OBJECTIVES:Refractory epilepsies (RE), as well as, the surgically correctable syndromes, are of great interest, since they affect the very young population of children and adolescents. The early diagnosis and treatment are very important in preventing the psychosocial disability. Therefore MRI and EEG are highly sensitive methods in the diagnosis and localization of epileptogenic focus, but also in pre-surgical evaluation of these patients. The aim of our study is to correlate the imaging findings of EEG, MRI and CT scan in refractory symptomatic epilepsies, and to determine their specificity in detecting the epileptogenic focus.METHODS:The study was prospective with duration of over two years, open-labelled, and involved a group of 37 patients that had been evaluated and diagnosed as refractory epilepsy patients. In the evaluation the type and frequency of seizures were considered, together with the etiologic factors and their association, and finally the risk for developing refractory epilepsy was weighted. EEG and MRI findings and CT scan results were evaluated for their specificity and sensitivity in detecting the epileptogenic focus, and the correlation between them was analyzed.RESULTS:Regarding the type of seizures considered in our study, the patients with PCS (partial complex seizures) dominated, as opposed to those with generalized seizures (GS) (D=1.178, p < 0.05). Positive MRI findings were registered in 28 patients (75.7%). Most of them were patients with hippocampal sclerosis, 12 (42.8%), and also they were found to have the highest risk of developing refractory epilepsy (RE) (Odds ratio = 5.7), and the highest association between the etiologic factor and refractory epilepsy (p < 0.01). In detecting the epileptogenic focus, a significant difference was found (p < 0.01) between MRI and CT scan findings, especially in patients with hippocampal sclerosis and cerebral malformations. There was a strong correlation between the MRI findings and the etiologic factor (R = 1), and for CT scan and etiologic factor an R=0.75 correlation. There was a significant difference between imaging methods MRI/CT (p < 0.1), and CT/EEG (p < 0.05) in detecting the etiologic factor, and little difference was noticed between findings of EEG/MRI.CONCLUSION:Our study confirms that for an accurate diagnosis of refractory epilepsy in patients, a combination of neuroimaging and neurophysiologic methods is required. MRI showed to be highly sensitive in detecting the etiologic factor in RE patients, whereas EEG was sensitive in localization of the epileptogenic focus, with high correlation between these two methods. An early diagnosis of these patients is very important in having a better therapeutic response and prognosis for them.