Diagnostic value of apparent diffusion coefficient lesion texture biomarkers in breast MRI

• Published in: Health and technology Vol. 10; no. 4; pp. 969 - 978
• Main Authors: Tsarouchi, Marialena I., Vlachopoulos, Georgios F., Karahaliou, Anna N., Costaridou, Lena I.
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.07.2020; Springer Nature B.V
• Subjects: Biological and Medical Physics; Biomarkers; Biomedical Engineering and Bioengineering; Biomedicine; Biophysics; Breast cancer; Cancer therapies; Classification; Computational Biology/Bioinformatics; Contrast agents; Diagnostic systems; Diffusion coefficient; Engineering; Females; Heterogeneity; Image segmentation; Lesions; Magnetic resonance imaging; Mammography; Medical diagnosis; Medical imaging; Medicine/Public Health; Morphology; Open source software; Original Paper; Patients; Quantitative analysis; R & D/Technology Policy; Radiomics; Registration; Standard deviation; Statistics; Texture; ADC map; Texture analysis; Breast cancer diagnosis; Imaging lesion heterogeneity; DWI
• ISSN: 2190-7188; 2190-7196
• DOI: 10.1007/s12553-020-00452-3

Quantitative assessment of breast intra-lesion heterogeneity in terms of contrast agent free Magnetic Resonance Imaging (MRI) approaches hold potential in breast cancer diagnosis. This study focuses on an Apparent Diffusion Coefficient (ADC) based approach, investigating the diagnostic role of 1 st and 2 nd order ADC statistics features, in differentiating benign from malignant breast lesion status. A total of 67 patients with 78 histologically verified breast lesions (40 benign and 38 malignant) was analyzed. ADC maps were generated for a slice representative of lesion largest diameter, considering intra Diffusion Weighted Imaging (DWI) sequence non rigid registration scheme. Lesion segments were defined by semi-automated Fuzzy C-Means (FCM) segmentation on high b-value diffusion images and propagated on ADC maps. 27 (11 1 st order statistics and 16 2 nd order statistics (texture) features were derived. To avoid overfitting a stepwise feature selection method was employed, while the discriminating ability of features was evaluated with univariate and multivariate Logistic Regression classification. The classification performance of the diagnostic model was evaluated by means of the Area Under Receiver Operating Characteristic curve (Az index). A combination of two features, one from 1 st order statistics (25 th Percentile) and one from 2 nd order statistics, (texture Entropy), achieved high classification performance (Az = 0.965 ± 0.024), suggesting both the diagnostic significance of 1 st order statistics and texture biomarkers of the ADC map representation.