Yannis A. Tolias

Image segmentation by a fuzzy clustering algorithm using adaptive spatially constrained membership functions

We present an adaptive fuzzy clustering scheme for image segmentation, the adaptive fuzzy clustering/segmentation (AFCS) algorithm. In AFCS, the nonstationary nature of images is taken into account by modifying the prototype vectors as functions of the sample location in the image. The inherent high interpixel correlation is modeled using neighborhood information. A multiresolution model is utilized for estimating the spatially varying prototype vectors for different window sizes. The fuzzy segmentations at different resolutions are combined using a data fusion process in order to compute the final fuzzy partition matrix. The results provide segmentations, having lower fuzzy entropy when compared to the possibilistic C-means algorithm, while maintaining the images main characteristics. In addition, due to the neighborhood model, the effects of noise in the form of single pixel regions are minimized.

On applying spatial constraints in fuzzy image clustering using a fuzzy rule-based system

A novel approach for enhancing the results of fuzzy clustering by imposing spatial constraints for solving image segmentation problems is presented. We have developed a Sugeno (185) type rule-based system with three inputs and 11 rules that interacts with the clustering results obtained by the well-known fuzzy c-means (FCM) and/or possibilistic c-means (PCM) algorithms. It provides good image segmentations in terms of region smoothness and elimination of the effects of noise. The results of the proposed rule-based neighborhood enhancement (RB-NE) system are compared to well-known segmentation algorithms using stochastic field modeling. They are found to be of comparable quality, while being of lower computational complexity.

Generalized fuzzy indices for similarity matching

The purpose of this paper is to introduce a new family of fuzzy similarity indices, referred to as the generalised Tversky index (GTI). The development of GTI is based on the theoretical findings by Amos Tversky regarding the human perception of similarity between different objects, as formulated by the Features Contrast model (FC). Although GTI was developed starting from Tverskys FC, it is shown that it provides a fuzzy extension and generalization of several widely used similarity indices like the Jaccard and Dice coefficients. The mathematical properties of two members of the GTI family, namely TIM and TIP, are studied and their interpretation of similarity is explained by comparison to other conventional indices.

An orthogonal least squares-based fuzzy filter for real-time analysis of lung sounds

Pathological discontinuous adventitious sounds (DAS) are strongly related with the pulmonary dysfunction. Its clinical use for the interpretation of respiratory malfunction depends on their efficient and objective separation from vesicular sounds (VS). In this paper, an automated approach to the isolation of DAS from VS, based on their nonstationarity, is presented. The proposed scheme uses two fuzzy inference systems (FISs), operating in parallel, to perform the task of adaptive separation, resulting in the orthogonal least squares-based fuzzy filter (OLS-FF). By applying the OLS-FF to fine/coarse crackles and squawks, selected from three lung sound databases, the coherent structure of DAS is revealed and they are efficiently separated from VS. The important time domain DAS features, related to diagnostic information, are preserved and their true location and structural morphology are automatically identified. When compared to previous works, the OLS-FF performs quite similarly, but with significantly lower computational load, resulting in a faster real-time clinical screening of DAS.

A fuzzy rule-based system for real-time separation of crackles from vesicular sounds

An automated way of revealing the diagnostic character of crackle by isolating them from vesicular sounds (VS) is presented in this paper. The proposed algorithm takes into account crackle nonstationarity and uses fuzzy rules in order to compose a fuzzy-based stationary-nonstationary filter (FST-NST). Applying the FST-NST filter to fine/coarse crackles, selected from three lung sound databases, the coherent structure of crackles is revealed and they are separated from VS. The resulted separation is accurate, objective, and of a high quality, since the FTST-NST filter automatically identifies the true location of crackles in the original signal and maintains their structure in its nonstationary output. Due to its simple and fast implementation it can easily be used as an on-line crackle identification system in clinical medicine.

Real-time separation of discontinuous adventitious sounds from vesicular sounds using a fuzzy rule-based filter

The separation of pathological discontinuous adventitious sounds (DAS) from vesicular sounds (VS) is of great importance to the analysis of lung sounds since DAS are related to certain pulmonary pathologies. An automated way of revealing the diagnostic character of DAS, by isolating them from VS, based on their nonstationarity, is presented. The proposed algorithm uses two adaptive network-based fuzzy inference systems to compose a generalized fuzzy rule-based stationary-nonstationary filter (GFST-NST). The training procedure of the fuzzy inference systems involves the outputs of the wavelet transform-based stationary-nonstationary filter (WTST-NST), proposed by Hadjileontiadis and Panas (1997). The basic idea of the GFST-NST was initially proposed by the authors with the introduction of the fuzzy rule-based stationary-nonstationary filter (FST-NST) (1997), tested with the separation of crackles from VS. The main contribution of this paper is the modification of the structure of the FST-NST filter to a serial-type fuzzy filter that, unlike the parallel operation of the FST-NST filter, sends a predicted stationary signal (VS) into the predictor of the nonstationary (DAS). Applying the GFST-NST filter to fine-coarse crackles and squawks, selected from three lung sound databases, the coherent structure of DAS is revealed and they are separated from VS. The separation performance of the GFST-NST filter was evaluated through quantitative and qualitative indexes that proved its efficiency and superiority against the FST-NST filter. When compared to the WTST-NST filter, the GFST-NST filter performed similarly in accuracy and objectiveness, but in a faster way. Thus, the GFST-NST filter combines the separation accuracy of the WTST-NST filter with the real-time implementation of the FST-NST filter, so it can easily be used in clinical medicine as a module of an integrated intelligent patient evaluation system.

ICASP: An Intensive-Care Acquisition and Signal Processing Integrated Framework

This paper presents an intensive-care acquisition and signal processing integrated framework in the area of intensive care units. The framework includes nearly all monitored biosignals in the intensive care, along with metadata and processing results. It is structured on two basic applications, i.e., the acquisition and the database one, running in two different PCs that are connected through a local area network, facilitating real-time data exchange between them. The analytical rundown shows that the proposed framework is a serious effort to give a complete clinical condition of a patient and a form of a diagnostic analysis implement in the intensive care by taking in real-time processing.

A hierarchical edge-stressing algorithm for adaptive image segmentation

In this paper we present a new multiresolution edge-stressing approach for segmenting images. Our algorithm utilises the wavelet transform to obtain a multiresolution representation of the image. The low frequency residuals of each stage of the wavelet transform are being segmented using an enhanced Gibbs Random Fields model that incorporates edge information provided by the high frequency residuals. The results of the application of our algorithm are visually more attractive than the segmentation results obtained by applying both the K-means algorithm and the Adaptive Clustering Segmentation algorithm by Pappas (1992).

FSMIQ: fuzzy similarity matching for image queries

In this paper we present FSMIQ, a novel image retrieval system that is based on fuzzy similarity metrics. FSMIQ uses shape and color information to generate a 2630 byte long information vector that describes the shape and color distribution in the image. This information is generated by the application of the discrete wavelet transform to the YIQ color space and picking the appropriate information by quantization of the Y channel coefficients and using fuzzy linguistic variables for color description. The information vectors that correspond to the images in a given database are used for the queries. Queries are carried out using the generalized Tversky index, a similarity index that is based on human similarity perception, which has been developed by the authors. Different retrieval results are calculated for shape and color; a final data fusion process takes place to provide the overall results. For examining the efficiency of FSMIQ, we use the AVRR and IAVRR metrics proposed by Flickner et al. (1995). Our experiments indicate very good performance, both visually and based on the aforementioned metrics.

An Adaptive Fuzzy Clustering Algorithm for Image Segmentation

Publisher Summary This chapter presents a novel adaptive fuzzy clustering scheme for image segmentation. In the proposed method, the non-stationary nature of images is taken into account by modifying the prototype vectors as functions of the sample location in the image, and the inherent high inter-pixel correlation is modeled using neighborhood information. A multi-resolution model is utilized for estimating the spatially varying prototype vectors for different window sizes. The fuzzy segmentations at different resolutions are combined using a data fusion process to compute the final fuzzy partition matrix. The results provide segmentations, having lower fuzzy entropy when compared to the possibilistic C-Means algorithm, while maintaining the images main characteristics. In addition, because of the neighborhood model, the effects of noise in the form of single pixel regions are minimized.