Visualizing whole-body treatment response heterogeneity using multi-parametric magnetic resonance imaging. Issue 4 (December 2016)

• Record Type: Journal Article
• Authors: Blackledge, Matthew D; Rata, Mihaela; Tunariu, Nina; Koh, Dow-Mu; George, Angela; Zivi, Andrea; Lorente, David; Attard, Gerhardt; de Bono, Johann S; Leach, Martin O; Collins, David J
• Other Names: Chen Ke guest-editor.; Lai Choi-Hong guest-editor.
• Is Part Of: Journal of algorithms & computational technology. Volume 10:Issue 4(2016)
• Page Start: 290
• Page End: 301
• Publication Date: 2016-12
• Subjects: Tumor heterogeneity -- whole-body magnetic resonance -- metastases -- apparent diffusion coefficient and fractional enhancement -- probabilistic image interpretation; Computer algorithms -- Periodicals; Numerical calculations -- Periodicals; Computer algorithms; Numerical calculations; Periodicals; 518.1
• DOI: 10.1177/1748301816668024
• Languages: English
• ISSNs: 1748-3018

Abstract

A novel post-processing methodology able to assess whole-body tumor heterogeneity in patients with metastatic disease is proposed. The method is demonstrated on paired pre- and post-treatment data sets obtained from an initial cohort of six patients with metastatic disease from primary prostate or ovarian cancers. Whole-body diffusion-weighted imaging and T1 -weighted contrast-enhanced imaging data were acquired covering the chest, abdomen, and pelvis. Joint histograms of Apparent Diffusion Coefficient and Fractional Enhancement values were calculated within volumes of interest and were modeled as a Gaussian mixture of two classes. Probability maps and volumetric estimates of the magnetic resonance data-derived classes providing visualization of pre- and post-treatment data are shown in three patient examples. This technique provided spatially heterogeneous characterization of regions following treatment as defined by the combined analysis of apparent diffusion coefficient and fractional enhancement. A new whole-body magnetic resonance data analysis has been demonstrated enabling visualization of intra-patient response heterogeneity in patients with metastatic cancer. Changes in the parameters of each subpopulation derived from this technique (apparent diffusion coefficient and fractional enhancement) reflect changes in the tissue properties of each subpopulation following treatment. Furthermore, the volume change of each population can be quantified. Such techniques may be essential for personalized anti-cancer therapy where there is a need to detect early drug-resistance and monitor heterogeneous response.