Page 101 - Quantitative Imaging of Small Tumours with Positron Emission Tomography

Chapter 4 Conclusion In [18F]FDHT PET-CT studies, noise-induced SUV variability leads to small increases in test-retest variability, which improves when excluding small lesions. Novel EARL2-compliant reconstruction increases lesion SUVs and marginally increases CNRs. However, it requires higher count statistics to preserve adequate precision. For SUVpeak normalized to bodyweight, test-retest variability remained below 30% when lesions <4.2 ml were excluded, which is generally acceptable for oncological [18F]-tracers. In contrast, SUVmax was substantially affected by count reduction and EARL2 reconstruction, hence its use should be avoided. Lesion detectability is only slightly impaired by reducing counts by 50%, which might not be clinically relevant in the mCRPC population. Taken together, with the current imaging procedure on an analogue PET- CT system count statistics are more than sufficient and could even be reduced by 50% without affecting diagnostic performance and at a small expense of reduced precision of response assessments. Acquisition time reduction is feasible for staging and response assessment purposes. 100