Chen Wu-fan

Automatic segmentation of coronary angiograms based on fuzzy inferring and probabilistic tracking

BackgroundSegmentation of the coronary angiogram is important in computer-assisted artery motion analysis or reconstruction of 3D vascular structures from a single-plan or biplane angiographic system. Developing fully automated and accurate vessel segmentation algorithms is highly challenging, especially when extracting vascular structures with large variations in image intensities and noise, as well as with variable cross-sections or vascular lesions.MethodsThis paper presents a novel tracking method for automatic segmentation of the coronary artery tree in X-ray angiographic images, based on probabilistic vessel tracking and fuzzy structure pattern inferring. The method is composed of two main steps: preprocessing and tracking. In preprocessing, multiscale Gabor filtering and Hessian matrix analysis were used to enhance and extract vessel features from the original angiographic image, leading to a vessel feature map as well as a vessel direction map. In tracking, a seed point was first automatically detected by analyzing the vessel feature map. Subsequently, two operators [e.g., a probabilistic tracking operator (PTO) and a vessel structure pattern detector (SPD)] worked together based on the detected seed point to extract vessel segments or branches one at a time. The local structure pattern was inferred by a multi-feature based fuzzy inferring function employed in the SPD. The identified structure pattern, such as crossing or bifurcation, was used to control the tracking process, for example, to keep tracking the current segment or start tracking a new one, depending on the detected pattern.ResultsBy appropriate integration of these advanced preprocessing and tracking steps, our tracking algorithm is able to extract both vessel axis lines and edge points, as well as measure the arterial diameters in various complicated cases. For example, it can walk across gaps along the longitudinal vessel direction, manage varying vessel curvatures, and adapt to varying vessel widths in situations with arterial stenoses and aneurysms.ConclusionsOur algorithm performs well in terms of robustness, automation, adaptability, and applicability. In particular, the successful development of two novel operators, namely, PTO and SPD, ensures the performance of our algorithm in vessel tracking.

Automated lung segmentation algorithm for CAD system of thoracic CT

Abstract Objective To design and test the accuracy and efficiency of our lung segmentation algorithm on thoracic CT image in computer-aided diagnostic (CAD) system, especially on the segmentation between left and right lungs. Methods We put forward the base frame of our lung segmentation firstly. Then, using optimal thresholding and mathematical morphologic methods, we acquired the rough image of lung segmentation. Finally, we presented a fast self-fit segmentation refinement algorithm, adapting to the unsuccessful left-right lung segmentation of thredsholding. Then our algorithm was used to CT scan images of 30 patients and the results were compared with those made by experts. Results Experiments on clinical 2-D pulmonary images showed the results of our algorithm were very close to the experts manual outlines, and it was very effective for the separation of left and right lungs with a successful segmentation ratio 94.8%. Conclusion It is a practicable fast lung segmentation algorithm for CAD system on thoracic CT image.

New approach to the automatic segmentation of coronary artery in X-ray angiograms

For the segmentation of X-ray angiograms (XRA), the essential feature and the prior knowledge of angiographic image were analyzed, and a multi-feature based fuzzy recognition (MFFR) algorithm was proposed to infer the local vessel structure in this paper. Guided by the prior knowledge of artery vessel, a probability tracking operator (PTO) can rapidly track along the artery tree, and walk across the weak region or gaps because of disturbance or preprocessing to angiographic image. Another, the accurate measurement of the vascular axis-lines and diameters can be synchronously implemented in the tracking process. To correctly evaluate the proposed method, a simulated image of CAT and some clinical XRA images were used in the experimentations. The algorithms performed better than the conventional one: given one start-point, on average 92.7% of the visible segments or branches was automatically delineated; the correctness ratio of vessel structure inference reached to 90.0% on the average.

A New Method for Detection of Pulmonary Nodules

In order to find the pulmonary nodules at the early stage of lung cancer or in duty check, we must detect the pulmonary nodules quickly and exhaustively, especially the small nodules which are often omitted by doctors. A new TMH (template-matching with Hessian matrix) method is proposed in this paper. First lung lobes are gained. Fhen FM (template-matching method) is used to detect the pulmonary nodules. Fhen the Hessian matrix is imported into the method. With this new method, we can decrease the true positive rate markedly. FMH method has been proved to be quick, complete and has low true positive rate.

Unsupervised segmentation of medical image based on difference of mutual information

In the scope of medical image processing, segmentation is important and difficult. There are still two problems which trouble us in this field. One is how to determine the number of clusters in an image and the other is how to segment medical images containing lesions. A new segmentation method called DDC, based on difference of mutual information (dMI) and pixon, is proposed in this paper. Experiments demonstrate that dMI shows one kind of intrinsic relationship between the segmented image and the original one and so it can be used to well determine the number of clusters. Furthermore, multi-modality medical images with lesions can be automatically and successfully segmented by DDC method.

Improved Algorithm for Motion Estimation in the Reconstruction of PROPELLER MRI Data

Algorithms to reliably and accurately extract inter-strip motion are crucial for motion artifacts suppression in PROPELLER MRI. The current algorithm is based on image registration through maximizing correlation coefficients of k-space data. To decouple rotational and translational estimation, only magnitude of k-space data is taken into account while performing rotational estimation. Because little data are contained in the central overlapped sampling area, the robustness and precision of the k-space based rotational estimation are degraded by the discarding of useful phase information. In this paper, an improved algorithm through optimizing correlation in image space is proposed. First, k-space data are transformed into image space by zero-padded reconstruction of each strip. Then the rotational and translational parameters are acquired simultaneously through maximizing correlation coefficients between these images with POWELL optimization. Phantom and in vivo imaging demonstrate that the proposed image-space-based algorithm is of a much higher accuracy than the k-space-based algorithm.

MR image segmentation by nonparametric model

Nonparametric Dirichlet Process Mixtures (MDP) model algorithm is applied to segment images, which can obtain the segmentation class numbers automatically without initialization. The algorithm is used to segment noisy natural images and magnetic resonance images with biasing field. Compared with classical Markov Field (MRF) segmentation, the nonparametric segmentation results show the greater performance. This method is also analyzed quantitatively on the belly magnetic resonance images. The Dice Similarity Coefficients (DSC) of all slices exceed 93%, which show that the proposed method is robust and accurate.

Improved Reconstruction of Parallel MR Data Using Smoothing Constraints

MR data acquisition efficiency can be greatly improved by the spatial sensitivity encoding using multiple coils in parallel MR imaging. However, when large reduce factor are chosen, scanning noise may be amplified and leads to serious artifacts in the image, due to the ill-conditioning in matrix inversion in the reconstruction. In this paper, a new framework, which can incorporate advanced edge-preserving smoothing techniques, is proposed for the regularized reconstruction of parallel MR data. Under the proposed framework, algorithm with smoothing constraints based on non-local means, which has the advantage of preserving the small structures well while filtering out noise, is presented for parallel imaging. Experiment on in-vivo brain imaging using the array of 8 coils shows that the noise and artifacts in the final reconstruction with large reduction factor can be better suppressed with the proposed algorithm.