Small Animal Veterinary Hospital Stommeln, D 50259, Pulheim, Germany.
Clinical Unit of Diagnostic Imaging, University of Veterinary Medicine, A 1210, Vienna, Austria.
BMC Vet Res. 2016 Aug 11;12(1):165. doi: 10.1186/s12917-016-0788-3.
Hippocampal necrosis in cats has been reported to be associated with epileptic seizures. Magnetic resonance imaging (MRI) features of temporal lobe (TL) abnormalities in epileptic cats have been described but MR images from epileptic and non-epileptic individuals have not yet been systematically compared. TL abnormalities are highly variable in shape, size and signal, and therefore may lead to varying evaluations by different specialists. The aim of this study was to investigate whether there were differences in the appearance of the TL between epileptic and non-epileptic cats, and whether there were any relationships between TL abnormalities and seizure semiologies or other clinical findings. We also investigated interobserver agreement among three specialists.
The MR images of 46 cats were reviewed independently by three observers, who were blinded to patient data, examination findings and the review of the other observers. Images were evaluated using a multiparametric scoring system developed for this study. Mann-Whitney U-tests and chi-square were used to analyse the differences between observers' evaluations. The kappa coefficient (k) and Fleiss' kappa coefficient were used to quantify interobserver agreement.
The overall interobserver agreement was moderate to good (k =0.405 to 0.615). The MR scores between epileptic and non-epileptic cats did not differ significantly. However, there was a significant difference between the MR scores of epileptic cats with and without orofacial involvement according to all three observers. Likewise, MR scores of cats with cluster seizures were higher than those of cats without clusters.
Cats presenting with recurrent epileptic seizures with orofacial involvement are more likely to have hippocampal pathologies, which suggests that TL abnormalities are not merely unspecific epileptic findings, but are associated with a certain type of epilepsy. TL signal alterations are more likely to be detected on FLAIR sequences. In contrast to severe changes in the TL which were described similarly among specialists, mild TL abnormalities may be difficult to interpret, thus leading to different assessments among observers.
已有报道称,猫的海马坏死与癫痫发作有关。已描述了癫痫猫颞叶(TL)异常的磁共振成像(MRI)特征,但尚未对癫痫和非癫痫个体的 MRI 图像进行系统比较。TL 异常在形状、大小和信号方面具有高度可变性,因此可能导致不同专家的评估结果不同。本研究旨在探讨癫痫和非癫痫猫之间 TL 外观是否存在差异,以及 TL 异常与癫痫发作半影或其他临床发现之间是否存在任何关系。我们还研究了三位专家之间的观察者间一致性。
三位观察者独立回顾了 46 只猫的 MRI 图像,他们对患者数据、检查结果以及其他观察者的回顾均不知情。使用为此项研究开发的多参数评分系统对图像进行评估。使用曼-惠特尼 U 检验和卡方检验分析观察者评估之间的差异。使用 Kappa 系数(k)和 Fleiss' Kappa 系数来量化观察者间一致性。
总体观察者间一致性为中度至良好(k =0.405 至 0.615)。癫痫和非癫痫猫的 MRI 评分无显著差异。然而,根据所有三位观察者的评估,有口腔面部受累的癫痫猫和无口腔面部受累的癫痫猫之间的 MRI 评分存在显著差异。同样,有簇状发作的猫的 MRI 评分高于无簇状发作的猫。
反复出现伴有口腔面部受累的癫痫发作的猫更有可能出现海马病变,这表明 TL 异常不仅仅是特异性癫痫发现,而是与特定类型的癫痫有关。TL 信号改变更有可能在 FLAIR 序列上检测到。与专家描述的 TL 严重变化相似,轻度 TL 异常可能难以解释,从而导致观察者之间的评估结果不同。