Ana Maria Mendonça

Segmentation of retinal blood vessels by combining the detection of centerlines and morphological reconstruction

This paper presents an automated method for the segmentation of the vascular network in retinal images. The algorithm starts with the extraction of vessel centerlines, which are used as guidelines for the subsequent vessel filling phase. For this purpose, the outputs of four directional differential operators are processed in order to select connected sets of candidate points to be further classified as centerline pixels using vessel derived features. The final segmentation is obtained using an iterative region growing method that integrates the contents of several binary images resulting from vessel width dependent morphological filters. Our approach was tested on two publicly available databases and its results are compared with recently published methods. The results demonstrate that our algorithm outperforms other solutions and approximates the average accuracy of a human observer without a significant degradation of sensitivity and specificity

An Automatic Graph-Based Approach for Artery/Vein Classification in Retinal Images

The classification of retinal vessels into artery/vein (A/V) is an important phase for automating the detection of vascular changes, and for the calculation of characteristic signs associated with several systemic diseases such as diabetes, hypertension, and other cardiovascular conditions. This paper presents an automatic approach for A/V classification based on the analysis of a graph extracted from the retinal vasculature. The proposed method classifies the entire vascular tree deciding on the type of each intersection point (graph nodes) and assigning one of two labels to each vessel segment (graph links). Final classification of a vessel segment as A/V is performed through the combination of the graph-based labeling results with a set of intensity features. The results of this proposed method are compared with manual labeling for three public databases. Accuracy values of 88.3%, 87.4%, and 89.8% are obtained for the images of the INSPIRE-AVR, DRIVE, and VICAVR databases, respectively. These results demonstrate that our method outperforms recent approaches for A/V classification.

Optic disc segmentation using the sliding band filter

BACKGROUND The optic disc (OD) centre and boundary are important landmarks in retinal images and are essential for automating the calculation of health biomarkers related with some prevalent systemic disorders, such as diabetes, hypertension, cerebrovascular and cardiovascular diseases. METHODS This paper presents an automatic approach for OD segmentation using a multiresolution sliding band filter (SBF). After the preprocessing phase, a low-resolution SBF is applied on a downsampled retinal image and the locations of maximal filter response are used for focusing the analysis on a reduced region of interest (ROI). A high-resolution SBF is applied to obtain a set of pixels associated with the maximum response of the SBF, giving a coarse estimation of the OD boundary, which is regularized using a smoothing algorithm. RESULTS Our results are compared with manually extracted boundaries from public databases (ONHSD, MESSIDOR and INSPIRE-AVR datasets) outperforming recent approaches for OD segmentation. For the ONHSD, 44% of the results are classified as Excellent, while the remaining images are distributed between the Good (47%) and Fair (9%) categories. An average overlapping area of 83%, 89% and 85% is achieved for the images in ONHSD, MESSIDOR and INSPIR-AVR datasets, respectively, when comparing with the manually delineated OD regions. DISCUSSION The evaluation results on the images of three datasets demonstrate the better performance of the proposed method compared to recently published OD segmentation approaches and prove the independence of this method when from changes in image characteristics such as size, quality and camera field of view.

Automatic localization of the optic disc by combining vascular and intensity information

This paper describes a new methodology for automatic location of the optic disc in retinal images, based on the combination of information taken from the blood vessel network with intensity data. The distribution of vessel orientations around an image point is quantified using the new concept of entropy of vascular directions. The robustness of the method for OD localization is improved by constraining the search for maximal values of entropy to image areas with high intensities. The method was able to obtain a valid location for the optic disc in 1357 out of the 1361 images of the four datasets.

Automatic segmentation of microaneurysms in retinal angiograms of diabetic patients

In this paper a method for automatic detection of microaneurysms in digital angiograms of the eye fundus is described. These lesions of the human retina, a characteristic of the earliest phases of diabetic retinopathy, present themselves in the angiographic images as small, round, hyperfluorescent objects. The proposed method includes initial pre-processing and enhancement steps, followed by object segmentation. In the final phase, microaneurysms are validated using two new criteria based on local intensity, contrast and shape relations. The combination of these local features with global image parameters makes possible a high degree of independence from image intensity characteristics.

Fibronectin-mediated endothelialisation of chitosan porous matrices

Chitosan (Ch) porous matrices were investigated regarding their ability to be colonized by human microvascular endothelial cells (HPMEC-ST1.6R cell line) and macrovascular endothelial cells namely HUVECs. Specifically we assessed if previous incubation of Ch in a fibronectin (FN) solution was effective in promoting endothelial cell (EC) adhesion to Ch matrices with different degrees of acetylation (DAs). Upon FN physiadsorption, marked differences were found between the two DAs investigated, namely DA 4% and 15%. While cell adhesion was impaired on Ch with DA 15%, ECs were able to not only adhere to Ch with DA 4%, but also to spread and colonize the scaffolds, with retention of the EC phenotype and angiogenic potential. To explain the observed differences between the two DAs, protein adsorption studies using (125)I-FN and immunofluorescent labelling of FN cell-binding domains were carried out. In agreement with the higher cell numbers found, scaffolds with DA 4% revealed a higher number of exposed FN cell-binding domains as well as greater ability to adsorb FN and to retain and exchange adsorbed FN in the presence of competitive proteins. These findings suggest that the DA is a key parameter modulating EC adhesion to FN-coated Ch by influencing the adsorbed protein layer.

Detection of Lung Nodule Candidates in Chest Radiographs

This paper presents an automated method for the selection of a set of lung nodule candidates, which is the first stage of a computer-aided diagnosis system for the detection of pulmonary nodules. An innovative operator, called sliding band filter (SBF), is used for enhancing the lung field areas. In order to reduce the influence of the blood vessels near the mediastinum, this filtered image is multiplied by a mask that assigns to each lung field point an a prioriprobability of belonging to a nodule. The result is further processed with a watershed segmentation method that divides each lung field into a set of non-overlapping areas. Suspicious nodule locations are associated with the regions containing the highest regional maximum values. The proposed method, whose result is an ordered set of lung nodule candidate regions, was evaluated on the 247 images of the JSRT database with very promising results.

A neural network approach for the automatic detection of microaneurysms in retinal angiograms

In this paper a neural network structure is used to develop a system capable of detecting microaneurysms locations in retinal angiograms. The LVQ (learning vector quantization) neural network is used to classify the input patterns into their desired classes using competitive layers. The neurons in the competitive layers compete among each other to produce subclasses. These subclasses are then combined to produce the desired output classes. The input vector of the neural network is derived from a grid of smaller image windows. The presence of microaneurysms in these windows is detected according to a novel multi-stage training procedure that has proved to be very effective.

A new similarity criterion for retinal image registration

Registration methods invariably demand the evaluation of similarity between image areas using criteria based on the maintenance of pixel relative intensities. Because of the severe brightness changes in retinal images, intensity dependent similarity measures are often unable to produce useful registration results. The main purpose of this paper is the presentation of a new similarity evaluation criterion, prepared to overcome the problems raised by important brightness modifications in the images to register. The new criterion is based on the detection of edge point localizations, which are used to assess the correspondence of the compared areas. Some attention was also dedicated to the measure calculation process, and, as an ultimate result, a two-stage implementation was adopted. The values generated by the similarity measure determination step are later used by two registration algorithms prepared to compensate for distinct geometric misalignments between the images.<<ETX>>

Automatic localization of the optic disc in retinal images based on the entropy of vascular directions

This paper proposes an automatic method for estimating the location of the optic disc in color images of the retina. The proposed methodology is founded in a new concept, the entropy of vascular directions, which proved to be a reliable measure for assessing the convergence of vessels around an image point. To improve the robustness of the method, the search for the maximum value of entropy is restricted to image areas with high intensity. This new method was evaluated in two publicly available databases, containing both normal and pathological images, and was able to obtain a valid location for the optic disc in 115 out of the 121 images of the two datasets.