Assessing tumor microstructure with time-dependent diffusion imaging: Considerations and feasibility on clinical MRI...

Area of Science:

• Biomedical Imaging
• Radiotherapy Research
• Quantitative MRI

Background:

• Quantitative imaging biomarkers (QIBs) aid radiotherapy (RT) guidance by characterizing tumor heterogeneity.
• Time-dependent diffusion MRI (TDD-MRI) offers enhanced specificity over traditional DW-MRI by assessing restricted diffusion and diffusional exchange.
• Complex TDD-MRI modeling and hardware requirements limit clinical applicability, particularly in RT settings.

Purpose of the Study:

• Investigate a simplified TDD-MRI analysis framework for detecting diffusion effects.
• Evaluate feasibility on conventional 1.5 T MRI and low-gradient 1.5 T MRI-Linac systems.
• Promote clinical applicability using standard gradient waveforms and moderate gradient strengths.

Main Methods:

• Simulated diffusion MRI signal in tumor-mimicking geometries to optimize experimental parameters.
• Implemented TDD-MRI using pulsed field gradient spin echo on conventional MRI and MRI-Linac.
• Acquired TDD-MRI data in green asparagus and 10 brain lesion patients to assess TDD contrast.

Main Results:

• Simulations showed TDD contrast differentiates restricted diffusion from diffusional exchange based on cell size.
• Conventional MRI achieved >5% TDD contrast in simulations and asparagus; MRI-Linac yielded <5% and ~2% for typical cancer cell sizes.
• Conventional MRI revealed distinct tumor sub-regions indicating dominant restricted diffusion or diffusional exchange in brain lesions.

Conclusions:

• TDD contrast maps on conventional MRI show tumor sub-regions, potentially indicating spatial heterogeneity.
• MRI-Linac TDD contrast trends matched conventional MRI but were near noise levels for common cancer cell sizes.
• On conventional MRI, TDD contrast can guide biophysical model selection for precise tumor characterization.