A Comparative Study of Neural Computing Approaches for Semantic Segmentation of Breast Tumors on Ultrasound Images

Luis Eduardo Aguilar-Camacho, Wilfrido Gómez-Flores, Juan Humberto Sossa-Azuela

Producción científica: Capítulo del libro/informe/acta de congresoContribución a la conferenciarevisión exhaustiva

2 Citas (Scopus)

Resumen

This paper compares two approaches for semantic segmentation of breast tumors on ultrasound. The first approach, called conventional, follows the typical pattern classification process to extract hand-crafted features, followed by pixel classification with a Multilayer Perceptron (MLP) network. The second approach, called convolutional, uses a Convolutional Neural Network (CNN) to learn features automatically. For evaluating both approaches, a breast ultrasound dataset with 1200 images is considered. Experimental results reveal that the CNNs called VGG16 and ResNet50 outperformed the conventional approach in various segmentation quality indices. Thus, extracting hand-crafted discriminant features is challenging since it depends on the problem domain and the designer’s skills. On the other hand, through transfer learning, it is possible to adjust a pre-trained CNN for addressing the problem of tumor segmentation satisfactorily. This performance is because CNN learns general features in its first layers, and more subtle features are activated as depth increases.

Idioma originalInglés
Título de la publicación alojada27th Brazilian Congress on Biomedical Engineering - Proceedings of CBEB 2020
EditoresTeodiano Freire Bastos-Filho, Eliete Maria de Oliveira Caldeira, Anselmo Frizera-Neto
EditorialSpringer Science and Business Media Deutschland GmbH
Páginas1649-1657
Número de páginas9
ISBN (versión impresa)9783030706005
DOI
EstadoPublicada - 2022
Evento27th Brazilian Congress on Biomedical Engineering, CBEB 2020 - Vitória, Brasil
Duración: 26 oct. 202030 oct. 2020

Serie de la publicación

NombreIFMBE Proceedings
Volumen83
ISSN (versión impresa)1680-0737
ISSN (versión digital)1433-9277

Conferencia

Conferencia27th Brazilian Congress on Biomedical Engineering, CBEB 2020
País/TerritorioBrasil
CiudadVitória
Período26/10/2030/10/20

Huella

Profundice en los temas de investigación de 'A Comparative Study of Neural Computing Approaches for Semantic Segmentation of Breast Tumors on Ultrasound Images'. En conjunto forman una huella única.

Citar esto