Breast cancer is one of the cancers that has a higher mortality rate among women and early detection increases the possibilities of cure, so its early detection is one of the best treatments for this serious disease. Microcalcifications are a type of lesion in the breast and its presence is highly correlated with the presence of cancer. In this paper we present a method for automatic detection of microcalcifications using digital image processing using a Gaussian filtering approach, which can enhance the contrast between microcalcifications and normal tissue present in a mammography, then apply a local thresholding algorithm witch allow the identification of suspicious microcalcifications. The classifier used to determine the degree of benign or malignant microcalcifications is the K-Nearest Neighbours (KNN) and the validation of the results was done using ROC curves.
Tópico:
AI in cancer detection
Citaciones:
0
Citaciones por año:
No hay datos de citaciones disponibles
Altmétricas:
No hay DOI disponible para mostrar altmétricas
Información de la Fuente:
FuenteDOAJ (DOAJ: Directory of Open Access Journals)