BioWebVis: ambiente web para citomorfometria utilizando imagens 3D

• Ano de defesa: 2017
• Autor(a) principal: Silva, Eduardo Henrique
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Uberlândia; Brasil; Programa de Pós-graduação em Ciência da Computação
• Palavras-chave em Português: Citomorfometria; Software de bioimagem; Visualização científica; Cytomorphometry; Bioimaging software; Scientific visualization; Computação; Desenvolvimento de software; Análise - Software; Armazenamento de dados; CNPQ::CIENCIAS EXATAS E DA TERRA
• Link de acesso: https://repositorio.ufu.br/handle/123456789/20685; http://doi.org/10.14393/ufu.di.2017.481

Bioimaging software is used to analyze microscopic images and assist users in their decision making. Most bioimaging software is available as local tools, requiring installation and, in some cases, configuration. The present dissertation aims to present the development of bioimaging software in the cloud, called BioWebVis. BioWebVis can be accessed in a web browser using the internet. This software is intended to assist pathologists, biologists and other specialists in decision making, minimizing the subjectivity of their evaluations. In the development, studies were carried out on free bioimaging software, which identified the positive and negative characteristics, allowing the selection of appropriate technologies to develop the environment. The work provided a bioimaging software in the cloud for cytomorphometric analysis using three-dimensional images. Two case studies were carried out to validate the environment. In each case study, different techniques were used to validate the environmental functionalities. The first one used cytomorphometry to classify the location of subcellular patterns in HeLa cells. This case study provided relevant results, because using the Quadratic Discriminant Analysis (QDA) classifier with 5 morphological attributes made possible to reach an accuracy of 97.59% in the classification of subcellular patterns. The second case study used cytomorphometry to analyze changes in the heterochromatin pattern in Amazonian turtle brains. Using the K-means algorithm the V-measure was equal to 1, indicating a perfect grouping. Preliminary studies indicate that heterochromatin cells have suffered a decrease in size with larger dosages of herbicides. The proposed environment was able to provide quantitative data for cytomorphometric analysis and also aided in the discovery of cellular patterns.