Title: The effect of inter-patient attenuation coefficient variability on segmented attenuation correction for PET
Abstract:Objectives: Using a segmented attenuation map for PET attenuation correction can introduce errors because of the differences in average attenuation coefficient (AC) between patients. We investigated t...Objectives: Using a segmented attenuation map for PET attenuation correction can introduce errors because of the differences in average attenuation coefficient (AC) between patients. We investigated this inter-patient variability for lung tissue, cortical bone and liver tissue and estimated the effect on reconstructed PET images.
Methods: Lungs, liver and cortical bone were segmented using thresholding and region-growing in 20 whole-body CT data sets. For each subject and each tissue type the average AC was calculated. The standard deviation of the average values was then calculated as a measure of inter-patient variability. A whole-body FDG scan of the NCAT phantom was simulated. Four 2 cm lesions were placed in the lungs, spine, liver and prostate. The data was reconstructed using 4 segmented attenuation maps: a reference attenuation map with the correct average AC assigned to each tissue class and 3 modified attenuation maps with a larger AC in the lungs, cortical bone or liver (+ 1 SD).
Results: The average AC for each patient is shown in the figure. The inter-patient variability is clearly largest in lung tissue (SD = 0.004 cm-1), followed by cortical bone (SD = 0.002 cm-1) and liver tissue (SD = 0.001 cm-1). The table shows the effect on the reconstructed images. The largest effect is seen for the lung lesion with incorrect lung attenuation values (8.2 %). Changing the bone or liver AC has little effect, even on the spine (1.5 %) or liver (2.1 %) lesion respectively.
Conclusions: Inter-patient variability of lung attenuation coefficients can have a significant effect on quantitation in PET. The effect of inter-patient variability on cortical bone and liver attenuation is small.Read More
Publication Year: 2010
Publication Date: 2010-05-01
Language: en
Type: article
Access and Citation
Cited By Count: 7
AI Researcher Chatbot
Get quick answers to your questions about the article from our AI researcher chatbot