The long axial field-of-view (AFOV) PET scanners with ultra-high sensitivity provide new opportunities for enhanced parametric imaging but suffer from the dramatically increased volume and complexity of dynamic data. This study reconstructed a high-quality direct Patlak Ki image from five frames sinograms without input function by a deep learning framework based on DeepPET to explore the potential of artificial intelligence reducing the acquisition time and the dependence of input function in parametric imaging. MethodsThis study is implemented on a large AFOV PET/CT scanner (Biograph Vision Quadra) and twenty patients were recruited with 18 F-Fluorodeoxyglucose ( 18 F-FDG) dynamic scans. During training and testing of the proposed deep learning framework, the last five frames (25 min, 40-65 min post-injection) sinograms were set as input and the reconstructed Patlak Ki images by a nested EM algorithm on the vendor were set as ground truth. To evaluate the image quality of predicted Ki images, mean square error (MSE), peak signal-to-noise ratio (PSNR), and structural similarity index measure (SSIM) were calculated. Meanwhile, a linear regression process was applied between predicted and true Ki means on avid malignant lesions and tumor volume of interests (VOIs). ResultsIn testing phase, the proposed method achieved excellent MSE of less than 0.003, high SSIM, and PSNR of ~0.98 and ~38 dB, respectively. Moreover, there had a high correlation (DeepPET: = 0.7525, self-attention DeepPET: =0.8382) between predicted Ki and traditionally reconstructed Patlak Ki means over eleven lesions. ConclusionsThe results show that the deep learning based method produced high-quality parametric images from small frames of projection data without input function. It has much potential to address the dilemma that the long scan time and dependency on input function that still hampers the clinical translation of dynamic PET.

Deep learning is an emerging reconstruction method for positron emission tomography (PET) that can tackle complex PET corrections in an integrated procedure. This study optimized the direct PET reconstruction from sinogram on a long axial field of view (LAFOV) PET. Methods:This study developed a new deep learning architecture to reduce the biases during direct reconstruction from sinograms to images. This architecture is based on an encoder-decoder network and perceptual loss is adopted with pre-trained convolutional layers. It is trained and tested on data of 80 patients acquired from recent Siemens Biograph Vision Quadra PET/CT. The patients were randomly split into a training dataset of 60 patients, the validation dataset of 10 patients, and the test data set of 10 patients. The 3D sinograms were converted into 2D sinogram slices and were used as input to the network, and the vendor reconstructed images were considered as ground truths.The proposed method was compared with DeepPET, a benchmark deep learning method for PET reconstruction. Results:Compared to the DeepPET, the proposed network significantly reduced the root-mean-squared error (rRMSE) from 0.63 to 0.6 (p<0.01), and the structural similarity index (SSIM) and peak signal-tonoise ratio (PSNR) were improved from 0.93 to 0.95(p<0.01) and from 82.02 to 82.36(p<0.01), respectively. The reconstruction time was approximately 10s per patient, which was shortened by 36 times compared with the reconstruction using the conventional method. The errors of average standardized uptake values (SUV) for lesions between ground truth and the predicted result was reduced from 33.5% to 18.7% (P=0.03), and the error of max SUV was reduced from 32.7% to 21.8% (P=0.02). Conclusion:The results demonstrated the feasibility of using deep learning to reconstruct images with acceptable image quality and short reconstruction time. It showed that the proposed method can improve the quality of deep learning-based reconstructed images without additional CT images for attenuation and scattering corrections. This learning-based approach may provide the potential to accommodate more complex corrections for LAFOV PET.