Pau-Choo Chung

An automatic diagnostic system for CT liver image classification

Computed tomography (CT) images have been widely used for liver disease diagnosis. Designing and developing computer-assisted image processing techniques to help doctors improve their diagnosis has received considerable interests over the past years. In this paper, a CT liver image diagnostic classification system is presented which will automatically find, extract the CT liver boundary and further classify liver diseases. The system comprises a detect-before-extract (DBE) system which automatically finds the liver boundary and a neural network liver classifier which uses specially designed feature descriptors to distinguish normal liver, two types of liver tumors, hepatoma and hemageoma. The DBE system applies the concept of the normalized fractional Brownian motion model to find an initial liver boundary and then uses a deformable contour model to precisely delineate the liver boundary. The neural network is included to classify liver tumors into hepatoma and hemageoma. It is implemented by a modified probabilistic neural network (PNN) [MPNN] in conjunction with feature descriptors which are generated by fractal feature information and the gray-level co-occurrence matrix. The proposed system was evaluated by 30 liver cases and shown to be efficient and very effective.

Robust radial basis function neural networks

Function approximation has been found in many applications. The radial basis function (RBF) network is one approach which has shown a great promise in this sort of problems because of its faster learning capacity. A traditional RBF network takes Gaussian functions as its basis functions and adopts the least-squares criterion as the objective function, However, it still suffers from two major problems. First, it is difficult to use Gaussian functions to approximate constant values. If a function has nearly constant values in some intervals, the RBF network will be found inefficient in approximating these values. Second, when the training patterns incur a large error, the network will interpolate these training patterns incorrectly. In order to cope with these problems, an RBF network is proposed in this paper which is based on sequences of sigmoidal functions and a robust objective function. The former replaces the Gaussian functions as the basis function of the network so that constant-valued functions can be approximated accurately by an RBF network, while the latter is used to restrain the influence of large errors. Compared with traditional RBF networks, the proposed network demonstrates the following advantages: (1) better capability of approximation to underlying functions; (2) faster learning speed; (3) better size of network; (4) high robustness to outliers.

Unsupervised texture segmentation via wavelet transform

Abstract In this paper, a mechanism for unsupervised texture segmentation is presented. The approach is based on the multiscale representation of the discrete (dyadic) wavelet transform which can be implemented by a fast iterative algorithm. For unsupervised segmentation it is generally difficult to determine the number of classes to be identified. The proposed approach offers an approach to circumvent this problem. Our method utilizes a set of high-frequency channel energies to characterize texture features, followed by a multi-thresholding technique for coarse segmentation. The coarsely segmented results at the same scale are incorporated by an intea-scale fusion procedure. A fine segmentation technique is then used to reclassify the ambiguously labeled pixels generated from the intea-scale fusion step. Finally, the number of texture classes is determined by an inter-scale fusion in which the segmentation results at multiple scales are integrated. The performance of this method is demonstrated by several experiments on synthetic images, natural textures from Brodatzs album and real-world textured images. Since the choice of wavelets is very extensive and open, we further explore various types of wavelets for texture segmentation. The time cost of the proposed method is also measured.

A daily behavior enabled hidden Markov model for human behavior understanding

This paper presents a Hierarchical Context Hidden Markov Model (HC-HMM) for behavior understanding from video streams in a nursing center. The proposed HC-HMM infers elderly behaviors through three contexts which are spatial, activities, and temporal context. By considering the hierarchical architecture, HC-HMM builds three modules composing the three components, reasoning in the primary and the secondary relationship. The spatial contexts are defined from the spatial structure, so that it is placed as the primary inference contexts. The temporal duration is associated to elderly activities, so activities are placed in the following of spatial contexts and the temporal duration is placed after activities. Between the spatial context reasoning and behavior reasoning of activities, a modified duration HMM is applied to extract activities. According to this design, human behaviors different in spatial contexts would be distinguished in first module. The behaviors different in activities would be determined in second module. The third module is to recognize behaviors involving different temporal duration. By this design, an abnormal signaling process corresponding to different situations is also placed for application. The developed approach has been applied for understanding of elder behaviors in a nursing center. Results have indicated the promise of the approach which can accurately interpret 85% of the elderly behaviors. For abnormal detection, the approach was found to have 90% accuracy, with 0% false alarm.

Polygonal approximation using a competitive Hopfield neural network

Abstract Polygonal approximation plays an important role in pattern recognition and computer vision. In this paper, a parallel method using a Competitive Hopfield Neural Network (CHNN) is proposed for polygonal approximation. Based on the CHNN, the polygonal approximation is regarded as a minimization of a criterion function which is defined as the arc-to-chord deviation between the curve and the polygon. The CHNN differs from the original Hopfield network in that a competitive winner-take-all mechanism is imposed. The winner-take-all mechanism adeptly precludes the necessity of determining the values for the weighting factors in the energy function in maintaining a feasible result. The proposed method is compared to several existing methods by the approximation error norms L2 and L∞ with the result that promising approximation polygons are obtained.

Identifying multiple abdominal organs from CT image series using a multimodule contextual neural network and spatial fuzzy rules

Identifying abdominal organs is one of the essential steps in visualizing organ structure to assist in teaching, clinical training, diagnosis, and medical image retrieval. However, due to partial volume effects, gray-level similarities of adjacent organs, contrast media affect, and the relatively high variations of organ position and shape, automatically identifying abdominal organs has always been a high challenging task. To conquer these difficulties, this paper proposes combining a multimodule contextual neural network and spatial fuzzy rules and fuzzy descriptors for automatically identifying abdominal organs from a series of CT image slices. The multimodule contextual neural network segments each image slice through a divide-and-conquer concept, embedded within multiple neural network modules, where the results obtained from each module are forwarded to other modules for integration, in which contextual constraints are enforced. With this approach, the difficulties arising from partial volume effects, gray-level similarities of adjacent organs, and contrast media affect can be reduced to the extreme. To address the issue of high variations in organ position and shape, spatial fuzzy rules and fuzzy descriptors are adopted, along with a contour modification scheme implementing consecutive organ region overlap constraints. This approach has been tested on 40 sets of abdominal CT images, where each set consists of about 40 image slices. We have found that 99% of the organ regions in the test images are correctly identified as its belonging organs, implying the high promise of the proposed method.

Contrast context histogram-An efficient discriminating local descriptor for object recognition and image matching

In this paper, we propose a new invariant local descriptor, called the contrast context histogram (CCH), for image matching and object recognition. By representing the contrast distributions of a local region, it serves as a distinctive local descriptor of the region. Our experiments demonstrate that contrast-based local descriptors can represent local regions with more compact histogram bins. Because of its high matching accuracy and efficient computation, the CCH has the potential to be used in a number of real-time applications.

A computer-aided design mammography screening system for detection and classification of microcalcifications

This paper presents a prototype of a computer-aided design (CAD) diagnostic system for mammography screening to automatically detect and classify microcalcifications (MCCs) in mammograms. It comprises four modules. The first module, called the Mammogram Preprocessing Module, inputs and digitizes mammograms into 8-bit images of size 2048x2048, extracts the breast region from the background, enhances the extracted breast and stores the processed mammograms in a data base. Since only clustered MCCs are of interest in providing a sign of breast cancer, the second module, called the MCCs Finder Module, finds and locates suspicious areas of clustered MCCs, called regions of interest (ROIs). The third module, called the MCCs Detection Module, is a real time computer automated MCCs detection system that takes as inputs the ROIs provided by the MCCs Finder Module. It uses two different window sizes to automatically extract the microcalcifications from the ROIs. It begins with a large window of size 64x64 to quickly screen mammograms to find large calcified areas, this is followed by a smaller window of size 8x8 to extract tiny, isolated microcalcifications. Finally, the fourth module, called the MCCs Classification Module, classifies the detected clustered microcalcifications into five categories according to BI-RADS (Breast Imaging Reporting and Data System) format recommended by the American College of Radiology. One advantage of the designed system is that each module is a separate component that can be individually upgraded to improve the whole system. Despite that it is still is a prototype system a preliminary clinical evaluation at TaiChung Veterans General Hospital (TCVGH) has shown that the system is very flexible and can be integrated with the existing Picture Archiving and Communications System (PACS) currently implemented in the Department of Radiology at TCVGH.

Color image segmentation using fuzzy C-means and eigenspace projections

In this paper, we propose two eigen-based fuzzy C-means (FCM) clustering algorithms to accurately segment the desired images, which have the same color as the pre-selected pixels. From the selected color pixels, we can first divide the color space into principal and residual eigenspaces. Combined eigenspace transform and the FCM method, we can effectively achieve color image segmentation. The separate eigenspace FCM (SEFCM) algorithm independently applies the FCM method to principal and residual projections to obtain two intermediate segmented images and combines them by logically selecting their common pixels. Jointly considering principal and residual eigenspace projections, we then suggest the coupled eigen-based FCM (CEFCM) algorithm by using an eigen-based membership function in clustering procedure. Simulations show that the proposed SEFCM and CEFCM algorithms can successfully segment the desired color image with substantial accuracy.

Mining Physiological Conditions from Heart Rate Variability Analysis

This article presents a successfully developed methodology for mining physiological conditions from heart rate variability (HRV) analysis. The application of HRV analysis in both research and clinical settings has seen rapid development in the past decades. Unlike previous research, this study employed features derived from longterm monitoring of HRV indices, as these trends can best reflect the autonomic nervous system dynamics influenced by various physiological conditions. We proposed two methods for mining physiological conditions from HRV trends: a decision-tree learning method and a hybrid learning method that combines feature selection, feature extraction, and classifier construction processes. The proposed methods have been validated through a clinical case study: severity classification for Parkinsons disease. Our approach yielded classification accuracy greater than 90.0%, and high sensitivity, specificity, positive predictive values (PPV), and negative predictive values (NPV).