Eduard Montseny

State of the art survey on MRI brain tumor segmentation.

Brain tumor segmentation consists of separating the different tumor tissues (solid or active tumor, edema, and necrosis) from normal brain tissues: gray matter (GM), white matter (WM), and cerebrospinal fluid (CSF). In brain tumor studies, the existence of abnormal tissues may be easily detectable most of the time. However, accurate and reproducible segmentation and characterization of abnormalities are not straightforward. In the past, many researchers in the field of medical imaging and soft computing have made significant survey in the field of brain tumor segmentation. Both semiautomatic and fully automatic methods have been proposed. Clinical acceptance of segmentation techniques has depended on the simplicity of the segmentation, and the degree of user supervision. Interactive or semiautomatic methods are likely to remain dominant in practice for some time, especially in these applications where erroneous interpretations are unacceptable. This article presents an overview of the most relevant brain tumor segmentation methods, conducted after the acquisition of the image. Given the advantages of magnetic resonance imaging over other diagnostic imaging, this survey is focused on MRI brain tumor segmentation. Semiautomatic and fully automatic techniques are emphasized.

Fuzzy Texture Unit and Fuzzy Texture Spectrum for texture characterization

An important problem that is necessary to face up to in texture characterization is that while the notion of texture is easy to grasp intuitively, it is extremely difficult to quantify. The Texture Spectrum methodology introduced by He and Wang [1990, Texture unit, texture spectrum and texture analysis, IEEE Trans. on Geoscience and Remote Sensing, 28(4) (1990) 509-512] was initially used as a texture filtering approach and has been applied to texture characterization and texture classification showing promising discrimination performance. In this paper we propose a texture characterization approach, based on the Texture Spectrum methodology that, for taking into account the vagueness introduced by noise and the different caption and digitation processes, makes use of fuzzy techniques for defining the Texture Unit Boxes.

On the Reliability Degree of Hue and Saturation Values of a Pixel for Color Image Classification

Variability of hue and saturation components is an important drawback for developing accurate color image segmentation algorithms based on HSI spaces: different colors can return similar color components because of the different illumination levels, and the components of a color change accordingly to the illumination level. To avoid these problems within color segmentation algorithms, we propose a method that, based on predicted hue and saturation deviations, provides the reliability degree of the H-S color components based obtained HIS values. The reliability degree has been used for improving color classification results

A new fuzzy approach to brain tumor segmentation

In this paper we present a fully automatic and unsupervised brain tumor segmentation method which considers human knowledge. The expert knowledge and the features derived from the MR images are coupled to define heuristic rules aimed to the design of the fuzzy approach. To assess the unsupervised and fully automatic segmentation, intensity-based objective measures are defined, and a new method for obtaining membership functions to suit the MRI data is introduced. The proposed approach is quantitatively comparable to the most accurate existing methods, even though the segmentation is done in 2D.

FUZZY-BASED ANALYSIS OF MICROSCOPIC COLOR CERVICAL PAP SMEAR IMAGES: NUCLEI DETECTION

The Pap Smear test, or cervico-vaginal cytology, is globally the most used and suitable method for screening cervical cancer precursor lesions, with a significant impact in reducing the incidence and mortality rates. However, Pap test suffers from subjective variability and no specificity, being the most controversial point the persistence of false negatives; i.e., normal cytological report for a woman with existing dysplasia, pre-malignant, or malignant lesions of the cervix. This is due in large part to the vast number of cells that must be reviewed by a technician for determining the possible existence of a small number of malignant or pre-malignant cells. Automated systems that include technician knowledge and interpretation could not only reduce sample examination time but also avoid misclassification of samples because of human errors. Here we present part of our ongoing work toward automation of cervical screening process. Specifically, since in cytological studies nuclei are considered the most informative regions, and an accurate segmentation is needed for extracting meaningful cell features, we propose an automated nuclei detection algorithm that integrates color information, cytopathologists knowledge, and fuzzy systems. Results have shown that besides a high performance and efficiency, the speed of the algorithm is very high.

A step edge detector algorithm based on symbolic analysis

This paper analyzes the noise error, in both the module and the argument in the calculation of the gradient vector of the images illumination function. The results show that the behavior of the argument is more robust than that of the module. A proposal is made for a symbolic analysis of the argument of the gradient vector to detect the contour of the objects in the image. This shows that it is possible to use smaller windows to calculate the gradient vector without infringing the contradiction proposed by Marr and Hildreth (1980) and Canny (1983). Finally, a description is given of an edge detection algorithm where their most important characteristics are: a) it introduces a symbolic analysis of the argument of the gradient vector to detect edges; and b) It uses smaller window to approximate the value of the gradient vector allowing to locate the edge with precision.

THREECOND: An Automated and Unsupervised Three Colour Fuzzy-Based Algorithm for Detecting Nuclei in Cervical Pap Smear Images

Visual examination and interpretation of microscopic images taken from the cervix are at the core for the detection and prevention of cervical cancer. However these visual processes are tedious and in many cases error-prone. This is why automated screening systems, interacting with the technologist, would be a tremendous improvement for reducing the likelihood of human errors. In this work we propose THREECOND, a three colour-based algorithm that integrates colour information, cyto-pathologists knowledge and fuzzy systems. This algorithm is designed to be integrated into the previously developed system [23], with the aim of improving its accuracy and efficiency for detecting and segmenting the nuclei of Pap smear images.

On quality assessment of corneal endothelium and its possibility to be used for surgical corneal transplantation

Transplantation of corneal tissue is a usual practice in hospitals. The analysis of microscopy images of donor corneal endothelium is routinely carried out at eye banks for the clinical assessment of cornea quality and suitability for transplantation. One of the main clinical parameters expressing the health of a cornea, and assessing their suitability as a human graft, is the cell density of its endothelium. Endothelium cell density is conventionally estimated by a long, tedious and error-prone manual counting procedure, carried out by experts who, according to a protocol, observe specimen images through an optical microscope. Besides a great subjectivity, this manual process causes a great disparity in the results derived from the protocol considered. Another additional and very important drawback is that images are often blurred and noisy, what makes very difficult the correct recognition of the cells. Taking into account aforementioned problems, this paper introduces a computer intelligence-based system for automatic segmentation of corneal endothelium images that in addition to facilitating technicians work of will reduce the disparity of results

Robustness and Performance Evaluation of the Fuzzy Texture Spectrum encoding

In a previous work we introduced the fuzzy texture spectrum (FTS) as a texture spectrum (TS)-based method wherein, for capturing the ambiguity of texture definitions and texture features variability of real images, unlike the original method the concepts of equal, greater, and smaller grey-levels were considered as fuzzy sets. In this paper both methods are analyzed and compared for robustness evaluation. To compare the FTS and the TS for texture characterization an entropy measure and a similarity-discriminatory criterion are considered for performance and robustness evaluation. Moreover, to make the fuzzy texture spectrum rotations invariant, texture units are grouped into classes.

A Fuzzy-based Automated Cells Detection System for Color Pap Smear Tests –-FACSDS–

There is a compelling need for automated cervical smear screening systems to improve the quality and cost/efficiency screening rate. Computer-assisted devices can reduce false negative Pap smear interpretations using computerized systems to assist the cytotechnologist in identifying Pap smear abnormalities and providing added value in their ability to consistently and objectively analyze all cells on slides without fatigue. However, automation of the process is a challenging problem due to the large variability in conventional Pap smears exhibiting no standard appearance and tremendous amount of data to be processed. Moreover, smear diagnostic may be obscured by benign conditions, overlapping cells, debris, inflammation, and no uniform staining.