Giovanni Venturi

Model generation and model matching of real images by a fuzzy approach

Abstract A method based on fuzzy sets for model representation and matching of real images in a knowledge-based system is presented. The detailed descriptions of the systems models, data structures, and matching mechanism, as well as the introduction to a method for the generation of symbolic models, are the main topics of the present paper. Models and data structures are based on the use of fuzzy restrictions. The process of model generation starts from a set of training images whose features are analysed to find discriminant descriptions of single objects and their mutual relationships. As an example of application, the efficiency of this approach has been tested using medical tomographic images acquired by the magnetic resonance technique. Results demonstrate the applicability of one ideal model to various real scenes of the same type. The systems performance and the errors incurred are evaluated. The robustness of the model and of the method has been proved both by processing images affected by noise and by changing segmentation threshold values in the preprocessing step.

A Fuzzy Model for the Processing and Recognition of MR Pathological Images

The methodologies for the generation of model describing two echoes of MR pathological images of the head are presented. A vocabulary set has been chosen and formalized consisting of attributes and relations for the characterization of the organs and tissues contained in the image. The analitic study of a training set of images, associated with expert aid for medical aspects has permitted the creation of the model whose robustness has been proved on a set of test images. The most important and innovative characteristics of the model are the hierarchical subdivision between organ-father and sub-organs, together with the distinction between anatomical and acquisition-dependent properties of the image.

A target-oriented adaptive segmentation method

An adaptive segmentation system for detection of simple structures in biomedical images is presented. The system adaptively selects the best segmentation algorithm on the basis of the segmentation status and of the segmentation goal, and executes different iterations. A geometric model can be specified or not. Use of a on-line evaluation of the segmentation status and the decision strategy are the most innovative characteristics of the system.

High-magnification image reconstruction from partial views

Observing specimen images at high magnification is often needed for histological analysis. Unfortunately, the higher the magnification, the more limited the field under observation. A single sample must be subdivided into several images, each representing a partial view, so that the global view and the geometrical correspondences among image components are missed. When using a conventional microscope, not equipped with a mechanical scanning mechanism, it is possible to overcome this drawback by exploited digital processing facilities. To this end, we have developed a method for high-magnification image reconstruction from partial views. Digital images of partial views from a single specimen are acquired and processed with image processing algorithms, in order to correct distortions and to eliminate overlapping. A global high-magnification image of the specimen is thus reconstructed.

Reconstruction of 3D structures and configurations from microscopical partial views

The purpose of the present article is to describe the reconstruction of 3D histological structures and of their spatial configurations by using a set of images obtained from histological serial sections acquired with a conventional microscope. A sequence of image processing algorithms is proposed and utilized for the reconstruction of a global slice from partial views, for the extraction of the contours of objects of interest, for a geometric registration of image planes, and for the final visualization of the spatial configurations of such objects.

IMAGE REGISTRATION BY A REGION-BASED APPROACH AND BY CORRECTION OF ELASTIC DEFORMATIONS**This work was supported by the CNR (Target Project on Biotechnology and Bioinstrumentation).

This paper addresses two main aspects of multisensory-image registration: rigid transformation and correction of nonlinear deformations.

Evaluation and error detection in digital image segmentation

The purpose of automatic segmentation is to extract interesting regions and contours from a digital image. Today a very large number of segmentation algorithms are available, whose efficiency is usually domain-dependent, i.e., they operate to different degrees of accuracy according to the parameters used which are tuned to specific application domains. A method for result evaluation and error detection in automatic segmentation is proposed. A mathematical and a physical description of possible errors are presented, and an algorithm for error detection is implemented. Three types of segmentation errors are analyzed: undersegmentation errors, oversegmentation errors, and boundary errors. An undersegmentation error occurs when pixels belonging to different semantic objects are grouped into a single region. Such errors are the most dang erous because they can invalidate the whole segmentation process. The oversegmentation error, on the contrary, occurs when a single semantic object is subdivided by segmentation into several regions. Small oversegmentation errors may be acceptable in many applications (especially in the medical field), as they can easily be rectified by merging object parts. A boundary error consists in a discrepancy between the boundaries of a semantic object and those of the segmented one. In real images, all these errors may often be encountered at the same time. The system implemented permits one to detect each type of error, at the pixel level, by referring to a manually segmented image obtained by an expert. It produces a report on segmentation results, for both a whole image and single regions.

Biomedical structure recognition by successive approximations

A knowledge-based system is presented that has been designed to overcome the difficulties generally encountered in the extraction of structures from biomedical images. Different segmentation methods are opportunistically applied and parameter values are automatically controlled through the use of models, data, and progressive results. Detected structures are assigned fuzzy membership values related to the reliability of recognition results. The application of the system to microscopic images is described. Peculiar features of the system include a high degree of tolerance to parameter variations high flexibility and a reduced processing time.

Model generation and performance analysis in a K-B matching process of medical images

The application of a system of the blackboard type, IBIS, devoted to the processing and interpretation of real digital images, to magnetic resonance (MR) tomographic images is discussed, and the generation of a symbolic model, together with the performance and robustness of the approach, is examined. Peculiarities are represented by the mechanism of the model generation, which starts from real cases as a training set, and by the satisfactory performance reached by the matching process (based on the fuzzy sets for uncertainty management) on test cases referring to a different set of patients. The specific application considered deals with transaxial slices that are acquired from the first and second echo of a multiecho sequence, at the superorbital level, and that show signs of a lesion.<<ETX>>

Exploitation Of Contour Information From Multi-echo MR Slices

A method for combining edge information from multi-echo images is presented. The edges are extracted separatly from the different channels and then they are added up. The result obtained in this way is processed by a thinning algorithm to eliminate rendundances and false contours and to mantain the complementary information of the original images. A parameter is calculated for each contour. It depends on intrinsic characteristics of the contour such length, direction, strength and variance. It expresses the fuzzy membership value of the contour and it is very useful to discriminate the importance of the edgel.