Cookies
O website necessita de alguns cookies e outros recursos semelhantes para funcionar. Caso o permita, o INESC TEC irá utilizar cookies para recolher dados sobre as suas visitas, contribuindo, assim, para estatísticas agregadas que permitem melhorar o nosso serviço. Ver mais
Aceitar Rejeitar
  • Menu
Publicações

Publicações por André Marçal

2015

A quantitative hybridization approach using 17 DNA markers for identification and clustering analysis of Ralstonia solanacearum

Autores
Albuquerque, P; Marcal, ARS; Caridade, C; Costa, R; Mendes, MV; Tavares, F;

Publicação
PLANT PATHOLOGY

Abstract
Ralstonia solanacearum (Rs) is a quarantine phytopathogenic bacterium accountable for heavy economic losses worldwide. Monitoring and eradication programmes required for this pathogen are dependent on the availability of time-and cost-efficient detection and typing methods. However, members of the Rs species complex are characterized by a high phenotypic and genetic diversity, which requires improved diagnostics methods. The currently available full genome sequences of several Rs strains allow for the selection of novel specific DNA markers using comparative genomics tools. In this work, 17 novel markers were selected based on Rs-specific protein domains and thoroughly validated for specificity and stability, both in silico and using 'wet lab' assays. Polymerase chain reaction-and hybridization-based validation assays revealed that the DNA regions selected as markers were unevenly distributed amongst the tested strains, with nine markers present throughout the species complex. The distribution of the remaining eight markers was highly variable between the different analysed strains and enabled the attainment of strain-specific dot blot hybridization patterns, particularly informative for typing. The average probability value of each strain being positive for each of the 17 markers was calculated by an algorithm and used to obtain a dendrogram representing hierarchical clustering analysis of Rs, according to the similarity of their hybridization patterns. This method should prove to be a robust and straightforward procedure for genotyping members of the Rs species complex. Furthermore, this quantitative hybridization approach will allow the construction of informative databases to determine new Rs genotypes and infer epidemiological patterns.

2013

Hierarchic Image Classification Visualization

Autores
Mesquita, TA; Marcal, ARS;

Publicação
IMAGE ANALYSIS AND RECOGNITION

Abstract
Image classification techniques are often used to reduce the large data volume content of an image to a simplified version - a thematic map, which can be more suitable from the user's point of view. However, the delimitation of specific regions using unsupervised classification techniques frequently generates an excessive number of clusters or classes. The resulting image can be simplified by a process of hierarchical aggregation of the initial classes, yielding a set of classified images. This set of thematic maps can provide a powerful insight into the image content, as long as an adequate visualization strategy is used. This paper presents methodologies for the visualization of hierarchically structured classified images.

2013

Identification of potential land cover changes on a continental scale using NDVI time-series from SPOT VEGETATION

Autores
Rodrigues, A; Marcal, ARS; Cunha, M;

Publicação
INTERNATIONAL JOURNAL OF REMOTE SENSING

Abstract
The identification of land cover changes on a continental scale is a laborious and time-consuming process. A new methodology is proposed based exclusively on SPOT VGT data, illustrated for the African Continent using GLC2000 as reference to select 26 distinct land cover types (classes). For each class, the normalized difference vegetation index (NDVI) time-series are extracted from SPOT VGT images and a hierarchical aggregation is done using two different methods: one that preserves the initial signatures throughout the hierarchical process, and another that recalculates the signatures for each aggregation level. The average classification agreement was above 89% using 26 classes. Reducing the number of classes improves classification agreement. In order to study the influence of temporal variability in the classification results, the methodology was applied on data from 1999, 2001, 2008, and 2010. With 26 classes, the best average classification agreement obtained was 94.5% with annual data, against 74.1% with interannual data.

2015

Automatic Analysis of Dot Blot Images

Autores
Caridade, CMR; Marcal, ARS; Albuquerque, P; Mendes, MV; Tavares, F;

Publicação
INTELLIGENT AUTOMATION AND SOFT COMPUTING

Abstract
This paper presents a method for the automatic analysis of macroarray (dot blot) images. The system developed receives as input a dot blot image, corrects it for grid rotation, identifies the visible markers and provides an evaluation of the status of each marker (ON/OFF). Two experiments were carried out to evaluate the detection and classification stages. A total of 222 test images were produced from 6 original dot blot images, with various rotations, translations, contrast and noise level. Over 7500 markers were identified automatically and compared to manual reference. The RMS error in positioning the molecular marker center was between 1.1 and 3.8 pixels and the marker radius error less than 4%. The automatic classification of markers (ON/OFF) was compared to the classification by 3 human experts, using 10 test images. The overall accuracy evaluated on 5118 markers was 94.0%. For those markers that had the same evaluation by all 3 experts, the classification accuracies were 96.6% (ON) and 95.9% (OFF).

2013

Evaluation of Features for Leaf Discrimination

Autores
Silva, PFB; Marcal, ARS; Almeida da Silva, RMA;

Publicação
IMAGE ANALYSIS AND RECOGNITION

Abstract
A number of shape features for automatic plant recognition based on digital image processing have been proposed by Pauwels et al. in 2009. A database with 15 classes and 171 leaf samples was considered for the evaluation of these measures using linear discriminant analysis and hierarchical clustering. The results obtained match the human visual shape perception with an overall accuracy of 87%.

2013

Colour-based dermoscopy classification of cutaneous lesions: An alternative approach

Autores
Silva, CSP; Marcal, ARS;

Publicação
Computer Methods in Biomechanics and Biomedical Engineering: Imaging and Visualization

Abstract
Dermoscopy (dermatoscopy or epiluminescence microscopy) is a non-invasive diagnostic technique for the in vivo observation of pigmented skin lesions used in dermatology. There is currently a great interest in the prospects of automatic image analysis methods for dermoscopy, both to provide quantitative information about a lesion, which can be of relevance for the dermatologist, and as a stand-alone early warning tool. The standard approach in automatic dermoscopic image analysis has usually three stages: (i) segmentation, (ii) feature extraction and selection and (iii) lesion classification. This study evaluates the potential of an alternative approach based on the Menzies method - presence of one or more of six colour classes, indicating that the lesion should be considered a potential melanoma. This method does not require stages (i) and (ii) - lesion segmentation and feature extraction. The identification of colour classes in dermoscopic images is a subjective task, which poses great challenges for an automatic implementation. The purpose of this work is to evaluate the potential discrimination between the six Menzies colour classes in dermoscopic red, blue and green (RGB) images. The Jeffries-Matusita and transformed divergence separability distances were used to evaluate the colour class separability for an experimental evaluation with 28 dermoscopic images. Considering the skin as an additional class, an image intensity calibration was applied to the data-set, which improved the rate of separable colour class pairs. A nonlinear cluster transformation allowed almost the total separation of each colour class in the feature space. Several neural networks in competition were used as classifiers, which lead to loss of arbitrariness and perfect knowledge of each cluster surface. The discrimination between the various Menzies colour classes in dermoscopic RGB images achieved 93% of sensibility, 62% of specificity and 74% of accuracy (averaged measures). These results indicate that it might be possible to evaluate a lesion based on the presence of Menzies colours in dermoscopic images, mimicking the human diagnosis. © 2013 Copyright Taylor and Francis Group, LLC.

  • 2
  • 15