We are working on fast and robust geometrical correction techniques (system, patient and sequence specific), image quality, image uniformity and patient immobilization require for MRI guided radiotherapy and quantitative MRI imaging for post respond measurements and adaptive radiation therapy.
Purpose: We are commissioning 3T MRI scanner for brain fractionated radiotherapy (RT) and stereotactic radio surgery (SRS). We defined our geometrical integrity baseline using big field of view phantom, QUASAR MRI3D, Modus Medical Devices Inc., Canada during commissioning and repeated bi-weekly for reproducibility. We also run weekly MRI ACR QC tests using standard ACR phantom. Using big phantom bi-weekly is time consuming, therefore, we propose to use weekly ACR phantom relative measurements for geometrical distortion evaluation.
Methods: We summarized ACR measurements during the initial seven weeks Figure 1. The results compared with big phantom at two different diameter volume surface (DVS), 320 mm and 203 mm to match with ACR phantom Figure 2.
Results: We summarized big phantom measurements during commissioning, Table 1 and 2. For 320 mm DVS, (X Mean= 0.4 0.29 mm, X Max= 1.45 mm, YMean= 0.21 0.18 mm, YMax= 0.94 mm, ZMean = 0.35 0.23 mm, ZMax=0.93 mm, Absolute Mean = 0.66 0.25 mm, Absolute Max= 1.55 mm) and 203 mm DVS, (X Mean= 0.25 mm, XMax = 0.3 mm, YMean= 0.2 mm, YMax= 0.3 mm, ZMean= 0.5 mm, ZMax= 0.5 mm, Absolute Mean= 0.72 mm, Absolute Max= 0.75 mm). We compare ACR measurements with baseline to define our relative baseline (r-baseline), Table 3. We reported 54% on Mean and 65% on Max differences at X-direction (r-baseline Mean= 0.84 mm and r-baseline Max= 1.44 mm), 42% on Mean and 65% on Max differences on Y-direction (r-baseline Mean= 0.49 mm, r-balance Max= 1.44 mm) and 17% on Mean and 43% on Max difference on Absolute measurement (r-baselineMean= 1.02 mm, r-baselineMax= 1.9 mm).
Conclusions: We will continue our bi-weekly big phantom measures for the next 6 months to confirm ACR measurements accuracy and reproducibility. Also, we are working on new 3D analysis approach for proper comparison.
https://drive.google.com/file/d/15aHFx7rR-T2iN1qgDteasLwQm6VNK_5P/view
https://drive.google.com/file/d/1rrehiyPKpg6D405DiCT3IS8ncbUEXIJJ/view