Dept. of Non-linear Systems, Saratov State University, Saratov, Russian Federation.
J Neurosci Methods. 2010 Dec 15;194(1):172-8. doi: 10.1016/j.jneumeth.2010.09.017. Epub 2010 Oct 7.
The automatic real-time detection of spike-wave discharges (SWDs), the electroencephalographic hallmark of absence seizures, would provide a complementary tool for rapid interference with electrical deep brain stimulation in both patients and animal models. This paper describes a real-time detection algorithm for SWDs based on continuous wavelet analyses in rodents. It has been implemented in a commercially available data acquisition system and its performance experimentally verified. ECoG recordings lasting 5-8h from rats (n=8) of the WAG/Rij strain were analyzed using the real-time SWD detection system. The results indicate that the algorithm is able to detect SWDs within 1s with 100% sensitivity and with a precision of 96.6% for the number of SWDs. Similar results are achieved for 24-h ECoG recordings of two rats. The dependence of accuracy and speed of detection on program settings and attributes of ECoG are discussed. It is concluded that the wavelet based real-time detecting algorithm is well suited for automatic, real-time detection of SWDs in rodents.
棘波放电（SWD）的自动实时检测是一种补充手段，可以在患者和动物模型中快速干扰电深部脑刺激。本文描述了一种基于连续小波分析的啮齿动物棘波放电实时检测算法。该算法已在商业可用的数据采集系统中实现，并通过实验验证了其性能。使用实时 SWD 检测系统对 WAG/Rij 大鼠的 ECoG 记录（n=8）进行了 5-8 小时的分析。结果表明，该算法能够在 1s 内以 100%的灵敏度检测到 SWD，并且 SWD 的数量的精度为 96.6%。对两只大鼠的 24 小时 ECoG 记录也得到了类似的结果。讨论了准确性和检测速度对程序设置和 ECoG 属性的依赖性。结论是，基于小波的实时检测算法非常适合啮齿动物棘波放电的自动实时检测。
