are used to fully recover the subspace bases using a novel linearized algorithm that takes the varying albedo field into account. The effectiveness of the linear Lambertian property is further investigated by using it for the problem of illumination-invariant face recognition using just one image. Attached shadows are incorporated in the model by a careful treatment of the inherent nonlinearity in Lambert's law. This enables us to extend our algorithm to perform face recognition in the presence of multiple illumination sources. Experimental results using standard data sets are presented.

From sample similarity to ensemble similarity: probabilistic distance measures in reproducing kernel Hilbert space.

This paper addresses the problem of characterizing ensemble similarity from sample similarity in a principled manner. Using reproducing kernel as a characterization of sample similarity, we suggest a probabilistic distance measure in the reproducing kernel Hilbert space (RKHS) as the ensemble similarity. Assuming normality in the RKHS, we derive analytic expressions for probabilistic distance measures that are commonly used in many applications, such as Chernoff distance (or the Bhattacharyya distance as its special case), Kullback-Leibler divergence, etc. Since the reproducing kernel implicitly embeds a nonlinear mapping, our approach presents a new way to study these distances whose feasibility and efficiency is demonstrated using experiments with synthetic and real examples. Further, we extend the ensemble similarity to the reproducing kernel for ensemble and study the ensemble similarity for more general data representations.

Image-based face recognition under illumination and pose variations.

We present an image-based method for face recognition across different illuminations and poses, where the term image-based means that no explicit prior three-dimensional models are needed. As face recognition under illumination and pose variations involves three factors, namely, identity, illumination, and pose, generalizations in all these three factors are desired. We present a recognition approach that is able to generalize in the identity and illumination dimensions and handle a given set of poses. Specifically, the proposed approach derives an identity signature that is illumination- and pose-invariant, where the identity is tackled by means of subspace encoding, the illumination is characterized with a Lambertian reflectance model, and the given set of poses is treated as a whole. Experimental results using the Pose, Illumination, and Expression (PIE) database demonstrate the effectiveness of the proposed approach.

Visual tracking and recognition using appearance-adaptive models in particle filters.

We present an approach that incorporates appearance-adaptive models in a particle filter to realize robust visual tracking and recognition algorithms. Tracking needs modeling interframe motion and appearance changes, whereas recognition needs modeling appearance changes between frames and gallery images. In conventional tracking algorithms, the appearance model is either fixed or rapidly changing, and the motion model is simply a random walk with fixed noise variance. Also, the number of particles is typically fixed. All these factors make the visual tracker unstable. To stabilize the tracker, we propose the following modifications: an observation model arising from an adaptive appearance model, an adaptive velocity motion model with adaptive noise variance, and an adaptive number of particles. The adaptive-velocity model is derived using a first-order linear predictor based on the appearance difference between the incoming observation and the previous particle configuration. Occlusion analysis is implemented using robust statistics. Experimental results on tracking visual objects in long outdoor and indoor video sequences demonstrate the effectiveness and robustness of our tracking algorithm. We then perform simultaneous tracking and recognition by embedding them in a particle filter. For recognition purposes, we model the appearance changes between frames and gallery images by constructing the intra- and extrapersonal spaces. Accurate recognition is achieved when confronted by pose and view variations.