Potentials and challenges of diffusion-weighted magnetic resonance imaging in radiotherapy

Abstract

Purpose: To review the potential and challenges of integrating diffusion weighted magnetic resonance imaging (DWI) into radiotherapy (RT).

Content: Details related to image acquisition of DWI for RT purposes are discussed, along with the challenges with respect to geometric accuracy and the robustness of quantitative parameter extraction. An overview of diffusion- and perfusion-related parameters derived from mono- and bi-exponential models is provided, and their role as potential RT biomarkers is discussed. Recent studies demonstrating potential of DWI in different tumor sites such as the head and neck, rectum, cervix, prostate, and brain, are reviewed in detail.

Conclusion: DWI has shown promise for RT outcome prediction, response assessment, as well as for tumor delineation and characterization in several cancer types. Geometric and quantification robustness is challenging and has to be addressed adequately. Evaluation in larger clinical trials with well designed imaging protocol and advanced analysis models is needed to develop the optimal strategy for integrating DWI in RT.

Fig. 1

Applying different models to extract quantitative parameters from the mean DWI signals derived within the gross tumor volume (GTV) of a oropharyngeal cancer patient. Blue: mono-exponential model (mExp) using high and low b-values (mExp_all), orange: mExp using only high b-values $>200$ $\mathrm{s}/{\mathrm{mm}}^2$ (mExp_high), green: intra-voxel incoherent motion (IVIM) model. Fit parameters are ADC = 1293 × ${10}^{-6}{\mathrm{mm}}^2/\mathrm{s}$ (mExp_all), $\text{ADC}=1018\times {10}^{-6}{\mathrm{mm}}^2/\mathrm{s}$ (mExp_high), and $f=0.21,D=958\times {10}^{-6}{\mathrm{mm}}^2/\mathrm{s}$, $D^{\ast }=16,500\times {10}^{-6}{\mathrm{mm}}^2/\mathrm{s}$ (IVIM). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 2

Dedicated positioning solution for MR imaging of head and neck cancer patients in RT treatment position. The components are a flat table top with an add-on for the fixation of a head and neck positioning mask, and coil holders to with flexible RF coils can be attached.