期刊
MEDICAL PHYSICS
卷 48, 期 9, 页码 5059-5071出版社
WILEY
DOI: 10.1002/mp.15063
关键词
brain imaging; deep learning; dynamic imaging; PET; recurrent neural network
资金
- Swiss National Science Foundation [320030_176052, 31003A_179373]
- Swiss National Science Foundation (SNF) [31003A_179373, 320030_176052] Funding Source: Swiss National Science Foundation (SNF)
The study evaluated a recurrent frame generation algorithm for predicting late frames in dynamic brain PET imaging, showing that the developed deep learning approach can effectively generate the trend of time-varying tracer biodistribution and significantly reduce scanning time.
Purpose We assess the performance of a recurrent frame generation algorithm for prediction of late frames from initial frames in dynamic brain PET imaging. Methods Clinical dynamic F-18-DOPA brain PET/CT studies of 46 subjects with ten folds cross-validation were retrospectively employed. A novel stochastic adversarial video prediction model was implemented to predict the last 13 frames (25-90 minutes) from the initial 13 frames (0-25 minutes). The quantitative analysis of the predicted dynamic PET frames was performed for the test and validation dataset using established metrics. Results The predicted dynamic images demonstrated that the model is capable of predicting the trend of change in time-varying tracer biodistribution. The Bland-Altman plots reported the lowest tracer uptake bias (-0.04) for the putamen region and the smallest variance (95% CI: -0.38, +0.14) for the cerebellum. The region-wise Patlak graphical analysis in the caudate and putamen regions for eight subjects from the test and validation dataset showed that the average bias for Ki and distribution volume was 4.3%, 5.1% and 4.4%, 4.2%, (P-value <0.05), respectively. Conclusion We have developed a novel deep learning approach for fast dynamic brain PET imaging capable of generating the last 65 minutes time frames from the initial 25 minutes frames, thus enabling significant reduction in scanning time.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据