Please use this identifier to cite or link to this item:
|Title:||Hippocampal sclerosis can be reliably detected by magnetic resonance imaging.||Austin Authors:||Jackson, Graeme D ;Berkovic, Samuel F ;Tress, B M;Kalnins, Renate M;Fabinyi, Gavin C ;Bladin, Peter F||Affiliation:||Department of Neurology, Austin Hospital, Heidelberg, Victoria, Australia||Issue Date:||1-Dec-1990||Publication information:||Neurology; 40(12): 1869-75||Abstract:||Two independent blinded observers reported the preoperative MRIs in a series of 81 consecutive patients with intractable temporal lobe epilepsy who were undergoing temporal lobectomy. We then compared the nature and lateralization of the MRI abnormalities with the pathologic diagnosis and the side of lobectomy. The MRI criteria of hippocampal sclerosis were an increased T2-weighted signal and the signal's confinement to a unilaterally small hippocampus. Imaging was performed in coronal and axial planes, specially orientated along and perpendicular to the long axis of the hippocampal body. We found diagnostic MRI abnormalities in 25 of the 27 cases with pathologically proven hippocampal sclerosis (sensitivity 93%, specificity 86%). In addition, we detected all 13 foreign tissue lesions on MRI. Overall, we detected lateralized lesions on MRI that correctly predicted the side of the epileptogenic temporal lobe in 72 cases (89%), with 2 possible errors. A learning effect in appreciating the relatively subtle MRI changes of hippocampal sclerosis was apparent in our later cases, as shown by an improved correlation between the 2 observers. This study demonstrates that hippocampal sclerosis can be identified on MRI with a high degree of sensitivity and specificity.||URI:||http://ahro.austin.org.au/austinjspui/handle/1/11469||URL:||https://pubmed.ncbi.nlm.nih.gov/2247236||Type:||Journal Article||Subjects:||Brain Neoplasms.pathology
Magnetic Resonance Imaging.methods
|Appears in Collections:||Journal articles|
Show full item record
checked on Dec 6, 2022
Items in AHRO are protected by copyright, with all rights reserved, unless otherwise indicated.