Room: AAPM ePoster Library
Purpose: Diffusion weighted imaging (DWI) is extremely sensitive to patient motion, requiring consistent anatomical position across b-value images for apparent diffusion coefficient (ADC) calculation. This work presents the impact of respiratory motion management (RMM) with patient biofeedback on DWI of liver SBRT cancer patients.
Methods: Eight patients were imaged pre- and post-treatment on a 1.5T MRI with navigator triggered acquisition and no RMM. Two patients were imaged once before and twice after treatment on a 3.0T MRI with a manually triggered sequence during biofeedback guided breath hold. Biofeedback was provided using an in-house developed RMM system and an MRI safe monitor displaying real-time breathing traces to the patient. Clinical imaging protocol was followed for the 1.5T machine (b = 50, 400, 800s/mm²) and an eight b-value protocol on the 3.0T scanner (b = 0-1000s/mm²). ADC values were calculated in ROIs using least squares fitting with all b-values. ROIs were drawn in the GTV, liver, and spleen for all patients, and liver dose regions receiving >50Gy, 25 to 50Gy, and <25Gy. Liver dome position was compared on coronal views for all b-values.
Results: SNR of b0 images was significantly greater on the 3.0T system than the 1.5T system as expected (Average 515 versus 95). Liver dome superior-inferior position varied by 3.1mm and 1.04mm on average without and with visual guidance respectively, with maximum values of 9.5mm and 3.4mm. The standard deviation of tumor ADC values was two times greater for images acquired on the 3.0T system than on the 1.5T system but similar in the spleen and low dose liver.
Conclusion: The RMM system was able to significantly reduce liver dome motion during DWI acquisition, possibly improving the localization of ADC values. The increased STD in the tumor may indicate increased tissue heterogeneity which is not visible with respiratory motion.
Not Applicable / None Entered.