J Deng, T Aguilera, R Hannan, A Godley*, University of Texas Southwestern Medical Ctr, Dallas, Texas

• 
  A Godley

MO-E115-IePD-F6-4 (Monday, 7/11/2022) 1:15 PM - 1:45 PM [Eastern Time (GMT-4)]

Exhibit Hall | Forum 6

Purpose: Motion estimation is important in radiation therapy simulation process to determine the internal target volume (ITV) for treatment planning/delivery. In current MRI simulation, single-shot cine imaging through mid-slice of tumor is acquired continuously to visualize and estimate the range of target movement. Development of time-resolved volumetric (4D) MRI relies on sparsely under-sampled radial K-space trajectory and complicated image reconstruction algorithms. However, this technique is not currently available for clinical use. In this study, we exploited a clinically available real-time 3D sequence (4D-THRIVE) with sub-second temporal resolution to estimate ITV based on sets of grouped projection images.

Methods: The 4D-THRIVE sequence accelerates dynamic body imaging through a keyhole method to achieve combined good spatial resolution (2.5ₓ2.5mm², 25 slices) and high temporal resolution (0.8 second/frame). Coronal/sagittal images covering the target were acquired on each patient under free breathing. The time-averaged image across all time points on each slice was generated for the clinician to draw a pair of lines across the motion-blurred target boundary in SI/RL or SI/AP directions on a selected slice. The signal intensities along each direction went through K-means (N=3) clustering to determine three motion statuses of the target. Three Maximum Intensity Projection (MIP) images were formed from the images in each of the three motion states. The position of the tumor across the three MIPs provided the ITV estimation. We tested the method on a motion phantom then on patients with abdominal and lung cancers.

Results: 4D-THRIVE MRI was successfully performed in 20 patients and provided real-time imaging of the entire tumor volume. Three MIP images generated with minimal clinician involvement demonstrated motion-induced target position changes that facilitate estimation of ITV margins.

Conclusion: We proposed a clinically available method of combining 4D-THRIVE with simple post-processing method for ITV estimation in RT planning, negating the need for 4DCT.

Keywords

Image Analysis, MRI, Targeted Radiotherapy

Taxonomy

IM/TH- MRI in Radiation Therapy: MRI for treatment planning

Contact Email