S Shen*, R Sullivan, C Schneider, R Popple, R Jacob, University of Alabama Birmingham, Birmingham, AL

• 
  S Shen

PO-GePV-M-123 (Sunday, 7/25/2021)   [Eastern Time (GMT-4)]

Purpose: IMR (Philips Healthcare) is widely used in diagnostic imaging. Use of IMR is not well documented in abdominal 4DCT for radiotherapy, where pixel counts are low. This study compared image quality and visibility between IMR and iDose4.

Methods: A Gammex Model 467 tissue phantom was used for CT calibration. Contrast and resolution were evaluated using Philips QA CT phantom. Ten patients had 4DCT abdominal scans for treatment planning, with 160 mAs used for each phase image compared to 305 mAs used for conventional 3D imaging. Images were reconstructed using IMR and iDose4. The objective measurements included noise (pixel HU histogram), line profile crossing target (boundary gradient). Two independent observers provided subjective assessment on image quality and visibility for contouring target and organs.

Results: For the tissue phantom, the mass density ranged 0 to 1.69 and HU ranged -1000 to 1170; the HU mean difference was -0.8 HU (STD 2.6 HU) between IMR(1) and iDose4(3). For the Philips QA phantom with multi plugs, HU ranged from 0 to 910 and, the HU mean difference was -0.9 HU (STD 0.8 HU) between IMR(1) and iDose4(3). There was no measurable difference in contrast and resolution using the Philips QA phantom, but IMR had less objective noise in rods and background compared to iDose4. For 4DCT in patients, there was no observable difference in target boundary gradient between IMR and iDose4. IMR had objective noise reduction (pixel histogram HU range) as much as 21% in targets and 29% in background tissue compared to that of iDose4. IMR had better image quality and visibility for contouring target and organ in all cases.

Conclusion: For motion management in treatment planning, abdominal 4DCT has low pixel counts. We found IMR had better image quality and visibility when compared to that of iDose4.

ePosters

Keywords

Not Applicable / None Entered.

Taxonomy

Not Applicable / None Entered.

Contact Email