Interventional Oncology
Leening P. Liu, B.S.
graduate student
University of Pennsylvania
Disclosure(s): No financial relationships to disclose
Rizza Pua, Ph.D.
post doc
University of Pennsylvania
Derick N. Rosario-Berrios, B.S.
graduate student
University of Pennsylvania
Olivia F. Sandvold, B.S.
graduate student
University of Pennsylvania
Michael C. Soulen, MD
Director Interventional Oncology
Abramson cancer Center, University of Pennsylvania
Nadav Shapira, Ph.D.
research associate
University of Pennsylvania
Peter B. Noël, Ph.D.
associate professor
University of Pennsylvania
Intra-procedural real-time temperature feedback for thermal ablation provides real-time monitoring as an alternative to the current standard of post-ablation contrast enhanced computed tomography (CT) for evaluating the ablation zone and reducing local recurrence. Not only would this feedback ensure direct visualization of temperature change critical to ablation success, it would also allow monitoring of surrounding critical structures subject to damage. We evaluate spectral CT thermometry for temperature monitoring by relating physical density and temperature through thermal volumetric expansion and assessing the resulting temperature sensitivity.
Materials and Methods: A liver mimicking phantom with similar attenuation and thermal properties of liver was synthesized and embedded with a fiber Bragg grating temperature sensor as a ground truth measurement. To determine the effect of temperature on physical density, the phantom was then scanned with a dual-layer spectral CT at varying dose levels (CTDIvol 5 – 56.8 mGy) while it was heated to 80 °C and cooled to 37 °C. Physical density maps were generated from clinically available spectral results and measured at the corresponding sensor location for every 5 °C. Temperature and physical density were related through thermal volumetric expansion and linearly fit to assess the relationship. In addition, temperature sensitivity was determined for various scanning and reconstruction parameters for optimization.
Results: Temperature accuracy using thermal volumetric expansion ranged from 1.5 to 6.0 °C, which stemmed from the strong relationship between physical density and temperature characterized by a slope of 0.000519 mL/g°/C and a R value of 0.9942. Accuracy improved with decreased noise as a result of higher radiation dose, larger slice thickness, smooth reconstruction kernel, and higher iterative reconstruction level.
Conclusion: Spectral physical density quantifications provide a precise and accurate relationship with temperature through the principle of thermal volumetric expansion. Changes in physical density provide real-time intra-procedural thermal maps to help avoid injury to adjacent structures and ultimately reduce local tumor recurrences after thermal ablation.