MENU

Click here to

×

Are you sure ?

Yes, do it No, cancel

A Technique to Automatically Create Synthetic CT for MRI-Guided Online Adaptive Radiation Therapy of Abdominal Tumors

R Thapa*, E Ahunbay , H Nasief , X Chen , X Li , Medical College of Wisconsin, Milwaukee, WI

Presentations

(Thursday, 7/18/2019) 10:00 AM - 12:00 PM

Room: 225BCD

Purpose: Air cavity in abdomen varies day-to-day and is generally difficult to automatically segment on MRI as it often has similar intensity as other low signal regions. The purpose of this study is to develop a technique to automatically identify air cavity and generate synthetic CT (sCT) from MRI for MRI-guided online adaptive radiation therapy (MRgOART) in abdomen.

Methods: The proposed technique starts with a reference CT with segmented gross air-regions, bones, and all relevant organs and is to be performed after a daily MRI is acquired in four steps: (1) populating all the contours from the reference CT to the daily MRI, (2) automatically segmenting air cavities in the gross air-regions by establishing a formula based on a MRI intensity threshold derived from phantom/patient data, (3) refining air boundary based on MRI texture feature classification, and (4) creating sCT by copying electron densities of air, bone and all other organs from the reference CT. This technique was applied to the CT and MRI sets of 10 pancreatic cancer patients. The obtained air cavities and sCTs were compared with those created carefully with a manual process. The dosimetric difference between the two sCT sets were investigated.

Results: Automatically generated air cavities using the newly developed technique agreed well with the manually drawn air cavities with an average dice coefficient of 90% among all cases tested. Dosimetric analysis showed no significant dosimetric difference in commonly used dose volume parameters between the automatically and manually created sCT. The differences for parameters PTV: V100, Dmax, Dmin, Dmean, and D95, and duodenum: V45Gy and D1cc were 0.29, 0.63, 0.69, 0.78, 0.21, 0.89, and 0.36% respectively.

Conclusion: The newly developed technique can accurately and automatically segment air cavities and generate synthetic CT from abdominal MRI, thus, can be implemented to improve MRgOART for abdominal tumors.

Keywords

Not Applicable / None Entered.

Taxonomy

Not Applicable / None Entered.

Contact Email