Click here to

Session: [Return to Session]

Development and Characterization of a Dynamic, Anatomically Realistic, Abdominal Phantom for Simulating Complex Deformable Motion in Cone-Beam CT

Y Liu1*, H Huang1, J Siewerdsen1, W Zbijewski1, C Weiss1, T Ehtiati2, A Sisniega1, (1) Johns Hopkins University, Baltimore, MD, (2) Siemens Medical Solutions USA

Presentations

TU-IePD-TRACK 1-3 (Tuesday, 7/27/2021) 12:30 PM - 1:00 PM [Eastern Time (GMT-4)]

Purpose: Cone-beam CT (CBCT) provides 3D guidance for interventional radiology and radiotherapy, but long acquisition time results in patient motion. Motion compensation methods require realistic models to replicate the complex interplay between image features and motion in CBCT. We report a phantom that presents realistic anatomical deformable motion via simple linear actuation.

Methods: The phantom incorporates numerous elements within a plastic cylindrical container (23 x 24 cm): ultrasound gel interstitium; synthetic liver (Syndaver Inc.); kidneys (SuperFlab, Bebig); aorta (3 cm PMMA cylinder); spine (4 cm Teflon tube and Delrin insert); and fat fascia (vegetable shortening - Crisco, P&G). 11 Teflon spheres (3 - 9.5 mm) and two cylindrical vessels were inserted into the liver to simulate contrast-enhanced lesions and vasculature. Key to deformable motion are two rubber apertures – one at the diaphragm connected to a linear actuator for SUP-INF motion and another to allow AP motion. CBCT images were acquired on a C-arm (body protocol, Cios Spin 3D, Siemens) with 4 - 16 mm amplitude of the linear actuator. CBCT images were registered to the static reference using Elastix (mutual information, 14³ knots) to assess the imparted deformation.

Results: X-ray attenuation of the simulated structures was accurate: 30 HU liver, 60 HU kidneys, 660 HU cortical (300 HU inner) bone, and -160 HU fat. Linear actuation yielded a rich variety of deformable motion of structures: ~30% AP in mid-anterior regions, while posterior regions remained nearly static. The magnitude and complexity of the motion field was controllable via the actuation amplitude – e.g., 0.1 - 3.8 mm and 0.3 - 12.5 mm, for 4 mm and 16 mm actuation.

Conclusion: A realistic abdominal phantom for deformable motion was designed and tested. Linear motion applied to the diaphragm generated a rich deformable motion field, enabling testing of CBCT motion compensation techniques.

Funding Support, Disclosures, and Conflict of Interest: Academic-industry collaboration with Siemens Healthineers.

ePosters

    Keywords

    Cone-beam CT, Motion Artifacts, Phantoms

    Taxonomy

    IM- Cone Beam CT: Phantoms - physical

    Contact Email

    Share: