Kangasmaa Tuija S, Kuikka Jyrki T, Vanninen Esko J, Mussalo Hanna M, Laitinen Tomi P, Sohlberg Antti O
Department of Radiation Therapy, Vaasa Central Hospital, Vaasa, Finland.
Nucl Med Commun. 2011 Nov;32(11):1040-5. doi: 10.1097/MNM.0b013e328349c765.
To test the potential of a new reconstruction algorithm with Monte Carlo-based scatter correction in half-time myocardial perfusion single-photon emission computed tomography (SPECT).
The mathematical four-dimensional NURBS-based Cardiac-Torso phantom and the SIMIND Monte Carlo simulation package were used to simulate full-time and half-time SPECT projection data. The data were reconstructed using the standard ordered subset expectation maximization-based algorithm and the new Monte Carlo-based algorithm. Defect contrast, myocardium versus ventricle contrast and resolution were calculated. In addition to the simulation studies, full-time and half-time SPECT projection data of 30 patients were reconstructed with the standard and the new method. The patient data were qualitatively evaluated by four nuclear medicine experts on a scale from 1 (poor quality) to 5 (high quality).
The new reconstruction method with half-time data produced higher contrast and better resolution in the simulations and also achieved higher qualitative scores in the patient study than the standard reconstruction with full-time data.
Half-time myocardial perfusion imaging using the new reconstruction algorithm with Monte Carlo-based scatter correction produced images with superior quality when compared with full-time imaging with standard reconstruction.
在半时相心肌灌注单光子发射计算机断层扫描（SPECT）中，测试一种基于蒙特卡罗散射校正的新重建算法的潜力。
使用基于数学四维非均匀有理B样条（NURBS）的心脏躯干模型和SIMIND蒙特卡罗模拟软件包来模拟全时相和半时相SPECT投影数据。使用基于标准有序子集期望最大化的算法和新的基于蒙特卡罗的算法对数据进行重建。计算缺损对比度、心肌与心室对比度以及分辨率。除了模拟研究外，还使用标准方法和新方法对30例患者的全时相和半时相SPECT投影数据进行重建。由四位核医学专家对患者数据进行定性评估，评分范围为1（质量差）至5（质量高）。
在模拟中，使用半时相数据的新重建方法产生了更高的对比度和更好的分辨率，并且在患者研究中也比使用全时相数据的标准重建获得了更高的定性评分。
与使用标准重建的全时相成像相比，使用基于蒙特卡罗散射校正的新重建算法的半时相心肌灌注成像产生的图像质量更优。
