Monique Rasigni

Detection of the foveal avascular zone on retinal angiograms using Markov random fields

This paper deals with the segmentation of the Foveal Avascular Zone (FAZ) in digital retinal angiograms. Retinal angiography is used for detection and progression in some eye pathologies. The proposed method consists of two-stages: Singular Value Decomposition (SVD) and FAZ segmentation using Markov Random Fields (MRF). The obtained results demonstrate that the method is encouraging as a first approach for location and evolution of FAZ in retinal images.

Study of surface roughness using microdensitometer analysis of electron micrographs of surface replicas. II: Autocovariance functions

Autocovariance functions (ACF’s) for rough surfaces of magnesium and silver deposits are deduced from surface profiles previously determined by using microdensitometer analysis of micrographs of surface-shadowed carbon replicas. It is shown that initial portions of the ACF’s for magnesium and silver surfaces have a Gaussian form, whereas the ACF for a grating is perfectly periodic. The rms roughness height δ and the autocovariance length σ are deduced for each surface. There is a linear relation between δ and σ for rough deposits of magnesium. For this metal, values of δ and σ are compared with corresponding values of Gesell et al. [ Phys. Rev. B7, 5141 ( 1972)], who have fitted their experimental data to a theoretical expression given by Elson and Ritchie [ Phys. Lett.33A, 255 ( 1970)]. Good agreement is obtained. Finally, the ACF’s of the surface slopes are computed, and it is shown that results obtained are consistent with results deduced from ACF’s for surface profiles.

Statistical-Based Tracking Technique for Linear Structures Detection: Application to Vessel Segmentation in Medical Images

Linear structures such as blood vessels in medical images are important features for computer-aided diagnosis and follow-up of many diseases. In this letter a new tracking-based segmentation method is proposed to detect blood vessels in retinal images. Bayesian segmentation with the Maximum a posteriori (MAP) Probability criterion is used for that purpose. Tests on simulated and retinal images are presented and compared with a vessel detection technique. Our method performs better results.

Validity of surface-roughness study using microdensitometer analysis of electron micrographs of surface replicas

We reconstruct the micrograph images from surface profiles previously obtained1 by using microdensitometer analysis of electron micrographs of surface replicas. These micrograph images compare favorably with the original images.

Mathematical morphology in computerized analysis of angiograms in age-related macular degeneration.

This paper deals with image processing of numerical retinal angiograms in order to facilitate diagnosis and follow-up in age-related macular degeneration (AMD) which is currently the main cause of blindness in industrialized countries. A computerized scheme using principally mathematical morphology operators is proposed for detecting and counting drusen, which are precursor lesions of the ocular fundus. In order to check the feasibility of this approach, results relative to 58 retinal images are compared with those given by three retinal specialists independently. From manual counting measures it is found that interobserver correlation coefficients lie in the range 0.71-0.78. On the other hand, a correlation coefficient of 0.89 is obtained when the average of the three expert countings is compared with the drusen number given by the computerized method. This coefficient is improved from 0.89 to 0.93 by processing only frames captured immediately after the appearance of a dye consecutive to intravenous sodium fluorescein injection. Compared to the manual analysis which is, among other things, tedious and time consuming, the computerized analysis is both quicker and more objective. Validation by the practitioner is however necessary, given possible detection errors. The proposed computerized scheme for detecting and counting drusen may be easily automated and so should prove useful in clinical studies which involve high volume analysis of retinal angiograms.

Statistical segmentation of regions of interest on a mammographic image

This paper deals with segmentation of breast anatomical regions, pectoral muscle, fatty and fibroglandular regions, using a Bayesian approach. This work is a part of a computer aided diagnosis project aiming at evaluating breast cancer risk and its association with characteristics (density, texture, etc.) of regions of interest on digitized mammograms. Novelty in this paper consists in applying and adapting Markov random field for detecting breast anatomical regions on digitized mammograms whereas most of previous works were focused on masses and microcalcifications. The developed method was tested on 50 digitized mammograms of the mini-MIAS database. Computer segmentation is compared to manual one made by a radiologist. A good agreement is obtained on 68% of the mini-MIAS mammographic image database used in this study. Given obtained segmentation results, the proposed method could be considered as a satisfying first approach for segmenting regions of interest in a breast.

Filtering noise on mammographic phantom images using local contrast modification functions

This paper deals with filtering signal-dependent noise on digitized mammographic phantom images using a direct contrast modification method. First, a local contrast is computed for each pixel depending on the statistical properties of its neighbourhood. An optimal modification contrast function is then applied. This function is found by solving an optimisation problem using the mean squared error as a criterion. At last the enhanced pixel value is calculated using an inverse local contrast method. Simulated images containing objects similar to those observed in the phantom are built with different contrast and Signal to Noise Ratio (SNR) levels. Noise reduction results obtained are then compared to those of classical noise filtering methods. This comparison shows that the developed method gives better results. Evaluation was also done on real phantom images with the help of radiologists. Good results obtained lead us to consider the developed method as a good preprocessing step for quality control in mammographic facilities using image processing techniques.

Statistical-based linear vessel structure detection in medical images

Linear structures such as blood vessels in medical images are important features for computer-aided diagnosis and follow-up of many diseases. In this letter a new tracking-based segmentation method is proposed to detect blood vessels in retinal angiorams. Bayesian segmentation with the Maximum a posteriori (MAP) Probability criterion is used for that purpose. First promising results are presented and discussed.

ENHANCEMENT OF MAMMOGRAPHIC PHANTOM FEATURES BY NOISE REDUCTION

A noise reduction scheme on digitized mammographic phantom images is presented. This algorithm is based on a direct contrast modification method with an optimal function, obtained by using the mean squared error as a criterion. Computer simulated images containing objects similar to those observed in the phantom are built to evaluate the performance of the algorithm. Noise reduction results obtained on both simulated and real phantom images show that the developed method may be considered as a good preprocessing step from the point of view of automating phantom film evaluation by means of image processing.

Feasibility study of automating breast phantom scoring using image processing techniques

Quality control in mammographic facilities has to be done periodically in order to ensure that the mammographic chain works properly. In particular global image quality is evaluated from a mammographic phantom film. A phantom is an object with the same anatomic shape and radiological response as an average dense-fleshed breast in which are embedded structures that mimic clinically relevant features such as microcalcifications, masses and fibers. This evaluation is done by visual observation of a phantom film and a global score is given depending on the number of objects seen by several observers. This paper presents the main results of a feasibility study of breast phantom scoring using digital image processing. First breast phantom films were digitized. For each category of structures, subimages were extracted from the digitized phantom. A noise reduction method was used as a pre-processing step. A local contrast enhancement was then applied. At last image segmentation was done. Noise reduction and contrast enhancement steps were both based on a direct contrast modification technique. Segmentation step was adapted to each embedded object. Nine digitized phantom films were studied and results show that an evaluation of mammographic facilities could be done using digital image processing.