Chia-Ling Tsai

The dual-bootstrap iterative closest point algorithm with application to retinal image registration

Motivated by the problem of retinal image registration, this paper introduces and analyzes a new registration algorithm called Dual-Bootstrap Iterative Closest Point (Dual-Bootstrap ICP). The approach is to start from one or more initial, low-order estimates that are only accurate in small image regions, called bootstrap regions. In each bootstrap region, the algorithm iteratively: 1) refines the transformation estimate using constraints only from within the bootstrap region; 2) expands the bootstrap region; and 3) tests to see if a higher order transformation model can be used, stopping when the region expands to cover the overlap between images. Steps 1): and 3), the bootstrap steps, are governed by the covariance matrix of the estimated transformation. Estimation refinement [Step 2)] uses a novel robust version of the ICP algorithm. In registering retinal image pairs, Dual-Bootstrap ICP is initialized by automatically matching individual vascular landmarks, and it aligns images based on detected blood vessel centerlines. The resulting quadratic transformations are accurate to less than a pixel. On tests involving approximately 6000 image pairs, it successfully registered 99.5% of the pairs containing at least one common landmark, and 100% of the pairs containing at least one common landmark and at least 35% image overlap.

Registration of Challenging Image Pairs: Initialization, Estimation, and Decision

Our goal is an automated 2D-image-pair registration algorithm capable of aligning images taken of a wide variety of natural and man-made scenes as well as many medical images. The algorithm should handle low overlap, substantial orientation and scale differences, large illumination variations, and physical changes in the scene. An important component of this is the ability to automatically reject pairs that have no overlap or have too many differences to be aligned well. We propose a complete algorithm including techniques for initialization, for estimating transformation parameters, and for automatically deciding if an estimate is correct. Keypoints extracted and matched between images are used to generate initial similarity transform estimates, each accurate over a small region. These initial estimates are rank-ordered and tested individually in succession. Each estimate is refined using the Dual-Bootstrap ICP algorithm, driven by matching of multiscale features. A three-part decision criteria, combining measurements of alignment accuracy, stability in the estimate, and consistency in the constraints, determines whether the refined transformation estimate is accepted as correct. Experimental results on a data set of 22 challenging image pairs show that the algorithm effectively aligns 19 of the 22 pairs and rejects 99.8 percent of the misalignments that occur when all possible pairs are tried. The algorithm substantially out-performs algorithms based on keypoint matching alone.

Model-based method for improving the accuracy and repeatability of estimating vascular bifurcations and crossovers from retinal fundus images

A model-based algorithm, termed exclusion region and position refinement (ERPR), is presented for improving the accuracy and repeatability of estimating the locations where vascular structures branch and cross over, in the context of human retinal images. The goal is two fold. First, accurate morphometry of branching and crossover points (landmarks) in neuronal/vascular structure is important to several areas of biology and medicine. Second, these points are valuable as landmarks for image registration, so improved accuracy and repeatability in estimating their locations and signatures leads to more reliable image registration for applications such as change detection and mosaicing. The ERPR algorithm is shown to reduce the median location error from 2.04 pixels down to 1.1 pixels, while improving the median spread (a measure of repeatability) from 2.09 pixels down to 1.05 pixels. Errors in estimating vessel orientations were similarly reduced from 7.2/spl deg/ down to 3.8/spl deg/.

Automated Retinal Image Analysis Over the Internet

Retinal clinicians and researchers make extensive use of images, and the current emphasis is on digital imaging of the retinal fundus. The goal of this paper is to introduce a system, known as retinal image vessel extraction and registration system, which provides the community of retinal clinicians, researchers, and study directors an integrated suite of advanced digital retinal image analysis tools over the Internet. The capabilities include vasculature tracing and morphometry, joint (simultaneous) montaging of multiple retinal fields, cross-modality registration (color/red-free fundus photographs and fluorescein angiograms), and generation of flicker animations for visualization of changes from longitudinal image sequences. Each capability has been carefully validated in our previous research work. The integrated Internet-based system can enable significant advances in retina-related clinical diagnosis, visualization of the complete fundus at full resolution from multiple low-angle views, analysis of longitudinal changes, research on the retinal vasculature, and objective, quantitative computer-assisted scoring of clinical trials imagery. It could pave the way for future screening services from optometry facilities.

An Uncertainty-Driven Hybrid of Intensity-Based and Feature-Based Registration with Application to Retinal and Lung CT Images

A new hybrid of feature-based and intensity-based registration is presented. The algorithm reflects a new understanding of the role of alignment error in the generation of registration constraints. This leads to an iterative process where distinctive image locations from the moving image are matched against the intensity structure of the fixed image. The search range of this matching process is controlled by both the uncertainty in the current transformation estimate and the properties of the image locations to be matched. The resulting hybrid algorithm is applied to retinal image registration by incorporating it as the main estimation engine within our recently published Dual-Bootstrap ICP algorithm. The hybrid algorithm is used to align serial and 4d CT images of the lung using a B-spline based deformation model.

A View-Based Approach to Registration: Theory and Application to Vascular Image Registration

This paper presents an approach to registration centered on the notion of a view--a combination of an image resolution, a transformation model, an image region over which the model currently applies, and a set of image primitives from this region. The registration process is divided into three stages: initialization, automatic view generation, and estimation. For a given initial estimate, the latter two alternate until convergence; several initial estimates may be explored. The estimation process uses a novel generalization of the Iterative Closest Point (ICP) technique that simultaneously considers multiple correspondences for each point. View-based registration is applied successfully to alignment of vascular and neuronal images in 2-d and 3-d using similarity, affine, and quadratic transformations.

Automatic Characterization of Classic Choroidal Neovascularization by Using AdaBoost for Supervised Learning

PURPOSE To provide a computer-aided visualization tool for accurate diagnosis and quantification of choroidal neovascularization (CNV) on the basis of fluorescence leakage characteristics. METHODS All image frames of a fluorescein angiography (FA) sequence are first aligned and mapped to a global space. To automatically determine the severity of each pixel in the global space and hence the extent of CNV, the system matches the intensity variation of each set of spatially corresponding pixels across the sequence with the targeted leakage pattern, learned from a sampled population graded by a retina specialist. The learning strategy, known as the AdaBoost algorithm, has 12 classifiers for 12 features that summarize the variation in fluorescence intensity over time. Given a new sequence, the severity map image is generated using the contribution scores of the 12 classifiers. Initialized with points of low and high severity, regions of CNV are delineated using the random walk algorithm. RESULTS A dataset of 33 FA sequences of classic CNV showed the average accuracy of CNV delineation to be 83.26%. In addition, the 30- to 60-second interval provided the most reliable information for differentiating CNV from the background. Using eight sequences of multiple visits of four patients for evaluation of the postphotodynamic therapy (PDT), the statistics derived from the segmented regions correlate closely with the clinical observed changes. CONCLUSIONS The clinician can easily visualize the temporal characteristics of CNV fluorescence leakage using the severity map, which is a two-dimensional summary of a complete FA sequence. The computer-aided tool allows objective evaluation and computation of statistical data from the automatic delineation for surgical assessment.

Retinal Vascular Tree Reconstruction With Anatomical Realism

Motivated by the goals of automatically extracting vessel segments and constructing retinal vascular trees with anatomical realism, this paper presents and analyses an algorithm that combines vessel segmentation and grouping of the extracted vessel segments. The proposed method aims to restore the topology of the vascular trees with anatomical realism for clinical studies and diagnosis of retinal vascular diseases, which manifest abnormalities in either venous and/or arterial vascular systems. Vessel segments are grouped using extended Kalman filter which takes into account continuities in curvature, width, and intensity changes at the bifurcation or crossover point. At a junction, the proposed method applies the minimum-cost matching algorithm to resolve the conflict in grouping due to error in tracing. The system was trained with 20 images from the DRIVE dataset, and tested using the remaining 20 images. The dataset contained a mixture of normal and pathological images. In addition, six pathological fluorescein angiogram sequences were also included in this study. The results were compared against the groundtruth images provided by a physician, achieving average success rates of 88.79% and 90.09%, respectively.

Disease-Oriented Evaluation of Dual-Bootstrap Retinal Image Registration

This paper presents a disease-oriented evaluation of two recent retinal image registration algorithms, one for aligning pairs of retinal images and one for simultaneously aligning all images in a set. Medical conditions studied include diabetic retinopathy, vein occlusion, and both dry and wet age-related macular degeneration. The multi-image alignment worked virtually flawlessly, missing only 2 of 855 images. Pairwise registration, the Dual-Bootstrap ICP algorithm, worked nearly as well, successfully aligning 99.5% of the image pairs having a sufficient set of common features and 78.5% overall. Images of retinas having an edema and pairs of images taken before and after laser treatment proved the most difficult to register.

Vascular Tree Construction with Anatomical Realism for Retinal Images

In this paper, we present a method to automatically extract the vessel segments and construct the vascular tree with anatomical realism from a color retinal image. The significance of the work is to assist in clinical studies of diagnosis of cardio-vascular diseases, such as hypertension,which manifest abnormalities in either venous and/or arterial vascular systems. To maximize the completeness of vessel extraction, we introduce vessel connectiveness measure to improve on an existing algorithm which applies multiscale matched filtering and vessel likelihood measure.Vessel segments are grouped using extended Kalman filter to take into consideration continuities in curvature, width,and color changes at the bifurcation or crossover point. The algorithm is tested on five images from the DRIVE database,a mixture of normal and pathological images, and the results are compared with the ground truth images provided by a physician. The preliminary results show that our method reaches an average success rate of 92.1%.