KOCHEN Sara Silvia
Diagnostic Performance of MRI Volumetry in Epilepsy Patients With Hippocampal Sclerosis Supported Through a Random Forest Automatic Classification Algorithm
PRINCICH, JUAN PABLO; DONNELLY-KEHOE, PATRICIO ANDRES; VALLEJO-AZAR, MARIANA NAHIR; PASCARIELLO, GUIDO ORLANDO; SEOANE, PABLO; VERON DO SANTOS, JOSE GABRIEL; COLLAVINI, SANTIAGO; NASIMBERA, ALEJANDRO HUGO; KOCHEN, SILVIA
Frontiers in Neurology
Frontier in Neurology
Año: 2021 vol. 12
Introduction: Several methods offer free volumetry services for MR data that adequatelyquantify volume differences in the hippocampus and its subregions. These methodsare frequently used to assist in clinical diagnosis of suspected hippocampal sclerosisin temporal lobe epilepsy. A strong association between severity of histopathologicalanomalies and hippocampal volumes was reported using MR volumetry with a higherdiagnostic yield than visual examination alone. Interpretation of volumetry results ischallenging due to inherent methodological differences and to the reported variabilityof hippocampal volume. Furthermore, normal morphometric differences are recognizedin diverse populations that may need consideration. To address this concern, wehighlighted procedural discrepancies including atlas definition and computation of totalintracranial volume that may impact volumetry results. We aimed to quantify diagnosticperformance and to propose reference values for hippocampal volume from twowell-established techniques: FreeSurfer v.06 and volBrain-HIPS.Methods: Volumetry measures were calculated using clinical T1 MRI from alocal population of 61 healthy controls and 57 epilepsy patients with confirmedunilateral hippocampal sclerosis. We further validated the results by a state-of-the-artmachine learning classification algorithm (Random Forest) computing accuracy andfeature relevance to distinguish between patients and controls. This validationprocess was performed using the FreeSurfer dataset alone, considering morphometricvalues not only from the hippocampus but also from additional non-hippocampalbrain regions that could be potentially relevant for group classification. Mean reference values and 95% confidence intervals were calculated for left and righthippocampi along with hippocampal asymmetry degree to test diagnostic accuracy.Results: Both methods showed excellent classification performance (AUC:> 0.914)with noticeable differences in absolute (cm3) and normalized volumes. Hippocampalasymmetry was themost accurate discriminator fromall estimates (AUC:1∼0.97). Similarresults were achieved in the validation test with an automatic classifier (AUC:>0.960),disclosing hippocampal structures as the most relevant features for group differentiationamong other brain regions.Conclusion: We calculated reference volumetry values from two commonly usedmethods to accurately identify patients with temporal epilepsy and hippocampalsclerosis. Validation with an automatic classifier confirmed the principal role of thehippocampus and its subregions for diagnosis.