Texture analysis can be used to improve skin sonography quantification capacities
Miniatura indisponível
Ficheiros
Data
2018
Título da revista
ISSN da revista
Título do Volume
Editora
Edições Universitárias Lusófonas
Resumo
A interpretação de imagens de sonogramas é um processo difícil, pois depende muito da experiência do observador e é limitado pela resolução de imagem dos scanners comerciais. A análise de textura oferece uma variedade de ferramentas de análise que avaliam a variação espacial e a distribuição dos níveis de cinza dentro de uma determinada imagem e as traduz em parâmetros quantitativos. Comparamos pele não tratada, hidratada e com eritema. Os parâmetros de análise de textura (entropia, contraste, correlação, energia, homogeneidade) foram calculados a partir das regiões de interesse selecionadas da epiderme, da derme e hipoderme. Entropia, contraste e homogeneidade foram maiores na epiderme mais ecogénica, enquanto correlação e energia foram maiores para a derme mais homogénea. O eco, entropia, contraste e correlação foram consistentemente menores em ambas as camadas da pele hidratada e eritema, enquanto a energia foi maior. Estes resultados sugerem que a análise de textura é útil para descrever quantitativamente as diferenças nas camadas da pele com base no seu perfil ecogénico.
Image interpretation of sonograms is a difficult process since it largely depends on the observer’s experience and is the limited by image resolution of commercial scanners. Texture analysis offers a range of analysis tools that assess the spatial variation and distribution of gray levels within a given image and translates them into quantitative parameters. We compared non-treated skin, hydrated, and erythema. Texture analysis parameters (entropy, contrast, correlation, energy, homogeneity) were calculated from the epidermis, dermis, and hypodermis selected regions of interest. Entropy, contrast and homogeneity were higher in the more echogenic epidermis, while correlation and energy were higher for the more homogenous dermis. The echo, entropy, contrast and correlation were found to be consistently lower on both layers of the hydrated skin and erythema, while energy was higher. These results suggest that texture analysis is useful to quantitatively describe the differences in skin layers based on their echogenic profile.
Image interpretation of sonograms is a difficult process since it largely depends on the observer’s experience and is the limited by image resolution of commercial scanners. Texture analysis offers a range of analysis tools that assess the spatial variation and distribution of gray levels within a given image and translates them into quantitative parameters. We compared non-treated skin, hydrated, and erythema. Texture analysis parameters (entropy, contrast, correlation, energy, homogeneity) were calculated from the epidermis, dermis, and hypodermis selected regions of interest. Entropy, contrast and homogeneity were higher in the more echogenic epidermis, while correlation and energy were higher for the more homogenous dermis. The echo, entropy, contrast and correlation were found to be consistently lower on both layers of the hydrated skin and erythema, while energy was higher. These results suggest that texture analysis is useful to quantitatively describe the differences in skin layers based on their echogenic profile.
Descrição
Biomedical and biopharmaceutical research : jornal de investigação biomédica e biofarmacêutica
Palavras-chave
MEDICINA, ULTRASSONOGRAFIA, PELE, HIDRATAÇÃO, ERITEMA, MEDICINE, ULTRASONOGRAPHY, SKIN, HYDRATION, ERYTHEMA
Citação
Macedo , A , Silva , H & Rodrigues , L M 2018 , ' Texture analysis can be used to improve skin sonography quantification capacities ' , Biomedical and biopharmaceutical research : jornal de investigação biomédica e biofarmacêutica . https://doi.org/10.19277/BBR.15.2.187