UCLA Journal of Radiation Oncology Issue 4 - Flipbook - Page 25
UCLA RADIATION ONCOLOGY JOURNAL
In March of 2019, the first UCLA Radiation Oncology patient was scanned using the 5DCT protocol
for treatment planning purposes. This initial clinical scan was the culmination of years of research
and development, beginning with a widely cited 2005 Red Journal publication by Dr. Low and
colleagues, describing a “Novel breathing motion model for radiotherapy.” Extensive work at UCLA
was done to validate the technique before employing it clinically, including an animal study and a
comparison against commercial 4DCT in a cohort of patients on an IRB-approved imaging study. To
date, over 200 UCLA Radiation Oncology patients have been scanned using the 5DCT protocol. Since
its clinical rollout, the increased accuracy and image quality afforded by UCLA’s 5DCT technique has
contributed to the success and growth of the department’s thoracic SBRT program.
In addition to bringing tangible improvements to our clinical workflow, the adoption of 5DCT has
enabled our group to pursue multiple new avenues of research. One of the greatest failures of
4DCT has been the impact on scientific studies of free-breathing patients. 4DCT uses significant
assumptions and approximations when analyzing images. These create well-established sorting
artifacts. Our current work in pulmonary biomechanical modeling and ventilation, as well as in
model-based cone-beam CT (CBCT), would be infeasible or greatly hindered with access to only
artifact-prone commercial 4DCT images.
Biomechanical models of human anatomy are critical in understanding physiology and have been
developed for applications ranging from deformable image registration to image-guided treatment
planning. We can use the motion characterization provided by 5DCT to calculate ventilation
regionally throughout the lungs. Using the 5DCT model, we can generate motion vectors from
exhalation to inhalation and calculate the Jacobian determinant, which measures the expansion of
the vector field. This can be used to define functional regions of the lungs for applications such as
function sparing treatment planning or COPD diagnosis. A second biomechanical property that can
be calculated from 5DCT is tissue elasticity. Using the motion vectors provided from 5DCT, we can
obtain elasticity with an iterative model. Tissue elasticity is also promising for functional sparing
treatment planning and disease characterization, and combining ventilation and elasticity could
provide a powerful, comprehensive tool.
25
