Program Information
Weekly MRI for Dose Assessment of Organs at Risk During Treatment of HN Cancer of the Oropharynx
K Ludwig1*, J Li1 , P Venigalla1 , J Zhang1 , X Tang2 , N Tyagi1 , S Fontenla1 , N Lee1 , Z Saleh2 , (1) Memorial Sloan Kettering Cancer Center, New York, New York, (2) Memorial Sloan Kettering West Harrison, West Harrison, NY
Presentations
SU-F-J-104 (Sunday, July 31, 2016) 3:00 PM - 6:00 PM Room: Exhibit Hall
Purpose:
Investigate the feasibility of using weekly MRI to assess dose to organs at risk utilizing deformable image registration.
Methods:
Sixteen H&N patients with oropharyngeal cancer were imaged on a 3T MR scanner using T2W and mDIXON sequence. Patients were imaged on a weekly basis in treatment position. Parotids (LP & RP), submandibular glands (LS, RS), and oral cavity (OC) were delineated on the weekly MR and reviewed by a board certified radiation oncologist. The original planning CT (pCT), RT-Dose, and RT-Structures were deformed and registered to each weekly MRIs. The deformed CTs and RT-Structures were imported to the treatment planning system (TPS) and rigidly registered to the pCT. Forward dose calculation of the original RT-Plan was used to estimate the delivered dose on the deformed CT. The dose volume histograms (DVH) statistics were performed to compare planned dose, deformed dose, and forward calculated dose. In addition, Dice similarity metric (DSM) was used to compare deformed and reference structures.
Results:
The average (min,max) DSM between deformed and reference structures was 0.71 (0.69,0.93); 0.70 (0.64,0.89); 0.65 (0.48,0.86); 0.63 (0.37,0.89); and 0.63 (0.58,0.87); for LP, RP, LS, RS, and OC respectively. The respective average relative structures volumes changed at a weekly rate of -4.99%; -4.40%; +3.45%; +1.46%; -1.39%, respectively. The percentage difference %(min,max) between estimated delivered dose and planned dose was +3.94 (-51.3,+30.5); +6.33 (-58.6,+82.7); +2.46 (-38.9,+37.6,); +2.38(-49.0,+28.9); +3.55(-17.0,+43.1).
Conclusion:
The recalculated dose based on weekly MRI deviated from planned dose for all OARs. Meanwhile, the deformed dose did not reflect the subtle changes in OARs as compared to the recalculated dose. This study demonstrates the feasibility of using weekly MRI to monitor volumetric changes which has important implications on actual delivered dose.
Contact Email: