Combining deep learning with a kinetic model to predict dynamic PET images and generate parametric images
Abstract Background Dynamic positron emission tomography (PET) images are useful in clinical practice because they can be used to calculate the metabolic parameters (K i ) of tissues using graphical methods (such as Patlak plots). K i is more stable than the standard uptake value and has a good refe...
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
SpringerOpen
2023-10-01
|
| Series: | EJNMMI Physics |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s40658-023-00579-y |
| _version_ | 1827708252232613888 |
|---|---|
| author | Ganglin Liang Jinpeng Zhou Zixiang Chen Liwen Wan Xieraili Wumener Yarong Zhang Dong Liang Ying Liang Zhanli Hu |
| author_facet | Ganglin Liang Jinpeng Zhou Zixiang Chen Liwen Wan Xieraili Wumener Yarong Zhang Dong Liang Ying Liang Zhanli Hu |
| author_sort | Ganglin Liang |
| collection | DOAJ |
| description | Abstract Background Dynamic positron emission tomography (PET) images are useful in clinical practice because they can be used to calculate the metabolic parameters (K i ) of tissues using graphical methods (such as Patlak plots). K i is more stable than the standard uptake value and has a good reference value for clinical diagnosis. However, the long scanning time required for obtaining dynamic PET images, usually an hour, makes this method less useful in some ways. There is a tradeoff between the scan durations and the signal-to-noise ratios (SNRs) of K i images. The purpose of our study is to obtain approximately the same image as that produced by scanning for one hour in just half an hour, improving the SNRs of images obtained by scanning for 30 min and reducing the necessary 1-h scanning time for acquiring dynamic PET images. Methods In this paper, we use U-Net as a feature extractor to obtain feature vectors with a priori knowledge about the image structure of interest and then utilize a parameter generator to obtain five parameters for a two-tissue, three-compartment model and generate a time activity curve (TAC), which will become close to the original 1-h TAC through training. The above-generated dynamic PET image finally obtains the K i parameter image. Results A quantitative analysis showed that the network-generated K i parameter maps improved the structural similarity index measure and peak SNR by averages of 2.27% and 7.04%, respectively, and decreased the root mean square error (RMSE) by 16.3% compared to those generated with a scan time of 30 min. Conclusions The proposed method is feasible, and satisfactory PET quantification accuracy can be achieved using the proposed deep learning method. Further clinical validation is needed before implementing this approach in routine clinical applications. |
| first_indexed | 2024-03-10T17:01:02Z |
| format | Article |
| id | doaj.art-9a52ce03ae6444ac8a06f00770e040f6 |
| institution | Directory Open Access Journal |
| issn | 2197-7364 |
| language | English |
| last_indexed | 2024-03-10T17:01:02Z |
| publishDate | 2023-10-01 |
| publisher | SpringerOpen |
| record_format | Article |
| series | EJNMMI Physics |
| spelling | doaj.art-9a52ce03ae6444ac8a06f00770e040f62023-11-20T10:58:07ZengSpringerOpenEJNMMI Physics2197-73642023-10-0110111510.1186/s40658-023-00579-yCombining deep learning with a kinetic model to predict dynamic PET images and generate parametric imagesGanglin Liang0Jinpeng Zhou1Zixiang Chen2Liwen Wan3Xieraili Wumener4Yarong Zhang5Dong Liang6Ying Liang7Zhanli Hu8Lauterbur Research Center for Biomedical Imaging, Shenzhen Institute of Advanced Technology, Chinese Academy of SciencesDepartment of Nuclear Medicine, National Cancer Center, National Clinical Research Center for Cancer, Cancer Hospital & Shenzhen Hospital, Chinese Academy of Medical Sciences and Peking Union Medical CollegeLauterbur Research Center for Biomedical Imaging, Shenzhen Institute of Advanced Technology, Chinese Academy of SciencesLauterbur Research Center for Biomedical Imaging, Shenzhen Institute of Advanced Technology, Chinese Academy of SciencesDepartment of Nuclear Medicine, National Cancer Center, National Clinical Research Center for Cancer, Cancer Hospital & Shenzhen Hospital, Chinese Academy of Medical Sciences and Peking Union Medical CollegeDepartment of Nuclear Medicine, National Cancer Center, National Clinical Research Center for Cancer, Cancer Hospital & Shenzhen Hospital, Chinese Academy of Medical Sciences and Peking Union Medical CollegeLauterbur Research Center for Biomedical Imaging, Shenzhen Institute of Advanced Technology, Chinese Academy of SciencesDepartment of Nuclear Medicine, National Cancer Center, National Clinical Research Center for Cancer, Cancer Hospital & Shenzhen Hospital, Chinese Academy of Medical Sciences and Peking Union Medical CollegeLauterbur Research Center for Biomedical Imaging, Shenzhen Institute of Advanced Technology, Chinese Academy of SciencesAbstract Background Dynamic positron emission tomography (PET) images are useful in clinical practice because they can be used to calculate the metabolic parameters (K i ) of tissues using graphical methods (such as Patlak plots). K i is more stable than the standard uptake value and has a good reference value for clinical diagnosis. However, the long scanning time required for obtaining dynamic PET images, usually an hour, makes this method less useful in some ways. There is a tradeoff between the scan durations and the signal-to-noise ratios (SNRs) of K i images. The purpose of our study is to obtain approximately the same image as that produced by scanning for one hour in just half an hour, improving the SNRs of images obtained by scanning for 30 min and reducing the necessary 1-h scanning time for acquiring dynamic PET images. Methods In this paper, we use U-Net as a feature extractor to obtain feature vectors with a priori knowledge about the image structure of interest and then utilize a parameter generator to obtain five parameters for a two-tissue, three-compartment model and generate a time activity curve (TAC), which will become close to the original 1-h TAC through training. The above-generated dynamic PET image finally obtains the K i parameter image. Results A quantitative analysis showed that the network-generated K i parameter maps improved the structural similarity index measure and peak SNR by averages of 2.27% and 7.04%, respectively, and decreased the root mean square error (RMSE) by 16.3% compared to those generated with a scan time of 30 min. Conclusions The proposed method is feasible, and satisfactory PET quantification accuracy can be achieved using the proposed deep learning method. Further clinical validation is needed before implementing this approach in routine clinical applications.https://doi.org/10.1186/s40658-023-00579-yParametric imagingImage generationDeep learningKinetic modelDynamic PET images |
| spellingShingle | Ganglin Liang Jinpeng Zhou Zixiang Chen Liwen Wan Xieraili Wumener Yarong Zhang Dong Liang Ying Liang Zhanli Hu Combining deep learning with a kinetic model to predict dynamic PET images and generate parametric images EJNMMI Physics Parametric imaging Image generation Deep learning Kinetic model Dynamic PET images |
| title | Combining deep learning with a kinetic model to predict dynamic PET images and generate parametric images |
| title_full | Combining deep learning with a kinetic model to predict dynamic PET images and generate parametric images |
| title_fullStr | Combining deep learning with a kinetic model to predict dynamic PET images and generate parametric images |
| title_full_unstemmed | Combining deep learning with a kinetic model to predict dynamic PET images and generate parametric images |
| title_short | Combining deep learning with a kinetic model to predict dynamic PET images and generate parametric images |
| title_sort | combining deep learning with a kinetic model to predict dynamic pet images and generate parametric images |
| topic | Parametric imaging Image generation Deep learning Kinetic model Dynamic PET images |
| url | https://doi.org/10.1186/s40658-023-00579-y |
| work_keys_str_mv | AT ganglinliang combiningdeeplearningwithakineticmodeltopredictdynamicpetimagesandgenerateparametricimages AT jinpengzhou combiningdeeplearningwithakineticmodeltopredictdynamicpetimagesandgenerateparametricimages AT zixiangchen combiningdeeplearningwithakineticmodeltopredictdynamicpetimagesandgenerateparametricimages AT liwenwan combiningdeeplearningwithakineticmodeltopredictdynamicpetimagesandgenerateparametricimages AT xierailiwumener combiningdeeplearningwithakineticmodeltopredictdynamicpetimagesandgenerateparametricimages AT yarongzhang combiningdeeplearningwithakineticmodeltopredictdynamicpetimagesandgenerateparametricimages AT dongliang combiningdeeplearningwithakineticmodeltopredictdynamicpetimagesandgenerateparametricimages AT yingliang combiningdeeplearningwithakineticmodeltopredictdynamicpetimagesandgenerateparametricimages AT zhanlihu combiningdeeplearningwithakineticmodeltopredictdynamicpetimagesandgenerateparametricimages |