Department of Mental Health, University of Aberdeen, UK.
Psychiatry Res. 2010 Nov 30;184(2):86-95. doi: 10.1016/j.pscychresns.2010.08.010. Epub 2010 Sep 28.
Psychiatric neuroimaging techniques are likely to improve understanding of the brain in health and disease, but studies tend to be small, based in one imaging centre and of unclear generalisability. Multicentre studies have great appeal but face problems if functional magnetic resonance imaging (fMRI) data from different centres are to be combined. Fourteen healthy volunteers had two brain scans on different days at three scanners. Considerable effort was first made to use similar scanning sequences and standardise task implementation across centres. The n-back cognitive task was used to investigate between- and within-scanner reproducibility and reliability. Both the functional imaging and behavioural results were in good accord with the existing literature. We found no significant differences in the activation/deactivation maps between scanners, or between repeat visits to the same scanners. Between- and within-scanner reproducibility and reliability was very similar. However, the smoothness of images from the scanners differed, suggesting that smoothness equalization might further reduce inter-scanner variability. Our results for the n-back task suggest it is possible to acquire fMRI data from different scanners which allows pooling across centres, when the same field strength scanners are used and scanning sequences and paradigm implementations are standardised.
精神神经影像学技术可能会增进对健康和疾病状态下大脑的理解，但这些研究往往规模较小，仅基于一个影像中心，且其普遍性也不清楚。多中心研究具有很大的吸引力，但如果要整合来自不同中心的功能磁共振成像（fMRI）数据，就会面临一些问题。14 名健康志愿者在三台扫描仪上于不同天进行了两次脑部扫描。首先，我们做出了相当大的努力，以在中心之间使用相似的扫描序列和标准化任务实施。n-back 认知任务用于研究扫描仪之间以及同一扫描仪的重复扫描之间的再现性和可靠性。功能成像和行为结果都与现有文献非常一致。我们没有发现扫描仪之间或同一扫描仪的重复扫描之间的激活/失活图谱有显著差异。扫描仪之间和内部的再现性和可靠性非常相似。然而，扫描仪的图像平滑度不同，这表明平滑度均衡可能会进一步降低扫描仪之间的变异性。我们对 n-back 任务的结果表明，当使用相同场强扫描仪并标准化扫描序列和范式实现时，从不同扫描仪获取 fMRI 数据并允许跨中心汇总是可能的。
