Visualización de vasos sanguíneos en imágenes de contraste basado en la descomposición de la transformada wavelet discreta.

The visualization of blood vessels is a fundamental task in the assessment of the state and health of biological tissues. Besides, blood vessel visualization is related to changes in the relative velocity of blood ow and has been used for diagnostic purposes in biomedical applications. The visualiza...

Descripción completa

Detalles Bibliográficos
Autor: Francisco Javier López Tiro
Tipo de recurso: tesis de maestría
Estado:Versión aceptada para publicación
Fecha de publicación:2019
País:México
Institución:Instituto Nacional de Astrofísica, Óptica y Electrónica
Repositorio:Repositorio Institucional del INAOE
Idioma:español
OAI Identifier:oai:inaoe.repositorioinstitucional.mx:1009/2014
Acceso en línea:http://inaoe.repositorioinstitucional.mx/jspui/handle/1009/2014
Access Level:acceso abierto
Palabra clave:info:eu-repo/classification/Inspec/Visualización
info:eu-repo/classification/Inspec/Contraste
info:eu-repo/classification/Inspec/Wavelet
info:eu-repo/classification/Inspec/Vasos sanguíneos
info:eu-repo/classification/cti/1
info:eu-repo/classification/cti/22
info:eu-repo/classification/cti/2209
info:eu-repo/classification/cti/220990
Descripción
Sumario:The visualization of blood vessels is a fundamental task in the assessment of the state and health of biological tissues. Besides, blood vessel visualization is related to changes in the relative velocity of blood ow and has been used for diagnostic purposes in biomedical applications. The visualization of blood vessels in speckle images has been approached from different approaches, with contrast representation being the most used to visualize blood vessels. However, the contrast representation has certain limitations due to the noise that increases with the depth of the blood vessel. The proposed model improves the visualization of blood vessels in-vitro in contrast images based on the decomposition of Transformed Wavelet Discrete as a noise reduction method. The main reason for using a noise reduction approach based on the information of the approximation and detail coefficients during the decomposition of the Discrete Wavelet Transform is the need for a model that allows high and low-frequency noise to be selectively separated in the region corresponding to the blood vessels. First, the contrast images are processed with the Discrete Wavelet Transform and a noise-based criterion, obtaining a low-frequency contrast image. The low-frequency contrast images are then processed using a morphological approach to obtain the region corresponding to the blood vessels. Finally, the contrast is recalculated using a low-frequency contrast map to improve the visualization of the blood vessels. A set of 112 contrast images in-vitro was used to evaluate the model's performance. The set contains straight and bifurcated blood vessels labeled and distributed in eight depth levels up to 1000μm. The proposed model improves the visualization of superficial blood vessels (up to 500μm depth), and allows the blood vessel region to be located with an average value of 55;43 ± 6;6% compared to the reference labels on the input set.