RT - Journal Article T1 - Assessment of Degree of Agreement between Automated and Manual Volumetry for Quantification of Hippocampal Volume on MRI Images of Epileptic Patients JF - sjimu YR - 2015 JO - sjimu VO - 23 IS - 4 UR - http://sjimu.medilam.ac.ir/article-1-2512-en.html SP - 279 EP - 293 K1 - Magnetic resonance imaging K1 - Manual volumetry K1 - Automated volumetry K1 - Hippocampus AB - Introduction: localizing of the epileptogenic zone is a key step in the preoperative assessment of patients with medically refractory medial temporal lobe epilepsy (mTLE). Quantitative measurement of hippocampal volume (HV) using a standard method of manual tracing is very time consuming, difficult and operator-dependent that is why several automated hippocampal volumetric methods are used. The purpose of this study was to assess the degree of agreement between the automated hippocampal volumetric methods and measurements obtained from standard method (manually) in mTLE patients. Materials & methods: This study was performed in10 patients with mTLE, and 10 healthy controls participated. For each case, the right and left hippocampal volumes were measured using manual tracing method and three automated hippocampal volumetric methods include FSL, FreeSurfer and atlas-based. Degree of agreement between these methods was analyzed using statistical tests of the Pearson correlation coefficient and the Intra-class Correlation Coefficient (ICC). Findings: The Pearson correlation coefficients of HV between the manual tracing and FreeSurfer methods were 0.80(right) and 0.88(left), between the manual tracing and FSL methods were 0.80(right) and 0.79(left), and between the manual tracing and atlas-based methods were 0.72(right) and 0.66(left). The ICCs of HV between the manual tracing and FreeSurfer methods were 0.80(right) and 0.87(left), between the manual tracing and FSL methods were 0.77(right) and 0.78(left), and between the manual tracing and atlas-based methods were 0.70(right) and 0.66(left). Discussion & Conclusion: Findings of this study showed that automated methods could be valuable tools for the evaluation of epileptic patients and preoperative assessment. LA eng UL http://sjimu.medilam.ac.ir/article-1-2512-en.html M3 ER -