Extracting and visualizing physiological parameters using dynamic contrast-enhanced magnetic resonance imaging of the breast

• Published in: Medical image analysis 2005-08, Vol.9 (4), p.315-329
• Main Authors: Armitage, Paul, Behrenbruch, Christian, Brady, Michael, Moore, Niall
• Format: Article
• Language: English
• Subjects: Algorithms; Area Under Curve; Bayes Theorem; Breast cancer; Breast Neoplasms - drug therapy; Breast Neoplasms - pathology; Chemotherapy assessment; Contrast Media - pharmacokinetics; Diagnosis, Differential; Female; Humans; Image Processing, Computer-Assisted - methods; Imaging, Three-Dimensional; Magnetic resonance imaging; Magnetic Resonance Imaging - methods; Phantoms, Imaging; Pharmacokinetic modelling
• ISSN: 1361-8415; 1361-8423
• DOI: 10.1016/j.media.2005.01.001

An analysis procedure is presented that enables the acquisition and visualization of physiologically relevant parameters using dynamic contrast-enhanced magnetic resonance imaging. The first stage of the process involves the use of a signal model that relates the measured magnetic resonance signal to the contrast agent concentration. Since the model requires knowledge of the longitudinal relaxation time T 1, a novel optimization scheme is presented which ensures a reliable measurement. Pharmacokinetic modelling of the observed contrast agent uptake is then performed to obtain physiological parameters relating to microvessel leakage permeability and volume fraction and the assumptions made in the derivation of these parameters are discussed. A simple colour representation is utilized that enables the relevant physiological information to be conveyed to the clinician in a visually efficient and meaningful manner. A second representation, based on vector maps, is also devised and it is demonstrated how this can be used for malignant tumour segmentation. Finally, the procedure is applied to 14 pre- and post-chemotherapy breast cases to demonstrate the clinical value of the technique. In particular, the apparent improved representation of tissue vascularity when compared to conventional methods and the implications for this in treatment assessment are discussed.