Shutter‐Speed DCE‐MRI Analyses of Human Glioblastoma Multiforme (GBM) Data
Ruiliang Bai, Bao Wang, Yinhang Jia, Zejun Wang, Charles S. Springer, Zhaoqing Li, Chuanjin Lan, Yi Zhang, Peng Zhao, Yingchao Liu
Abstract
Background The shutter‐speed model dynamic contrast‐enhanced (SSM‐DCE) MRI pharmacokinetic analysis adds a metabolic dimension to DCE‐MRI. This is of particular interest in cancers, since abnormal metabolic activity might happen. Purpose To develop a DCE‐MRI SSM analysis framework for glioblastoma multiforme (GBM) cases considering the heterogeneous tissue found in GBM. Study Type Prospective. Subjects Ten GBM patients. Field Strength/Sequence 3T MRI with DCE‐MRI. Assessments The corrected Akaike information criterion (AIC c ) was used to automatically separate DCE‐MRI data into proper SSM versions based on the contrast agent (CA) extravasation in each pixel. The supra‐intensive parameters, including the vascular water efflux rate constant ( k bo ), the cellular efflux rate constant ( k io ), and the CA vascular efflux rate constant ( k pe ), together with intravascular and extravascular–extracellular water mole fractions ( p b and p o , respectively) were determined. Further error analyses were also performed to eliminate unreliable estimations on k io and k bo . Statistical Tests Student's t ‐test. Results For tumor pixels of all subjects, 88% show lower AIC c with SSM than with the Tofts model. Compared to normal‐appearing white matter (NAWM), tumor tissue showed significantly larger p b (0.045 vs. 0.011, P < 0.001) and higher k pe (3.0 × 10 −2 s −1 vs. 6.1 × 10 −4 s −1 , P < 0.001). In the contrast, significant k bo reduction was observed from NAWM to GBM tumor tissue (2.8 s −1 vs. 1.0 s −1 , P < 0.001). In addition, k bo is four orders and two orders of magnitude greater than k pe in the NAWM and GBM tumor, respectively. These results indicate that CA and water molecule have different transmembrane pathways. The mean tumor k io of all subjects was 0.57 s −1 . Data Conclusion We demonstrate the feasibility of applying SSM models in GBM cases. Within the proposed SSM analysis framework, k io and k bo could be estimated, which might be useful biomarkers for GBM diagnosis and survival prediction in future. Level of Evidence 4 Technical Efficacy Stage 1 J. Magn. Reson. Imaging 2020;52:850–863.