Segmentation and quantification of COVID-19 infections in CT using pulmonary vessels extraction and deep learning.

At the end of 2019, the World Health Organization (WHO) reported pneumonia that started in Wuhan, China, as a global emergency problem. Researchers quickly advanced in research to try to understand this COVID-19 and sough solutions for the front-line professionals fighting this fatal disease. One of the tools to aid in the detection, diagnosis, treatment, and prevention of this disease is computed tomography (CT). CT images provide valuable information on how this new disease affects the lungs of patients. However, the analysis of these images is not trivial, especially when researchers are searching for quick solutions. Detecting and evaluating this disease can be tiring, time-consuming, and susceptible to errors. Thus, in this study, we aim to automatically segment infections caused by COVID19 and provide quantitative measures of these infections to specialists, thus serving as a support tool. We use a database of real clinical cases from Pedro Ernesto University Hospital of the State of Rio de Janeiro, Brazil. The method involves five steps: lung segmentation, segmentation and extraction of pulmonary vessels, infection segmentation, infection classification, and infection quantification. For the lung segmentation and infection segmentation tasks, we propose modifications to the traditional U-Net, including batch normalization, leaky ReLU, dropout, and residual block techniques, and name it as Residual U-Net. The proposed method yields an average Dice value of 77.1% and an average specificity of 99.76%. For quantification of infectious findings, the proposed method achieves results like that of specialists, and no measure presented a value of ρ < 0.05 in the paired t-test. The results demonstrate the potential of the proposed method as a tool to help medical professionals combat COVID-19. fight the COVID-19.

Breast cancer diagnosis from histopathological images using textural features and CBIR.

Currently, breast cancer diagnosis is an extensively researched topic. An effective method to diagnose breast cancer is to use histopathological images. However, extracting features from these images is a challenging task. Thus, we propose a method that uses phylogenetic diversity indexes to characterize images for creating a model to classify histopathological breast images into four classes - invasive carcinoma, in situ carcinoma, normal tissue, and benign lesion. The classifiers used were the most robust ones according to the existing literature: XGBoost, random forest, multilayer perceptron, and support vector machine. Moreover, we performed content-based image retrieval to confirm the classification results and suggest a ranking for sets of images that were not labeled. The results obtained were considerably robust and proved to be effective for the composition of a CADx system to help specialists at large medical centers.

Segmentação de massas em imagens mamográficas utilizando índices de similaridade / Mass segmentation in mammographic images using similarity indixes

Uma forma de segmentação de massas em imagens mamográficas é pela análise bilateral dos pares de mamografias. Sabe-se que mamografias da mama esquerda e direita apresentam alto grau de simetria e quando há uma diferença brusca entre os pares, pode-se considerar algo suspeito. OBJETIVO: Uma metodologia para segmentação de massas baseado em análise bilateral de mamografias usando técnicas de similaridade de espécies para encontrar regiões assimétricas. Materiais e MÉTODOS: Fluxo de cinco etapas: Materiais, Pré-processamento de imagens, Registro de imagens,Segmentação de regiões assimétricas e Filtragem de regiões. RESULTADOS: Os resultados preliminares mostram que essa metodologia é promissora na detecção de regiões assimétricas apresentando 95% de acerto na etapa de segmentação e 90,8% após a filtragem de regiões. CONCLUSÃO: Os índices de similaridade mostram-se promissores na tarefa de encontrar regiões suspeitas em pares de mamografias, além de formalização de técnicas para filtragem de regiões que não são massas.

One way of segmenting masses in mammographic images is the bilateral analysis of mammograms pairs. It isknown that mammograms of the left and right breast, has a high degree of symmetry and when there is an abrupt difference between the pairs may be considered something suspicious. OBJECTIVE: A methodology to segmentation mass based on bilateralanalysis of mammograms using species similarity techniques to find asymmetric regions. Materials and METHODS: A five-step flow: Materials, Pre-processing, Image Registration, Segmentation of asymmetric regions and Filtering regions. RESULTS: Preliminary results show that this method is promising in detecting asymmetrical regions showing 95%accuracy in segmentation step and 90.8% after filtering regions. CONCLUSION: The similarity indices show promise in the task of finding suspicious areas in mammograms pairs, there is also formalization of techniques to filtering regions.

Análise temporal de lesões em mamografias utilizando índices de similaridade / Temporal analysis of mammogram lesions using similarity indexes

Uma forma de verificar a malignidade de lesões em mamografias é o acompanhamento periódico, analisando mudanças em medições de geometria (forma) e textura (tecido). Uma das medidas de forma mais utilizadas é a taxa de crescimento. No entanto, somada a medidas de tecido, obtém-se informações úteis sobre o desenvolvimento interno da lesão. OBJETIVOS: Uma metodologia para estabelecer uma correspondência entre lesões em mamografias de tempos diferentes e analisar as mudanças no tecido através de índices de similaridade. MÉTODOS: Executado em cinco etapas: Aquisição das Imagens, Pré-processamento, Registro de Imagens, Correspondência entre as Lesões e Análise Temporal de Texturas. RESULTADOS: Os resultados preliminares mostram que essa metodologia é promissora na detecção de mudanças no tecido das lesões. CONCLUSÃO: Os índices de similaridade se mostraram eficientes na quantificação de mudanças na textura e podem ser usados como informações para auxiliar o acompanhamento e diagnóstico de doenças associadas as lesões.

One way to verify the malignancy of breast lesions is the temporal analysis measurement geometry (shape) and texture (tissue). In this sense, one of the most used form measures is the growth rate. However, in addition to tissue measurements over time, you get useful information about their behavior. OBJECTIVES: A methodology for establishing a correspondence between injuries at different times and analyze changes in tissue through similarity indices. METHODS: Executed in five steps: image acquisition, preprocessing, Image Registration, Correspondence between Lesions and Temporal Analysis of Lesions Texture. RESULTS: Preliminary results show that this method is promising for detecting changes in tissue lesions. CONCLUSION: The similarity indices were effective in quantitating changes in texture and can be used as information to assist the monitoring and diagnosis of lesions associated diseases.