Catálogo de publicaciones - libros

In the diagnosis of malignant melanoma, a skin cancer, the degree of irregularity along the skin lesion border is an important diagnostic factor. This paper presents a new measure of border irregularity based on conditional entropy. The measure was tested on 98 skin lesions of which 16 were malignant melanoma. The ROC analysis showed that the measure is 70% sensitive and 84% specific in discriminating the malignant and benign lesions. These results compare favourably with other measures and indicate that conditional entropy captures some distinguishing features in the boundary of malignant lesions.

This paper proposes an automatic hepatic tumor segmentation method of a computed tomography (CT) image using composite hypotheses. The liver structure is first segmented using histogram transformation, multi-modal threshold, maximum a posteriori decision, and binary morphological filtering. Hepatic vessels are removed from the liver because hepatic vessels are not related to tumor segmentation. In order to find an optimal threshold, composite hypotheses and minimum total probability error are used. Then a hepatic tumor is segmented by using the optimal threshold value. In order to test the proposed method, 272 slices from 10 patients were selected. Experimental results show that the proposed method is very useful for diagnosis of the normal and abnormal liver.

Segmentation is a crucial task in medical image processing. Snakes or Active Contour Models (ACM) are valuable tools to segment images. However, they need a good initialization, which is usually provided manually by an expert. In order to achieve a reliable automation of prostate segmentation in ultrasound images, morphological techniques have been used in this work to automatically generate the initial snake. The accuracy of the proposed approach is verified by testing several images. The automated segmentation of the prostate can be done in the majority of the cases without user interaction.

A reliable method to segment instruments in endoscope images is required as part of an enhanced reality system for minimally invasive surgery of the spine. Numerous characteristics of these images make typical intensity or model constraints for segmentation impractical. Rather, line-structure concepts are used to exploit the high length-to-diameter ratio expected of surgical instruments. A Bayesian selection scheme is proposed, and is shown to reliably differentiate these target objects from other line-like background structures.

A new algorithm for an effective and automatic segmentation of the carotid wall in ultrasonic images is proposed. It combines the speed of thresholding algorithms with the accuracy, flexibility and robustness of a successful geometric active contour model which incorporates an optimal image segmentation model in a level set framework. Due to the multiphase nature of these images, a sequential minimum cross entropy thresholding is used to get a first approximation of the segments, reducing the problem to a two phase segmentation. This thresholding solution is then used as a starting point for a two phase piecewise constant version of a geometric active contour model to reduce noise, smooth contours, improve their position accuracy and close eventual gaps in the carotid wall.

A method is proposed to segment digital posterior-anterior chest X-ray images. The segmentation is achieved through the registration of a deformable prior model, describing the anatomical structures of interest, to the X-ray image. The deformation of the model is performed using a deformation grid. A coarse matching of the model is done using anatomical landmarks automatically extracted from the image, and maps of oriented edges are used to guide the deformation process, optimized with a probabilistic genetic algorithm. The method is applied to extract the ribcage and delineate the mediastinum and diaphragms. The segmentation is needed for defining the lungs region, used in computer-aided diagnosis systems.

X-ray images, such as mammograms, often contain high-intensity radiopaque artifacts in the form of horizontal or vertical stripes, often the result of the digitization process. These artifacts can contribute to difficulties in segmentation and enhancement algorithms. This paper presents an algorithm to suppress stripe artifacts based on weighted median filtering and shows how it affects post-processing segmentation.

Thin-Layer Chromatography images are used to detect and identify the presence of specific oligosaccharides, expressed by the existence, at different positions, of bands in the gel image. 1D gaussian deconvolution, commonly used for band detection, does not produce good results due to the large curvature observed in the bands. To overcome this uncertainty on the band position, we propose a novel feature extraction methodology that allows an accurate modeling of curved bands. The features are used to classify the data into two different classes, to differentiate normal from pathologic cases. The paper presents the developed methodology together with the analysis and discussion of the results.

Image and statistical analysis are two important aspects of microarray technology. Of these, gridding is necessary to accurately identify the location of each spot while extracting spot intensities from the microarray images and automating this procedure permits high-throughput analysis. In this paper, an automatic gridding and spot quantification technique is proposed, which takes a microarray image (or a sub-grid) as input, and makes no assumptions about the size of the spots, and number of rows and columns in the grid. The proposed method is based on a weighted energy maximization algorithm that utilizes three different energy functions. The method has been found to effectively detect the grids on microarray images drawn from databases from GEO, Stanford genomic laboratories and on some images obtained from private repositories.

In this paper a novel approach to the impulsive noise removal in color images is presented. The proposed technique employs the switching scheme based on the impulse detection mechanism using the so called concept. Compared to the vector median filter, the proposed technique consistently yields better results in suppressing both the random-valued and fixed-valued impulsive noise. The main advantage of the proposed noise detection framework is its enormous computational speed, which enables efficient filtering of large images in real-time applications. The proposed filtering scheme has been successfully applied to the denoising of the cDNA microarray images. Experimental results proved that the new filter is capable of removing efficiently the impulses present in multichannel images, while preserving their textural features.