2020-488 AUTOMATED TREATMENT PLANNING FOR BREATHING MOTION

SUMMARY

UCLA researchers in the Department of Radiation Oncology have developed a novel computed tomography (CT)  scanning protocol to not only solve the problems of image artifacts of commercial 4DCT, but also enable automation in treatment planning and quantification of the CT images beyond what is currently available.

BACKGROUND

Radiation therapy targeting lung and upper abdominal cancers uses various breathing/respiratory motion techniques including specialized CT scanning methods. Respiratory motion is significantly challenging for radiation therapy due to its irregularity and consequential uncertainty of tumor and normal organ positions, resulting in larger than necessary target volumes and potentially systematic errors in radiotherapy.

Breathing motion in radiotherapy is commonly managed with four-dimensional computed tomography (4DCT). Accurate quantification of 4DCT data is difficult due to breathing irregularities amongst patients; as a result, 4DCT is susceptible to subjective biases and errors in targeting tumor volume for radiotherapy. . There is a critical need for a more accurate CT scanning protocol that accurately measures breathing motion without relying on individual patient breathing rates.

INNOVATION

Dr. Daniel Low and colleagues in the Department of Radiation Oncology at UCLA have developed a novel CT scanning protocol that resolves the shortcomings of commercial 4DCT. This new CT scan protocol uses a model-based CT (MBCT or 5DCT) data, including images, motion model and breathing proxy to inform treatment planning decisions and optimized treatment solutions for lung cancer and upper abdominal cancer patients

APPLICATIONS

  • Lung CT scan
  • Radiation therapy
  • Upper abdominal cancer CT scans

ADVANTAGES

  • Machine learning approach to inform treatment planning and approaches
  • Able to predict the impact of breathing motion on the treatment

STATE OF DEVELOPMENT

Have numerous papers on model-based CT (5DCT) but there is no published work on automating treatment planning using breathing motion models

Related Papers:

T. H. Dou, D. H. Thomas, D. O'Connell, J.M. Lamb, P. Lee, and D.A. Low. A Method for Assessing Ground-Truth Accuracy of the 5DCT Technique, Int J Radiat Oncol Biol Phys. 93(4): 925–933 (2015)

Patent Information:
For More Information:
Earl Weinstein
Associate Director of Business Development
eweinstein@tdg.ucla.edu
Inventors:
Daniel Low