Cookies
Usamos cookies para melhorar nosso site e a sua experiência. Ao continuar a navegar no site, você aceita a nossa política de cookies. Ver mais
Fechar
  • Menu
Publicações

Publicações por Ana Maria Mendonça

2015

3D lung nodule candidate detection in multiple scales

Autores
Novo, J; Goncalves, L; Mendonca, AM; Campilho, A;

Publicação
2015 14th IAPR International Conference on Machine Vision Applications (MVA)

Abstract
Lung cancer is mainly diagnosed by the identification of malignant nodules in the lung parenchyma. For that purpose, the identification of all the possible structures that could be suspicious of lung nodules became a crucial task in any lung cancer computer aided diagnosis (CAD) system. In this paper, a new approach for lung nodule candidate identification is proposed. This method uses a 3D medialness Hessian-based filtering to identify round shape structures that could be identified as nodules. This technique, that demonstrated its accuracy in lung vesselness extraction, provides clearer candidates than other approaches, providing less response in the presence of noise artifacts and returns a better continuity in vessels, mostly responsible for false positives. That way, they will be better distinguishable from the nodules in posterior analysis. This approach was validated in 120 scans from the LIDC/IDRI image database. They include 212 nodules with diameters in the range 3 mm to 30 mm. The results demonstrate that our approach is capable of identifying most of the nodules and include less false positives than other approaches, facilitating a posterior task for false positive removal.

2015

Assessment of Retinal Vascular Changes Through Arteriolar-to-Venular Ratio Calculation

Autores
Dashtbozorg, B; Mendonca, AM; Campilho, A;

Publicação
IMAGE ANALYSIS AND RECOGNITION (ICIAR 2015)

Abstract
The Arteriolar-to-Venular Ratio (AVR) is an index used for the early diagnosis of diseases such as diabetes, hypertension or cardiovascular pathologies. This paper presents three automatic approaches for the estimation of the AVR in retinal images that result from the combination of different methodologies in some of the processing phases used for AVR estimation. Each one of these methods includes vessel segmentation, vessel caliber estimation, optic disc detection or segmentation, region of interest determination, vessel classification into arteries and veins and finally AVR calculation. The values produced by the proposed methods on 40 images of the INSPIRE-AVR dataset were compared with a ground-truth obtained by two medical experts using a semi-automated system. The results showed that the measured AVRs are not statistically different from the reference, with mean errors similar to those achieved by the two experts, thus demonstrating the reliability of the herein proposed approach for AVR estimation.

2017

A Deep Neural Network for Vessel Segmentation of Scanning Laser Ophthalmoscopy Images

Autores
Meyer, MariaInes; Costa, Pedro; Galdran, Adrian; Mendonça, AnaMaria; Campilho, Aurelio;

Publicação
Image Analysis and Recognition - 14th International Conference, ICIAR 2017, Montreal, QC, Canada, July 5-7, 2017, Proceedings

Abstract
Retinal vessel segmentation is a fundamental and well-studied problem in the retinal image analysis field. The standard images in this context are color photographs acquired with standard fundus cameras. Several vessel segmentation techniques have been proposed in the literature that perform successfully on this class of images. However, for other retinal imaging modalities, blood vessel extraction has not been thoroughly explored. In this paper, we propose a vessel segmentation technique for Scanning Laser Opthalmoscopy (SLO) retinal images. Our method adapts a Deep Neural Network (DNN) architecture initially devised for segmentation of biological images (U-Net), to perform the task of vessel segmentation. The model was trained on a recent public dataset of SLO images. Results show that our approach efficiently segments the vessel network, achieving a performance that outperforms the current state-of-the-art on this particular class of images. © Springer International Publishing AG 2017.

2017

Improving convolutional neural network design via variable neighborhood search

Autores
Araujo, T; Aresta, G; Almada Lobo, B; Mendonca, AM; Campilho, A;

Publicação
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

Abstract
An unsupervised method for convolutional neural network (CNN) architecture design is proposed. The method relies on a variable neighborhood search-based approach for finding CNN architectures and hyperparameter values that improve classification performance. For this purpose, t-Distributed Stochastic Neighbor Embedding (t-SNE) is applied to effectively represent the solution space in 2D. Then, k-Means clustering divides this representation space having in account the relative distance between neighbors. The algorithm is tested in the CIFAR-10 image dataset. The obtained solution improves the CNN validation loss by over 15% and the respective accuracy by 5%. Moreover, the network shows higher predictive power and robustness, validating our method for the optimization of CNN design. © Springer International Publishing AG 2017.

2017

Objective quality assessment of retinal images based on texture features

Autores
Remeseiro, B; Mendonca, AM; Campilho, A;

Publicação
2017 International Joint Conference on Neural Networks, IJCNN 2017, Anchorage, AK, USA, May 14-19, 2017

Abstract

2017

Adversarial Synthesis of Retinal Images from Vessel Trees

Autores
Costa, Pedro; Galdran, Adrian; Meyer, MariaInes; Mendonça, AnaMaria; Campilho, Aurelio;

Publicação
Image Analysis and Recognition - 14th International Conference, ICIAR 2017, Montreal, QC, Canada, July 5-7, 2017, Proceedings

Abstract
Synthesizing images of the eye fundus is a challenging task that has been previously approached by formulating complex models of the anatomy of the eye. New images can then be generated by sampling a suitable parameter space. Here we propose a method that learns to synthesize eye fundus images directly from data. For that, we pair true eye fundus images with their respective vessel trees, by means of a vessel segmentation technique. These pairs are then used to learn a mapping from a binary vessel tree to a new retinal image. For this purpose, we use a recent image-to-image translation technique, based on the idea of adversarial learning. Experimental results show that the original and the generated images are visually different in terms of their global appearance, in spite of sharing the same vessel tree. Additionally, a quantitative quality analysis of the synthetic retinal images confirms that the produced images retain a high proportion of the true image set quality. © Springer International Publishing AG 2017.

  • 1
  • 7