Computer-aided detection of brain metastases using a three-dimensional template-based matching algorithm

The purpose of this work was to develop an algorithm for detecting brain metastases in magnetic resonance imaging (MRI), emphasizing the reduction of false positives. Firstly, three-dimensional templates were cross-correlated with the brain volume. Afterwards, each lesion candidate was segmented in...

Full description

Saved in:
Bibliographic Details
Published in2014 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society Vol. 2014; pp. 2384 - 2387
Main Authors Perez-Ramirez, Ursula, Arana, Estanislao, Moratal, David
Format Conference Proceeding Journal Article
LanguageEnglish
Published United States IEEE 01.01.2014
Subjects
Aged
Algorithms
Brain Neoplasms - diagnosis
Brain Neoplasms - secondary
Databases, Factual
Diagnosis, Computer-Assisted - methods
Female
Humans
Image segmentation
Imaging, Three-Dimensional
Level set
Magnetic resonance imaging
Magnetic Resonance Imaging - methods
Male
Sensitivity
Three-dimensional displays
Tumors
Online AccessGet full text
ISSN1094-687X
1557-170X
DOI10.1109/EMBC.2014.6944101

Cover

More Information
Summary:The purpose of this work was to develop an algorithm for detecting brain metastases in magnetic resonance imaging (MRI), emphasizing the reduction of false positives. Firstly, three-dimensional templates were cross-correlated with the brain volume. Afterwards, each lesion candidate was segmented in the three orthogonal views as a previous step to remove elongated structures such as blood vessels. In a database containing 19 patients and 62 brain metastases, detection algorithm showed a sensitivity of 93.55%. After applying the method for false positive reduction, encouraging results were obtained: false positive rate per slice decreased from 0.64 to 0.15 and only one metastasis was removed, leading to a sensitivity of 91.94%.
ISSN:1094-687X
1557-170X
DOI:10.1109/EMBC.2014.6944101