Processamento de imagens digitais para identificação e contagem de núcleos nas fases da mitose em lâminas de teste de Allium cepa

• Ano de defesa: 2018
• Autor(a) principal: Rattmann, Amilton Carlos
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Positivo; Brasil; Pós-Graduação; Programa de Pós-Graduação em Gestão Ambiental; UP
• Palavras-chave em Português: Gestão ambiental; Algoritmo computacional; Morfologia - Matemática; Processamento de imagens digitais; CNPQ::CIENCIAS BIOLOGICAS; CNPQ::ENGENHARIAS::ENGENHARIA SANITARIA
• Link de acesso: https://repositorio.cruzeirodosul.edu.br/handle/123456789/2512

Laboratory tests for environmental research and monitoring have employed both plant and animal organisms in assessing the effects of pollutants in the environment. The use of Allium cepa has been well documented by the scientific community in recent decades for the evaluation of the effects of substances on changes in cellular structure and genetic alterations, due to the sensitivity and reliability of this organism in the assays. The purpose of this work was to develop an algorithm for the processing of the digital images of A. cepa slides captured by high resolution camera coupled to laboratory microscope, performing the identification and counting of cells at each stage of the mitotic cycle, without proposing changes to the method currently used in the preparation of the slides. The preprocessed images from space and spatial frequency domain filters were subject to global thresholding to separate the nucleus and cytoplasm regions. The decision weighted on a statistical basis of the morphological properties of the cellular nuclei was used for the classification, allowing counting of nuclear phases and determination of the mitotic index per slide. The images of each slide were processed for about 60 seconds, identifying an average of 95% of the nuclei. and correctly classified on average 63.8% of the nuclei, showing no significant differences in mitotic indexes, at the 5% level, between manual counting and automatic counting. In addition to the immediate results, the technique is expected to promote test standardization, greater efficiency in the use of human resources and equipment, reduction of the time to obtain the results, lower operational cost in the process, and greater reliability in the test results.