271 - Assessing Multi-Modality Imaging of Radiation Damage to Bone Marrow through Rigid Registration and Deformable Image Registration

C. Eckrich¹, J. J. Meudt², D. Shanmuganayagam², J. Shah³, A. McMillan⁴, K. Li⁵, A. Pirasteh⁴, J. Miller¹, and M. Lawless¹; ¹Department of Human Oncology, University of Wisconsin Hospitals and Clinics, Madison, WI, ²University of Wisconsin-Madison, Department of Animal Science, Madison, WI, ³Siemens Healthineers, Cary, NC, ⁴Department of Radiology, University of Wisconsin School of Medicine and Public Health, Madison, WI, ⁵Department of Medical Physics, University of Wisconsin School of Medicine and Public Health, Madison, WI

Purpose/Objective(s): To assess the change in bone marrow (BM) after radiation using multi-energy CT (MECT), proton density fat fraction MRI (FFMRI), and 18F-fluorothymidine (FLT) PET through both rigid registration (RR) and deformable image registration (DIR). Materials/

Methods: Six human-sized Wisconsin Miniature Swine^TM underwent Dual Spiral (DS) and Fast kVp Switching (FKS) MECT scanning, FFMRI, and FLT PET. 6 or 8 Gy was delivered in a single fraction to BM volumes in the lumbosacral spine and pelvis using an MRgRT system. Imaging was repeated at 7, 28, and 56 days post-irradiation. BM changes were assessed by quantifying differences in image signal in the bones at different time points for each modality. The measured signals from the imaging modalities include Hounsfield unit (HU) for DS, water concentration (mg/cc) for FKS, percent fat fraction (%FF) for MR, and SUV_mean on PET. An RR was performed for each bony ROI to account for variation in positioning between the different time points. DIR was also performed to assess its robustness in longitudinal studies with the imaging modalities.
Results: Table 1 shows the change in the ROI mean relative to the baseline data from day 0 for the unirradiated spine (US), irradiated spine (IS), and irradiated pelvis (IP). For MECT, the signal decreased over time for the IS and IP, while in MR, the signal increased in those same regions. Large changes were observed seven days post-irradiation for PET but were transient. The max(mean) differences between DIR and RR for each modality were -24.63(4.08) HU for DS, 11.68(3.0) mg/cc for FKS, -9.85(-2.9) %FF for MR, and 2.39(0.1) SUV_mean for PET.
Conclusion: The investigation into various imaging modalities revealed their effectiveness in evaluating changes in BM after irradiation. PET and MR scans clearly showed reduced BM activity in the irradiated areas. Changes in MECT were less conspicuous qualitatively but still exhibited statistically significant changes relative to the baseline. DIR produced similar results to the RR for PET and MECT, but unwanted distortions were seen in the DIR for MR, altering the calculation of longitudinal signal change. The difference in mean values between baseline and days 7, 28, and 56 for unirradiated spine (US), irradiated spine (IS), and irradiated pelvis (IP) were analyzed for each modality. The partial irradiated volumes are not included in the IS and IP. The US includes C3 and L1-L3, while the IS typically includes L4-L7. The IP includes S1 and both ilia. Abstract 271 – Table 1

	DS CT (HU)			FKS CT (mg/cc)				MRI (FF)				PET (SUV)
		Days				Days				Days				Days
	7	28	56		7	28	56		7	28	56		7	28	56
US	1.94	11.20*	4.06		-2.88	9.92	1.75		-3.69*	-9.29*	-2.71		4.20*	0.90	0.44
IS	-15.56*	-12.61*	-21.82*		-13.02*	-15.13*	-18.32*		3.90*	25.79*	28.80*		-1.25*	-0.93*	-0.87*
IP	-9.29*	-6.29*	-11.78		-4.11	-5.35*	-6.07		-0.53	12.11*	11.30*		-0.33*	-0.35*	-0.51*
													*significant (p<0.05)