Related Experiment Video
Updated: May 23, 2025
Computer-assisted Large-scale Visualization and Quantification of Pancreatic Islet Mass, Size Distribution and Architecture
Published on: March 4, 2011
Feasibility of simplified Patlak parametric imaging with scaled population-based input function on pancreatic cancer
Zhixin Hao1, Haiqiong Zhang1, Yonghong Dang1
1Department of Nuclear Medicine, State Key Laboratory of Complex Severe and Rare Diseases, Center for Rare Diseases Research, Beijing Key Laboratory of Molecular Targeted Diagnosis and Therapy in Nuclear Medicine, Peking Union Medical College Hospital, Chinese Academy of Medical Science and Peking Union Medical College, Beijing, 100730, China.
Background:
This study evaluates the feasibility of using a simplified Patlak parametric imaging technique with a scaled population-based input function (sPBIF) in pancreatic cancer.
Methods:
Twenty-six patients underwent multi-bed, multi-pass [18F]FDG PET/CT scans, from which both dynamic and static PET images were reconstructed. Patlak parametric images were generated from the dynamic PET series using both the image-derived input function (IDIF) and the sPBIF. The consistency between IDIF and sPBIF was evaluated by comparing the area under the curve (AUC) and Patlak parameters derived from both input functions. The detectability of pancreatic lesions, assessed by tumor-to-background ratio (TBR) and contrast-to-noise ratio (CNR), was compared between SUV and Patlak parametric images. Additionally, the correlation between clinicopathological features and PET parameters, including SUV and Patlak values, was analyzed.
Results:
We found good agreement between the AUC for IDIF and sPBIF with correlation coefficients of 0.87 and 0.93 for the 0-30 min and 0-50 min intervals, respectively. The Patlak parameters from IDIF and sPBIF presented correlation coefficients higher than 0.94. The SUV and Patlak Ki exhibited correlation coefficients greater than 0.92 and 0.73 in malignant and benign pancreatic lesions, respectively. The SUV and Patlak V0 correlated with correlation coefficients higher than 0.75 in benign lesions, but exhibited only a weak correlation in malignant lesions. The TBR of Patlak Ki was significantly higher in malignant lesions compared to SUV. However, the CNR of Patlak Ki was lower due to increased noise in the parametric images. Most clinicopathological features showed weak correlation with PET parameters, except for a marginal classification of lesion differentiation by the maximum Ki value.
Conclusions:
The sPBIF approach enables the acquisition of additional Patlak parametric images alongside static SUV imaging in pancreatic cancer patients. Ki parametric imaging provided higher contrast than static imaging for detecting pancreatic lesions.
