Laboratory of Analysis and Processing of Medical and Dental Images
 
Escola de Engenharia de São Carlos - Universidade de São Paulo
 
SEL | USP  
 
  /Trabalhos/Projetos Desenvolvidos
 
 

Microcalcifications segmentation techniques in dense breast images (ref. 28)

A comparison among 3 segmentation techniques regarding their performance in detecting microcalcifications in dense breast images was performed. A fourth technique joining the best results from those 3 others was proposed. The first technique (Wallet, 1997) uses a high-boost filter for image background suppression. The second one (Nappi, 1999) uses images difference with a median filter and a local threshold in order to detect microcalcifications. And the 3 rd (Nishikawa, 1992) applies a matched filter for background image suppression as well as for small signals attenuation, besides local threshold and morphological operators. The hybrid procedure proposed has summed the results from the other 3 in an unique image in order to increase the detection rates.

 
 
 
 
 
 
 

©2004 Laboratório de Análise e Processamento de Imagens Médicas e Odontológicas.