Ding Wenxiang, Wang Hanzhong, Qiao Xiaoya, Li Biao, Huang Qiu
School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, 200240, China.
Department of Nuclear Medicine, Ruijin Hospital, Shanghai Jiao Tong University, Shanghai, 200240, China.
Eur J Nucl Med Mol Imaging. 2025 Mar;52(4):1448-1459. doi: 10.1007/s00259-024-07012-1. Epub 2024 Dec 17.
Prolonged scanning durations are one of the primary barriers to the widespread clinical adoption of dynamic Positron Emission Tomography (PET). In this paper, we developed a deep learning algorithm that capable of predicting dynamic images from dual-time-window protocols, thereby shortening the scanning time.
This study includes 70 patients (mean age ± standard deviation, 53.61 ± 13.53 years; 32 males) diagnosed with pulmonary nodules or breast nodules between 2022 to 2024. Each patient underwent a 65-min dynamic total-body [F]FDG PET/CT scan. Acquisitions using early-stop protocols and dual-time-window protocols were simulated to reduce the scanning time. To predict the missing frames, we developed a bidirectional sequence-to-sequence model with attention mechanism (Bi-AT-Seq2Seq); and then compared the model with unidirectional or non-attentional models in terms of Mean Absolute Error (MAE), Bias, Peak Signal-to-Noise Ratio (PSNR), and Structural Similarity (SSIM) of predicted frames. Furthermore, we reported the comparison of concordance correlation coefficient (CCC) of the kinetic parameters between the proposed method and traditional methods.
The Bi-AT-Seq2Seq significantly outperform unidirectional or non-attentional models in terms of MAE, Bias, PSNR, and SSIM. Using a dual-time-window protocol, which includes a 10-min early scan followed by a 5-min late scan, improves the four metrics of predicted dynamic images by 37.31%, 36.24%, 7.10%, and 0.014% respectively, compared to the early-stop protocol with a 15-min acquisition. The CCCs of tumor' kinetic parameters estimated with recovered full time-activity-curves (TACs) is higher than those with abbreviated TACs.
The proposed algorithm can accurately generate a complete dynamic acquisition (65 min) from dual-time-window protocols (10 + 5 min).
长时间扫描是动态正电子发射断层扫描（PET）在临床广泛应用的主要障碍之一。在本文中，我们开发了一种深度学习算法，能够根据双时窗协议预测动态图像，从而缩短扫描时间。
本研究纳入了2022年至2024年间诊断为肺结节或乳腺结节的70例患者（平均年龄±标准差，53.61±13.53岁；32例男性）。每位患者均接受了65分钟的动态全身[F]FDG PET/CT扫描。模拟使用早期停止协议和双时窗协议进行采集以缩短扫描时间。为了预测缺失的帧，我们开发了一种带有注意力机制的双向序列到序列模型（Bi-AT-Seq2Seq）；然后在预测帧的平均绝对误差（MAE）、偏差、峰值信噪比（PSNR）和结构相似性（SSIM）方面，将该模型与单向或无注意力模型进行比较。此外，我们报告了所提出方法与传统方法之间动力学参数的一致性相关系数（CCC）的比较。
Bi-AT-Seq2Seq在MAE、偏差、PSNR和SSIM方面显著优于单向或无注意力模型。与15分钟采集的早期停止协议相比，使用双时窗协议（包括10分钟早期扫描和5分钟晚期扫描）分别将预测动态图像的四个指标提高了37.31%、36.24%、7.10%和0.014%。用恢复的全时活度曲线（TAC）估计的肿瘤动力学参数的CCC高于用缩短的TAC估计的CCC。
所提出的算法能够从双时窗协议（10 + 5分钟）准确生成完整的动态采集（65分钟）。
