The Impact of a Bayesian Penalized-likelihood Reconstruction Algorithm on Detection of Small Liver Lesions.

Autor: Mateva, Gabriela, Kostadinova, Irena, Garcheva, Marina
Předmět:
Zdroj: Nuclear Medicine Seminars; 2019 Supplement 1, Vol. 5, p41-42, 2p
Abstrakt: Aim: To demonstrate the benefit of implementing different reconstruction methods for better lesion detection and for increased interpretation confidence. Method: We present the case of 51 years old female patient diagnosed with sigmoid colon adenocarcinoma two years earlier. As initial treatment a hemicolectomy was performed, followed by 6 courses FOLFOX. One year after the treatment completion, the levels of CA19-9 started to elevate steadily up to 71.29 U/L. Three consecutive contrast enhanced CT scans were performed in the course of the year, and all were interpreted as negative findings for relapse. After the negative diagnostic contrast enhanced CT scans, a FDG scan was performed at our hospital on a Discovery IQ five ring PET/CT scanner. After i.v. administration of 233 MBq of FDG a low dose CT acquisition was followed by an emission scan at two minutes per bed in 3D mode. Two PET reconstructions were performed using ordered subset expectation maximization (OSEM) algorithm and bayesian penalized likelihood reconstruction algorithm (Q.Clear), which improves signal to noise and signal to background ratios. Results: Q.Clear reconstruction showed prominent uptake within the first liver segment, which was not well defined and visible on the standard OSEM reconstruction. If we apply the PERCIST criteria to this lesion based on the SUVpeak measured on the Q.Clear reconstruction (lesion SUVpeak 3.3, liver SUVmean = 1.73 with SD=0.14), the lesion should be considered as measurable. However if we apply it to the OSEM reconstruction, (SUVpeak=2.41, liver SUVmean=1.78 with SD=0.11) the lesion would be considered as non measurable. Finally, based on the Q.Clear reconstruction we interpreted the finding as a solitary liver metastases. The patient was referred for liver resection, followed by chemotherapy. The histopathology report confirmed the diagnosis - metastases of moderately differentiated adenocarcinoma. Five months later a follow-up PET/CT scan was performed and it showed complete response. Conclusion: The use of Q.Clear reconstruction with FDG PET/CT could be used in patients with elevated CA19-9 levels and negative results form other imaging techniques to identify early and reliably small lesions in the liver. In our case Q.Clear increased the confidence and specificity of the final result and helped for the choice of further treatment. [ABSTRACT FROM AUTHOR]
Databáze: Complementary Index