(Sunday, 7/12/2020) [Eastern Time (GMT-4)]
Room: AAPM ePoster Library
Purpose: An IRB approved prospective clinical trial was developed to test longitudinal multi-parametric MRI (MP-MRI) (DWI, DCE) features used as imaging biomarkers to assess treatment response for prostate radiotherapy. The study aims to characterize longitudinal changes of MRI imaging features between the dominant intraprostatic lesion (DIL) and normal prostate over the radiation treatment and stability post-treatment.
Methods: MRIs were taken at four time points: pretreatment, mid-treatment, end of treatment, and two months post treatment. This study currently has seven patients enrolled. Radiation treatment consisted of 78Gy prescribed to the entire prostate. The extended Tofts model was used for pharmacokinetic modeling. Parametric maps were generated with dynamic contrast enhanced (DCE) data for volume transfer constants Ktrans and Kep. An experienced radiologist contoured DILs on each MRI modality (T2, ADC, Ktrans and Kep). A normal prostate contour was drawn on the contralateral side to calculate relative DIL response. Spearman’s Rank correlation coefficient identified relevant trends in 300 features extracted using the Cancer Imaging Phenomics Toolkit. Relevant features were identified using a pretreatment to mid-treatment threshold change of 40% for T2, Kep, Ktrans, and a 20% threshold for ADC.
Results: Over the four imaging modalities, 62 features were identified that related dose response of the DIL relative to the normal prostate. These features either trended monotonically over time (such as the histogram texture energy), or there was an initial change from the pretreatment to mid-treatment that subsequently recovered back towards the pretreatment value in the successive scans (gray level busyness). Strong trends were found in the histogram texture energies, which showed up to a 65% changes between the pretreatment and mid-treatment scans.
Conclusion: This study identified relevant features that correlate tumor response to radiation dose, and provided metrics to track tumor changes and subsequent post treatment stability.
Funding Support, Disclosures, and Conflict of Interest: Funding support: Research Scholar Grant, RSG-15-137-01-CCE from the American Cancer Society
Keywords
MRI, Quantitative Imaging, Functional Imaging
Taxonomy
Contact Email
