Premium
Combined use of T 2 ‐weighted MRI and T 1 ‐weighted dynamic contrast–enhanced MRI in the automated analysis of breast lesions
Author(s) -
Bhooshan Neha,
Giger Maryellen,
Lan Li,
Li Hui,
Marquez Angelica,
Shimauchi Akiko,
Newstead Gillian M.
Publication year - 2011
Publication title -
magnetic resonance in medicine
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.696
H-Index - 225
eISSN - 1522-2594
pISSN - 0740-3194
DOI - 10.1002/mrm.22800
Subject(s) - receiver operating characteristic , pattern recognition (psychology) , segmentation , artificial intelligence , feature (linguistics) , mathematics , breast mri , nuclear medicine , computer aided diagnosis , computer science , feature selection , medicine , breast cancer , mammography , statistics , cancer , linguistics , philosophy
A multiparametric computer‐aided diagnosis scheme that combines information from T 1 ‐weighted dynamic contrast–enhanced (DCE)‐MRI and T 2 ‐weighted MRI was investigated using a database of 110 malignant and 86 benign breast lesions. Automatic lesion segmentation was performed, and three categories of lesion features (geometric, T 1 ‐weighted DCE, and T 2 ‐weighted) were automatically extracted. Stepwise feature selection was performed considering only geometric features, only T 1 ‐weighted DCE features, only T 2 ‐weighted features, and all features. Features were merged with Bayesian artificial neural networks, and diagnostic performance was evaluated by ROC analysis. With leave‐one‐lesion‐out cross‐validation, an area under the ROC curve value of 0.77 ± 0.03 was achieved with T 2 ‐weighted‐only features, indicating high diagnostic value of information in T 2 ‐weighted images. Area under the ROC curve values of 0.79 ± 0.03 and 0.80 ± 0.03 were obtained for geometric‐only features and T 1 ‐weighted DCE‐only features, respectively. When all features were considered, an area under the ROC curve value of 0.85 ± 0.03 was achieved. We observed P values of 0.006, 0.023, and 0.0014 between the geometric‐only, T 1 ‐weighted DCE‐only, and T 2 ‐weighted‐only features and all features conditions, respectively. When ranked, the P values satisfied the Holm–Bonferroni multiple‐comparison test; thus, the improvement of multiparametric computer‐aided diagnosis was statistically significant. A computer‐aided diagnosis scheme that combines information from T 1 ‐weighted DCE and T 2 ‐weighted MRI may be advantageous over conventional T 1 ‐weighted DCE‐MRI computer‐aided diagnosis. Magn Reson Med, 2011. © 2011 Wiley‐Liss, Inc.