Tranulis C, Guéguen B, Pham-Scottez A, Vacheron M N, Cabelguen G, Costantini A, Valero G, Galinovski A
Service Hospitalo-Universitaire de Santé Mentale et de Thérapeutique Pr H. Loo, Pr J.P. Olié, Centre Hospitalière Sainte-Anne, 1, rue Cabanis, Paris 75014, France.
Neurophysiol Clin. 2006 Jan-Feb;36(1):1-7. doi: 10.1016/j.neucli.2006.01.005. Epub 2006 Feb 2.
Motor threshold (MT) is an important parameter for the practice of transcranial magnetic stimulation. Our goal was to compare three methods to estimate MT in a clinical setting.
Comparison of three MT estimation algorithms: 1) the Rossini-Rothwell method consists in lowering stimulus intensity until only five positive responses out of 10 trials are recorded, defining MT; 2) the Mills-Nithi method considers the MT as the mean of an upper threshold (10 positive out of 10 trials) and a lower threshold (0 out of 10 trials); 3) the supervised parametric method estimates the MT by fitting (mathematically and graphically) a sigmoid function on raw data obtained by stimulation at variable intensities. Six MT estimations (two per method) were recorded in a single session in 10 healthy subjects.
The within-subject variation of MT (expressed as % of the mean MT+/-standard deviation) during a single session was of 8.5+/-7.2% for the Rossini-Rothwell method, 8.7+/-5.7% for the Mills-Nithi method and 9.5+/-4.0% for the supervised parametric method. No significant differences in variability of MT estimation were found between the methods, but the Rossini-Rothwell method was significantly shorter (half the number of stimuli compared to the two other methods).
In our setting, Rossini-Rothwell method was superior to the two other methods. The variability of MT estimation measured in our study is important, yet acceptable for clinical applications. However, this variability can be a source of considerable errors in excitability studies and should be a focus of future research.
运动阈值（MT）是经颅磁刺激实践中的一个重要参数。我们的目标是在临床环境中比较三种估计MT的方法。
比较三种MT估计算法：1）罗西尼 - 罗斯韦尔方法是降低刺激强度，直到在10次试验中仅记录到5次阳性反应，以此定义MT；2）米尔斯 - 尼蒂方法将MT视为上限阈值（10次试验全部为阳性）和下限阈值（10次试验全部为阴性）的平均值；3）监督参数方法通过对在可变强度刺激下获得的原始数据拟合（数学和图形方式）一个S形函数来估计MT。在10名健康受试者的单次实验中记录了六种MT估计值（每种方法两个）。
在单次实验中，罗西尼 - 罗斯韦尔方法的MT受试者内变异（以平均MT的百分比±标准差表示）为8.5±7.2%，米尔斯 - 尼蒂方法为8.7±5.7%，监督参数方法为9.5±4.0%。各方法之间在MT估计变异性方面未发现显著差异，但罗西尼 - 罗斯韦尔方法明显更短（与其他两种方法相比，刺激次数减半）。
在我们的实验环境中，罗西尼 - 罗斯韦尔方法优于其他两种方法。我们研究中测量的MT估计变异性是重要的，但在临床应用中是可接受的。然而，这种变异性在兴奋性研究中可能是相当大误差的来源，应成为未来研究的重点。
