目的： 探讨一体化¹⁸F-FDG PET/MRI（以下简化为PET/MRI）检查在癫痫精准定位中的应用价值。方法： 入组25例药物难治型癫痫患者,在术前发作间期行颅脑一体化PET/MRI检查,并在检查完成1个月内通过立体定向脑电图（stereot-actic electroencephalography, SEEG）或手术病理检查明确致癫痫灶。以SEEG及手术病理方法为金标准,对单一MRI形态学、PET形态分析法、PET/MRI融合成像3种方法检出、定位致痫灶的灵敏度和特异度进行对照研究;通过MI Neurology软件对致癫痫灶与正常PET/MRI脑代谢数据库进行配准比对,获得各脑区的标准差（standard deviation, SD）值,对比病变部位与对侧正常对照脑区SD值、平均标准化摄取值（mean standardized uptake value, SUVmean）值的差异。结果： 单一MRI形态学方法定位致癫痫灶的灵敏度为37.5%（9/24）,特异度为100%（4/4）,有15处致痫灶单一MRI形态学无法显示,其中14例病灶无结构异常,因而MRI无法显示。单一PET形态学方法确诊16处癫痫灶。结合PET（包括半定量PET）进行PET/MRI融合成像,可以检出12处MRI阴性病例,PET/MRI检查定位致癲痫灶的灵敏度为91.6%（22/24）,明显高于单一MRI形态学（P<0.05）。PET/MRI融合成像减少了1例PET假阳性诊断,其成像特异度为100%,同时MRI可以清晰显示解剖结构,为手术提供精准定位。致癫痫灶SD值为-6.16±2.26,健侧SD值为-0.72±0.89,致癫痫灶的SD值明显低于健侧对照区（P<0.01）。结论： 一体化PET/MRI检查充分融合了2种显像方法的优势,对于无结构异常的致痫灶,结合PET（包括定量PET）可以发现MRI检查阴性的病灶,而MRI分辨率高,可清晰显示解剖结构,PET/MRI同机可对病灶的部位和范围进行精准定位,并可以进行功能融合成像,为手术方案制定提供了有力信息,在癫痫个体化精准医疗中显示出巨大应用潜力。

Objective: To analyze the clinical value of simultaneous ¹⁸F-FDG PET/MR imaging(PET/MRI) in precise localization of epileptogenic lesion. Methods: Twenty-five patients with drug-refractory epilepsy were enrolled.All patients underwent simultaneous PET/MR during the interval of epilepsy. The epileptogenic foci were identified by stereo-tactic electroencephalography (SEEG) or surgical pathology in all patients within one month after PET/MRI. The sensitivity and specificity of three imaging modalities including single MRI, single PET and PET/MRI in localizing epileptogenic foci were compared. The standard deviation (SD) values of each brain region were assessed by matching the epileptogenic foci with the normal PET/MR brain metabolic database by MI Neurology software. The differences of SD values and mean standardized uptake value mean (SUVmean) between the lesions and the oppositesite brain regions were compared. Results: The sensitivity and specificity of single MRI in locating epileptogenic focus were 37.5% (9/24) and 100% (4/4), respectively. There were 15 epileptogenic focus which could not be displayed by single MR, among which there were fourteen focus with no structure abnormal. PET/MRI can detected twelve of these fourteen lesions and the sensitivity of PET/MRI in localizing epileptogenic foci was 91.6% (22/24), which was significantly higher than that of single MRI (P< 0.05). The specificity of PET/MRI was 100%. The combination of PET and MRI reduced false-positive diagnosis. At the same time, MRI could provide a very clear anatomical structure and help foraccurate localization. The SD value of epileptogenic foci was -6.16+2.26 and that of opposite healthy side was -0.72+0.89. The SD value of epileptogenic lesion was significantly lower than that of control area (P<0.01). Conclusions: Simultaneous PET/MRI fully combines the advantages of two imaging methods. For the nonstructural abnormal epileptogenic foci, combined MRI with PET (including quantitative PET) can detect the MRI-negative foci, and with the high resolution and clear defining of anatomical structure, MRI combined with PET can accurately evaluate the location and scope of the foci on the same machine, which provides great help to design surgical therapy plan and denotes great application prospect in the individualized and precise medical treatment of epilepsy.