Utility of respiratory motion correction and effects on dosimetry in imaging with integrated Y-90 PET/MRI after radioembolization of liver tumors.

PURPOSE: In this study we aimed to investigate the utility of respiratory motion correction using the Q.Static algorithm in integrated Y-90 PET/MR imaging and its impact on dosimetric accuracy, image quality, and dose-volume histograms following radioembolization of the liver tumors.

METHODS

This retrospective single-center analysis included 23 patients treated with Y-90 resin microspheres and imaged with Y-90 PET/MRI. Respiratory motion-corrected (Q.Static) and standard non-corrected PET reconstructions were compared for dosimetric variables, image quality (signal-to-noise ratio, SNR), and activity quantification. Tumor dose-volume histograms (D1-D99) and absorbed doses were calculated with absolute quantification and local dose deposition method. Bland-Altman plots and regression analyses were employed to assess differences between the two methods and investigate correlations with tumor volume.

RESULTS

Q.Static sequences resulted in 7.08 % (95 % CI: [+4.82 %] - [+9.34 %]) higher tumor, 9.38 % (95 % CI: [+7.55 %] - [+10.67 %]) higher normal tissue absorbed doses compared to standard reconstructions, resulting in similar T/N ratios. In addition, Q.Static images demonstrated significantly higher D1, D5, and D10 values but lower D90 and D95 values. SNR was 33.62 % lower, and extrahepatic activity was 46.65 % higher in Q.Static sequences, reflecting noise-related degradation. Tumor volume significantly influenced dose variability, with smaller lesions showing greater differences between methods.

CONCLUSION

Respiratory motion correction in Y-90 PET/MRI may improve contrast and dose quantification in smaller lesions but may reduce image quality in low-activity cases due to noise. While promising for accurate dosimetry, its impact on clinical outcomes and response prediction warrants further studies.