1 February 2002, vol. 8, no. 1, pp. 52-58(7)
Rovaris M; Holtmannspötter M; Rocca M A; Iannucci G; Codella M; Viti B; Campi A; Comi G; Yousry T A; Filippi M
 Neuroimaging Research Unit, Scientific Institute and University H San Raffaele, Milan, Italy
 Department of Radiology, Klinikum Grosshadern, Munich, Germany
 Department of Neuroradiology, Scientific Institute and University H San Raffaele, Milan, Italy
 Clinical Trials Unit, Department of Neuroscience, Scientific Institute and University H San Raffaele, Milan, Italy
This study was performed to assess how established diagnostic criteria for brain magnetic resonance imaging (MRI) interpretation in cases of suspected multiple sclerosis (MS) (Barkhof's criteria) would perform in the distinction of MS from other diseases and whether other MR techniques (cervical cord imaging and brain magnetization transfer imaging [MTI]), might help in the diagnostic work-up of these patients.
We retrospectively identified 64 MS and 59 non-MS patients.
The latter group included patients with systemic immune-mediated disorders (SID; n=44) and migraine (n=15).
All patients had undergone MRI scans of the brain (dual echo and MTI) and of the cervical cord (fast short-tau inversion recovery).
The number and location of brain T2-hyperintense lesions and the number and size of cervical cord lesions were assessed.
Brain images were also postprocessed to quantify the total lesion volumes (TLV) and to create histograms of magnetization transfer ratio (MTR) values from the whole of the brain tissue.
Barkhof's criteria correctly classified 108/123 patients, thus showing an accuracy of 87.8%. ''False negative'' MS patients were 13, while 2 patients with systemic lupus erythematosus (SLE) were considered as ''false positives''.
Using brain T2 TLV, nine of the ''false negative'' patients were correctly classified.
Correct classification of 10 MS patients and both the SLE patients was possible based upon the presence or absence of one cervical cord lesion.
Two MS patients with negative Barkhof's criteria and no cervical cord lesions were correctly classified based on their brain MTR values.
Overall, only one MS patient could not be correctly classified by any of the assessed MR quantities.
These preliminary data support a
more extensive use of cervical cord MRI and brain MTI to differentiate
between MS and other disorders in case of inconclusive findings on T2-weighted
MRI scans of the brain.
© 2002 ingenta