A novel adaptive level set segmentation method.

The adaptive distance preserving level set (ADPLS) method is fast and not dependent on the initial contour for the segmentation of images with intensity inhomogeneity, but it often leads to segmentation with compromised accuracy. And the local binary fitting model (LBF) method can achieve segmentation with higher accuracy but with low speed and sensitivity to initial contour placements. In this paper, a novel and adaptive fusing level set method has been presented to combine the desirable properties of these two methods, respectively. In the proposed method, the weights of the ADPLS and LBF are automatically adjusted according to the spatial information of the image. Experimental results show that the comprehensive performance indicators, such as accuracy, speed, and stability, can be significantly improved by using this improved method.

[A graph cuts-based interactive method for segmentation of magnetic resonance images of meningioma].

For accurate segmentation of the magnetic resonance (MR) images of meningioma, we propose a novel interactive segmentation method based on graph cuts. The high dimensional image features was extracted, and for each pixel, the probabilities of its origin, either the tumor or the background regions, were estimated by exploiting the weighted K-nearest neighborhood classifier. Based on these probabilities, a new energy function was proposed. Finally, a graph cut optimal framework was used for the solution of the energy function. The proposed method was evaluated by application in the segmentation of MR images of meningioma, and the results showed that the method significantly improved the segmentation accuracy compared with the gray level information-based graph cut method.

Bayesian statistical reconstruction for low-dose X-ray computed tomography using an adaptive-weighting nonlocal prior.

How to reduce the radiation dose delivered to the patients has always been a important concern since the introduction of computed tomography (CT). Though clinically desired, low-dose CT images can be severely degraded by the excessive quantum noise under extremely low X-ray dose circumstances. Bayesian statistical reconstructions outperform the traditional filtered back-projection (FBP) reconstructions by accurately expressing the system models of physical effects and the statistical character of the measurement data. This work aims to improve the image quality of low-dose CT images using a novel AW nonlocal (adaptive-weighting nonlocal) prior statistical reconstruction approach. Compared to traditional local priors, the proposed prior can adaptively and selectively exploit the global image information. It imposes an effective resolution-preserving and noise-removing regularization for reconstructions. Experimentation validates that the reconstructions using the proposed prior have excellent performance for X-ray CT with low-dose scans.

[A fast approach for level set segmentation].

To propose an optimal level set approach for fast medical image segmentation. By confining the computation quantity of the level sets function and using the image characteristics, we improved the efficiency of segmentation and decreased the parameter setting in some degree for DSA vascular segmentation.

[Automatic feature extraction and new method for retrieval from CT image database].

OBJECTIVE: To propose a new method for content-based retrieval from medical CT image database on the basis of automatically extracted features of the images. METHODS: An automatic feature extraction method is proposed based on expectation-maximization algorithm. A CT image is represented by a set of regions, each of which is characterized by a fuzzy regional feature vector reflecting the grey level, texture, shape, and the cumulative distribution histogram feature of the region of interest (ROI) to efficiently describe the difference between the ROIs. RESULTS: Compared with the submitted query image, the target images were retrieved in the order of similarity calculated by the proposed similarity measures. CONCLUSION: The proposed technique for CT image retrieval is suitable for clinical application, with greater precision and efficiency for retrieval than the conventional methods.

Automatic segmentation of cardiac magnetic resonance images using knowledge base.

OBJECTIVE: To study the automated implementation of cardiac magnetic resonance image (MRI) segmentation. METHODS: By training the feature parameters of the images and establishing a knowledge base, an efficient method for extracting and using prior knowledge was proposed. RESULTS and CONCLUSION: Through extracting and using prior knowledge of cardiac MRI, automation of cardiac MRI segmentation can be well accomplished.