RESUMO
Early diagnosis of potentially malignant disorders, such as oral epithelial dysplasia, is the most reliable way to prevent oral cancer. Computational algorithms have been used as an auxiliary tool to aid specialists in this process. Usually, experiments are performed on private data, making it difficult to reproduce the results. There are several public datasets of histological images, but studies focused on oral dysplasia images use inaccessible datasets. This prevents the improvement of algorithms aimed at this lesion. This study introduces an annotated public dataset of oral epithelial dysplasia tissue images. The dataset includes 456 images acquired from 30 mouse tongues. The images were categorized among the lesion grades, with nuclear structures manually marked by a trained specialist and validated by a pathologist. Also, experiments were carried out in order to illustrate the potential of the proposed dataset in classification and segmentation processes commonly explored in the literature. Convolutional neural network (CNN) models for semantic and instance segmentation were employed on the images, which were pre-processed with stain normalization methods. Then, the segmented and non-segmented images were classified with CNN architectures and machine learning algorithms. The data obtained through these processes is available in the dataset. The segmentation stage showed the F1-score value of 0.83, obtained with the U-Net model using the ResNet-50 as a backbone. At the classification stage, the most expressive result was achieved with the Random Forest method, with an accuracy value of 94.22%. The results show that the segmentation contributed to the classification results, but studies are needed for the improvement of these stages of automated diagnosis. The original, gold standard, normalized, and segmented images are publicly available and may be used for the improvement of clinical applications of CAD methods on oral epithelial dysplasia tissue images.
Assuntos
Redes Neurais de Computação , Camundongos , Animais , Aprendizado de Máquina , Algoritmos , Neoplasias Bucais/diagnóstico por imagem , Neoplasias Bucais/patologia , Processamento de Imagem Assistida por Computador/métodos , Bases de Dados Factuais , Lesões Pré-Cancerosas/diagnóstico por imagem , Lesões Pré-Cancerosas/patologia , Língua/patologia , Língua/diagnóstico por imagem , Humanos , Mucosa Bucal/patologia , Mucosa Bucal/diagnóstico por imagemRESUMO
Multiple sequence alignment (MSA) is one of the most important tasks in bioinformatics and it can be used to prediction of structures or functions of unknown proteins and to phylogenetic tree reconstruction. There are many heuristics to perform multiple sequence alignment, as Progressive Alignment, Ant Colony, Genetic Algorithms, among others. Along the years, some tools were proposed to perform MSA and MSA-GA is one of them. The MSA-GA is a tool based on Genetic Algorithm to perform multiple sequence alignment and its results are generally better than other well-known tools in bioinformatics, as Clustal W. The COFFEE objective function was implemented in the MSA-GA in order to allow it to produce better alignments to less similar sequence sets of proteins. Nonetheless, the COFFEE objective function is not suited do perform multiple sequence alignment of nucleotides. Thus, we have modified the COFFEE objective function, previously implemented in the MSA-GA, to allow it to obtain better results also to sequences of nucleotides. Our results have shown that our approach has achieved better results in all cases when compared with standard COFFEE and most of cases when compared with WSP for all test cases from BAliBase and BRAliBase. Moreover, our results are more reliable because their standard deviations have less variation.
Assuntos
Algoritmos , DNA/genética , Proteínas/genética , Alinhamento de SequênciaRESUMO
INTRODUCTION: The term "Fractal" is derived from the Latin fractus meaning "irregular" or "broken" considering the observed structure with a non-integer dimension. There are many studies which employed the Fractal Dimension (FD) as a diagnostic tool. One of the most common methods for its study is the "Box Counting Method". OBJECTIVE: The aim of the present study was to try to establish the contribution of FD in the quantification of myocardial cellular rejection after cardiac transplantation. METHODS: Microscopic digital images were captured at 800x600 resolution (magnification 100x). FD was calculated with the aid of "ImageJ software" with adaptations. The classification of the degrees of rejection was in agreement with the "International Society for Heart and Lung Transplantation" (ISHLT 2004). The final report of the degree of rejection was confirmed and redefined after an exhaustive review of the slides by an external experienced pathologist. 658 slides were evaluated with the following distribution among the degrees of rejection (R): 335 (0R); 214 (1R); 70 (2R); 39 (3R). The data were statistically analyzed with Kruskal-Wallis tests and ROC curves being considered significant values of P < 0.05. RESULTS: There was significant statistical difference between the various degrees of rejection with the exception of R3 versus R2. The same trend was observed in applying the ROC curve. CONCLUSION: FD may contribute to the assessment of myocardial cellular rejection. Higher values are directly associated with progressively higher degrees of rejection. This may help in decision making of doubtful cases and those which contemplate the intensification of immunosuppressive medication.
Assuntos
Fractais , Rejeição de Enxerto/patologia , Transplante de Coração/patologia , Biópsia , Humanos , Interpretação de Imagem Assistida por Computador , Variações Dependentes do Observador , Valor Preditivo dos Testes , Curva ROC , Sensibilidade e EspecificidadeRESUMO
INTRODUÇÃO: O termo "fractal" é derivado do latim fractus, que significa "irregular" ou "quebrado", considerando a estrutura observada como tendo uma dimensão não-inteira. Há muitos estudos que empregaram a Dimensão Fractal (DF) como uma ferramenta de diagnóstico. Um dos métodos mais comuns para o seu estudo é a "Box-plot counting" (Método de contagem de caixas). OBJETIVO: O objetivo do estudo foi tentar estabelecer a contribuição da DF na quantificação da rejeição celular miocárdica após o transplante cardíaco. MÉTODOS: Imagens microscópicas digitalizadas foram capturadas na resolução 800x600 (aumento de 100x). A DF foi calculada com auxílio do "software ImageJ", com adaptações. A classificação dos graus de rejeição foi de acordo com a "Sociedade Internacional de Transplante Cardíaco e Pulmonar" (ISHLT 2004). O relatório final do grau de rejeição foi confirmado e redefinido após exaustiva revisão das lâminas por um patologista experiente externo. No total, 658 lâminas foram avaliadas, com a seguinte distribuição entre os graus de rejeição (R): 335 (0R), 214 (1R), 70 (2R), 39 (3R). Os dados foram analisados estatisticamente com os testes Kruskal-Wallis e curvas ROC sendo considerados significantes valores de P < 0,05. RESULTADOS: Houve diferença estatística significativa entre os diferentes graus de rejeição com exceção da 3R versus 2R. A mesma tendência foi observada na aplicação da curva ROC. CONCLUSÃO: ADF pode contribuir para a avaliação da rejeição celular do miocárdio. Os valores mais elevados estiveram diretamente associados com graus progressivamente maiores de rejeição. Isso pode ajudar na tomada de decisão em casos duvidosos e naqueles que possam necessitar de intensificação da medicação imunossupressora.
INTRODUCTION: The term "Fractal" is derived from the Latin fractus meaning "irregular" or "broken" considering the observed structure with a non-integer dimension. There are many studies which employed the Fractal Dimension (FD) as a diagnostic tool. One of the most common methods for its study is the "Box Counting Method". OBJECTIVE: The aim of the present study was to try to establish the contribution of FD in the quantification of myocardial cellular rejection after cardiac transplantation. METHODS: Microscopic digital images were captured at 800x600 resolution (magnification 100x). FD was calculated with the aid of "ImageJ software" with adaptations. The classification of the degrees of rejection was in agreement with the "International Society for Heart and Lung Transplantation" (ISHLT 2004). The final report of the degree of rejection was confirmed and redefined after an exhaustive review of the slides by an external experienced pathologist. 658 slides were evaluated with the following distribution among the degrees of rejection (R): 335 (0R); 214 (1R); 70 (2R); 39 (3R). The data were statistically analyzed with Kruskal-Wallis tests and ROC curves being considered significant values of P < 0.05. RESULTS: There was significant statistical difference between the various degrees of rejection with the exception of R3 versus R2. The same trend was observed in applying the ROC curve. CONCLUSION: FD may contribute to the assessment of myocardial cellular rejection. Higher values are directly associated with progressively higher degrees of rejection. This may help in decision making of doubtful cases and those which contemplate the intensification of immunosuppressive medication.