Clinical Imaging Research Centre, Yong Loo Lin School of Medicine, National University of Singapore, Singapore.
Clinical Imaging Research Centre, Yong Loo Lin School of Medicine, National University of Singapore, Singapore; Joint Department of Medical Imaging, University Health Network, Toronto, Canada; Siemens Medical Solutions USA, Inc., Molecular Imaging, Knoxville, TN, USA.
Phys Med. 2021 Jan;81:285-294. doi: 10.1016/j.ejmp.2020.11.027. Epub 2020 Dec 16.
To conduct a simplified lesion-detection task of a low-dose (LD) PET-CT protocol for frequent lung screening using 30% of the effective PETCT dose and to investigate the feasibility of increasing clinical value of low-statistics scans using machine learning.
We acquired 33 SD PET images, of which 13 had actual LD (ALD) PET, and simulated LD (SLD) PET images at seven different count levels from the SD PET scans. We employed image quality transfer (IQT), a machine learning algorithm that performs patch-regression to map parameters from low-quality to high-quality images. At each count level, patches extracted from 23 pairs of SD/SLD PET images were used to train three IQT models - global linear, single tree, and random forest regressions with cubic patch sizes of 3 and 5 voxels. The models were then used to estimate SD images from LD images at each count level for 10 unseen subjects. Lesion-detection task was carried out on matched lesion-present and lesion-absent images.
LD PET-CT protocol yielded lesion detectability with sensitivity of 0.98 and specificity of 1. Random forest algorithm with cubic patch size of 5 allowed further 11.7% reduction in the effective PETCT dose without compromising lesion detectability, but underestimated SUV by 30%.
LD PET-CT protocol was validated for lesion detection using ALD PET scans. Substantial image quality improvement or additional dose reduction while preserving clinical values can be achieved using machine learning methods though SUV quantification may be biased and adjustment of our research protocol is required for clinical use.
使用有效 PETCT 剂量的 30%进行简化的低剂量（LD）PET-CT 协议下的肺部频繁筛查的病灶检测任务，并研究使用机器学习增加低统计扫描临床价值的可行性。
我们从 SD PET 扫描中获取了 33 张 SD PET 图像，其中 13 张是实际 LD（ALD）PET，以及 7 个不同计数水平的模拟 LD（SLD）PET 图像。我们采用图像质量转移（IQT），这是一种机器学习算法，可以执行补丁回归，将低质量图像的参数映射到高质量图像。在每个计数水平上，从 23 对 SD/SLD PET 图像中提取的补丁用于训练三个 IQT 模型——全局线性、单棵树和随机森林回归，补丁大小为 3 和 5 个体素。然后，使用这些模型根据每个计数水平的 10 个未见受试者的 LD 图像来估计 SD 图像。在匹配的病灶存在和病灶不存在的图像上进行病灶检测任务。
LD PET-CT 协议在灵敏度为 0.98 和特异性为 1 的情况下产生了病灶可检测性。使用 5 个立方体素大小的随机森林算法可以在不影响病灶可检测性的情况下进一步降低 11.7%的有效 PETCT 剂量，但 SUV 低估了 30%。
使用 ALD PET 扫描验证了 LD PET-CT 协议的病灶检测性能。虽然 SUV 定量可能存在偏差，需要调整我们的研究方案以用于临床，但可以使用机器学习方法实现图像质量的显著改善或额外的剂量减少，同时保持临床价值。
