Abordagem de visão computacional para contagem de unidades formadoras de colônias de micro-organismos

• Ano de defesa: 2019
• Autor(a) principal: Boukouvalas, Dimitria Theophanis
• Orientador(a): Araújo, Sidnei Alves de
• Banca de defesa: Araújo, Sidnei Alves de , Vieira, Tiago Figueiredo , Prates, Renato Araújo , Deana, Alessandro Melo
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Nove de Julho
• Programa de Pós-Graduação: Programa de Pós-Graduação em Informática e Gestão do Conhecimento
• Departamento: Informática
• País: Brasil
• Palavras-chave em Português: contagem de colônias; diluição serial; granulometria por correlação cruzada; descritores de forma baseados em região; segmentação
• Palavras-chave em Inglês: colony count; serial dilution; cross-correlation granulometry; region-based shape descriptors; segmentation
• Área do conhecimento CNPq: CIENCIA DA COMPUTACAO::SISTEMAS DE COMPUTACAO
• Link de acesso: http://bibliotecatede.uninove.br/handle/tede/3041

Quantification of colony forming units (CFU) on microbial cultures prepared according to the standard spread plate technique is a daily laboratory routine that requires a large amount of material, human and time resources, as it is performed manually in Petri dishes containing a single dilution of the sample. On the other hand, SP-SDS (Single Plate Serial Dilution Spotting) is a widely used technique that allows great reduction in the use of material resources and time, since it uses several dilutions of a sample in the same Petri dish. There are different approaches in the literature for automation of CFU quantification that are based on images of standard spread plate Petri dishes with low variation of CFU features and captured under controlled lighting conditions, which is not the actual situation of laboratories. In view of this, in this study, a computational vision approach was proposed for automated quantification of CFU in Petri dishes prepared by the SP-SDS technique and acquired under real laboratory conditions, which is composed of three steps: extraction of regions of interest (Petri dish dilutions), counting of isolated CFU by region-based shape descriptors, and counting of CFU in agglomerates by cross-correlation granulometry. In the experiments carried out, it was verified that the proposed approach allows the quantification of CFU with accuracy of 99.5%, precision of 99.7% sensitivity of 99.2% and specificity of 98.3%. These results were superior to the results of the Granul and OpenCFU approaches, with which the proposed approach was compared. In addition, a user-friendly software for the quantification of CFU was conceived from the proposed approach.