Ching-Wen Yang

Generalized constrained energy minimization approach to subpixel target detection for multispectral imagery

Subpixel detection in multispectral imagery presents a chal- lenging problem due to relatively low spatial and spectral resolution. We present a generalized constrained energy minimization (GCEM) ap- proach to detecting targets in multispectral imagery at subpixel level. GCEM is a hybrid technique that combines a constrained energy mini- mization (CEM) method developed for hyperspectral image classification with a dimensionality expansion (DE) approach resulting from a gener- alized orthogonal subspace projection (GOSP) developed for multispec- tral image classification. DE enables us to generate additional bands from original multispectral images nonlinearly so that CEM can be used for subpixel detection to extract targets embedded in multispectral im- ages. CEM has been successfully applied to hyperspectral target detec- tion and image classification. Its applicability to multispectral imagery is yet to be investigated. A potential limitation of CEM on multispectral imagery is the effectiveness of interference elimination due to the lack of sufficient dimensionality. DE is introduced to mitigate this problem by expanding the original data dimensionality. Experiments show that the proposed GCEM detects targets more effectively than GOSP and CEM without dimensionality expansion.

Detection of spectral signatures in multispectral MR images for classification

Presents a new spectral signature detection approach to magnetic resonance (MR) image classification. It is called constrained energy minimization (CEM) method, which is derived from the minimum variance distortionless response in passive sensor array processing. It considers a bank of spectral channels as an array of sensors where each spectral channel represents a sensor and object spectral signature in multispectral MR images are viewed as signals impinging upon the array. The strength of the CEM lies on its ability in detection of spectral signatures of interest without knowing image background. The detected spectral signatures are then used for classification. The CEM makes use of a finite impulse response (FIR) filter to linearly constrain a desired object while minimizing interfering effects caused by other unknown signal sources. Unlike most spatial-based classification techniques, the proposed CEM takes advantage of spectral characteristics to achieve object detection and classification. A series of experiments is conducted and compared with the commonly used c-means method for performance evaluation. The results show that the CEM method is a promising and effective spectral technique for MR image classification.

Band Expansion-Based Over-Complete Independent Component Analysis for Multispectral Processing of Magnetic Resonance Images

Independent component analysis (ICA) has found great promise in magnetic resonance (MR) image analysis. Unfortunately, two key issues have been overlooked and not investigated. One is the lack of MR images to be used to unmix signal sources of interest. Another is the use of random initial projection vectors by ICA, which causes inconsistent results. In order to address the first issue, this paper introduces a band-expansion process (BEP) to generate an additional new set of images from the original MR images via nonlinear functions. These newly generated images are then combined with the original MR images to provide sufficient MR images for ICA analysis. In order to resolve the second issue, a prioritized ICA (PICA) is designed to rank the ICA-generated independent components (ICs) so that MR brain tissue substances can be unmixed and separated by different ICs in a prioritized order. Finally, BEP and PICA are combined to further develop a new ICA-based approach, referred to as PICA-BEP to perform MR image analysis.

3-D localization of clustered microcalcifications using cranio-caudal and medio-lateral oblique view

This paper presents a 3D localization method to register clustered microcalcifications on mammograms from cranio-caudal (CC) and medio-lateral oblique (MLO) views. The method consists of three major components: registration of clustered microcalcifications in CC and MLO views, 3D localization of clustered microcalcifications and 3D visualization of clustered microcalcifications. The registration is performed based on three features, gradient, energy and local entropy codes that are independent of spatial locations of microcalcifications in two different views and are prioritized by discriminability in a binary decision tree. The 3D localization is determined by a sequence of coordinate corrections of calcified pixels using the breast nipple as a controlling point. Finally, the 3D visualization implements a virtual reality modeling language viewer (VRMLV) to view the exact location of the lesion as a guide for needle biopsy. In order to validate our proposed 3D localization system, a set of breast lesions, which appear both in mammograms and in MR Images is used for experiments where the depth of clustered microcalcifications can be verified by the MR images.

Unsupervised orthogonal subspace projection approach to magnetic resonance image classification

MR images and remotely sensed images share similar image structures and characteristics because they are acquired remotely as image sequences by spectral channels at different wavelengths. As a result, techniques developed for one may be also applicable to the other. In the past, we have witnessed that some techniques that were developed for magnetic resonance imaging (MRI) found great success in remote sensing image applications. Unfortunately, the opposite direction is yet to be investigated. In this paper, we present an application of one successful remote sensing image classification technique, called orthogonal subspace projection (OSP), to magnetic resonance image classification. Unlike classical image classification techniques, which are designed on a pure pixel basis, OSP is a mixed pixel classification technique that models an image pixel as a linear mixture of different material substances assumed to be present in the image data, then estimates the abundance fraction of each individual material substance within a pixel for classification. Technically, such mixed pixel classification is performed by estimating the abundance fractions of material substances resident in a pixel, rather than assigning a class label to it as usually done in pure-pixel-based classification techniques such as a minimum-distance or nearest-neighbor rule. The advantage of mixed pixel classification has been demonstrated in many applications in remote sensing image processing. The MRI experiments reported in this paper further show that mixed pixel classification may have advantages over the pure pixel classification.

Computer-aided diagnostic detection system of venous beading in retinal images

The detection of venous beading in retinal images provides an early sign of diabetic retinopathy and plays an important role as a preprocessing step in diagnosing ocular diseases. We present a computer-aided diagnostic system to automatically detect venous bead- ing of blood vessels. It comprises of two modules, referred to as the blood vessel extraction module (BVEM) and the venus beading detection module (VBDM). The former uses a bell-shaped Gaussian kernel with 12 azimuths to extract blood vessels while the latter applies a neural network-based shape cognitron to detect venous beading among the extracted blood vessels for diagnosis. Both modules are fully computer- automated. To evaluate the proposed system, 61 retinal images (32 beaded and 29 normal images) are used for performance evaluation.

Quantitative analysis in clinical applications of brain MRI using independent component analysis coupled with support vector machine

To effectively perform quantification of brain normal tissues and pathologies simultaneously, independent component analysis (ICA) coupled with support vector machine (SVM) is investigated and evaluated for effective volumetric measurements of normal and lesion tissues using multispectral MR images.

Independent component analysis for magnetic resonance image analysis

Independent component analysis (ICA) has recently received considerable interest in applications of magnetic resonance (MR) image analysis. However, unlike its applications to functional magnetic resonance imaging (fMRI) where the number of data samples is greater than the number of signal sources to be separated, a dilemma encountered in MR image analysis is that the number of MR images is usually less than the number of signal sources to be blindly separated. As a result, at least two or more brain tissue substances are forced into a single independent component (IC) in which none of these brain tissue substances can be discriminated from another. In addition, since the ICA is generally initialized by random initial conditions, the final generated ICs are different. In order to resolve this issue, this paper presents an approach which implements the over-complete ICA in conjunction with spatial domain-based classification so as to achieve better classification in each of ICA-demixed ICs. In order to demonstrate the proposed over-complete ICA, (OC-ICA) experiments are conducted for performance analysis and evaluation. Results show that the OC-ICA implemented with classification can be very effective, provided the training samples are judiciously selected.

HIERARCHICAL FAST TWO-DIMENSIONAL ENTROPIC THRESHOLDING ALGORITHM USING A HISTOGRAM PYRAMID

Entropic thresholding provides an alternative view to the de- sign rationale of conventional thresholding. Abutaleb proposed a second-order entropic thresholding approach to improve Puns first-order entropic thresholding by introducing a 2-D gray-level histogram to take into account the spatial correlation. Abutalebs (1989) method was fur- ther modified by Brink (1992). However, a major drawback of Abutaleb- Brinks method is very high computational cost. Recently, Chen et al. (1994) developed a fast efficient 2-D algorithm that reduces computa- tional complexity from O(L 4 )t o O ( L 8/3 ), where L is the total number of gray levels. A hierarchical fast 2-D entropic thresholding algorithm using a gray-level histogram pyramid is presented that can be viewed as a generalization of Chen et al.s algorithm. The new algorithm consists of a 2-D gray-level histogram pyramid build-up procedure expanding Ab- utalebs 2-D gray-level histogram to a histogram pyramid, and a thresh- olding process applying a modified version of Chen et al.s algorithm to the histogram pyramid layer by layer from top to bottom. As a result, the computational complexity of Chen et al.s algorithm can be further re- duced to the optimal complexity, O(L 2 ). The experiments show that the computer time of the new algorithm is only one tenth of that required for Chen et al.s algorithm, which is a significant saving.

A hierarchical model for pacs

In this paper, a hierarchical model for Picture Archiving and Communication Systems (HPACS) is presented and implemented at Taichung Veterans General Hospital (TCVGH) in Taiwan. Despite the fact that the HPACS is built on the architecture of the second generation PACS, it offers many improved features and has advantages over the second generation PACS, such as the user security control, fast resource dispatch and efficient resource management. This HPACS can be used as a reference model for a hospital with any scale-size. The real implementation of HPACS is currently undertaken in the Taichung Veterans General Hospital (TCVGH), Taiwan, Republic of China and consists of four phases with the first two phases already completed. It is the first pilot system ever to be implemented successfully in a large-scale hospital in Taiwan. The experiences have illustrated the great promise of the HPACS in the future.