Demet KINAY, Hülya YETİŞ, Reha TOLUN, Candan GÜRSES, Betül BAYKAN, Ayşen GÖKYİĞİT

İ.Ü. İstanbul Tıp Fakültesi, Nöroloji Anabilim Dalı

Keywords: Hippocampal volumetry, Magnetic resonance imaging, Temporal lob epilepsy

Abstract

The purpose of this study was to establish normative values for MR imaging-based volumetric measurements of the hippocampal formations and to evaluate the accuracy and reproducibility of these measurements before applying this technique to the patients with medically refractory temporal lobe epilepsy who are surgical candidates.in our validation studies we estimated intra and interobserver reliability using a multipl person, test-retest paradigm . Far assessment of interobserver reliability, a set of mesurements obtained by one of us from 20 healthy volunteers was compared with the measurements obtained by another of us. Far assessment of test-retest reliability of the volum measurements, the 10 MR imagings were remeasured by one of us afrer at least 3 months . When the observers trained together and agreed on anatomic landmarks , interobserver differences was reduced. The average interobserver difference was 92 (SD=206.5) mm³ for right HF measurements and 92.7 (SD= 198.5) mm³ for left HF measurements . The average intraobserver differences was 38 (SD= 84 .7) mm³ for right HF measurements and 57 (SD= 107.4) mm³ for lefr HF measurements. These average error values were well under 1 SD of the normal control mean for right and left HF ( 1SD= 546 and 474 mm³ , respectively) volume. Interobserver and intraobserver correlation coefficients far left and right hippocampal volume ranged from 0.92 to 0.98. in healthy volunteers aged 20-38 years, the mean right and left hippocampal volumes were 4119 mm3 and 4039 mm³ , respectively. The mean hippocampal formation asymmetry index (L-R/L +R ratio) which was established to determine the degree of asymmetry between sides was -0.009 (SD=0.024). We have not found a signifıcant rightleft hippocampal volumetric discrepancy in normals. The results demonstrate that MR-based volumemeasurements can be made with high precision and reproducibility in experienced centers when using appropriate technical principles.