Clinicians and neuroradiologists can differentiate multiple sclerosis (MS) from neuromyelitis optica spectrum disorder (NMOSD) and MOG antibody-associated disorder (MOGAD) by the location of demyelinated lesions on brain magnetic resonance imaging (MRI), according to findings published in the Journal of Clinical Neuroscience.

Researchers at the New York University Grossman School of Medicine in New York City analyzed the locations of demyelinated brain lesions in 82 patients diagnosed with either MS (n=51), NMOSD (n=23), or MOGAD (n=8). They required the presence of at least 3 brain lesions on MRI scans for each patient.

Two board-certified neuroradiologists and one neuroradiology fellow reviewed the MRIs blinded to the clinical history of the patients. They recorded the locations of lesions larger than 3 mm in length and visible on at least 2 consecutive scans. Using this data, the researchers determined the frequency of lesions in specific locations, comparing the location of the lesions for each patient with their confirmed diagnosis.

Continue Reading

They found that patients with MS were more likely than patients with a different disease to develop lesions in 3 locations—the periventricular white matter in Dawson’s finger configuration (P =.009), the cerebellar hemisphere (P =.02), and the anterior temporal horn (P <.0001). Statistical analysis revealed that these locations differentiated MS from NMOSD and MOGAD with 87.1% sensitivity and 72.5% specificity.

Read more about NMOSD differential diagnosis

The researchers then performed a validation test with 97 new patients, 51 with MS, 25 with NMOSD, and 21 with MOGAD. They confirmed that patients with MS more frequently developed lesions in those same 3 locations—Dawson’s finger (P <.0001), the cerebellar hemisphere (P =.02), and the anterior temporal horn (P <.0001). When comparing the first test with the second test, the validation test was 76.3% accurate in differentiating MS from NMOSD and MOGAD, with a 23.7% misclassification rate.

“Our study results can be used as a check and an adjunct to neuroradiologists’ gestalt review of cases and for improving the diagnostic accuracy of radiologists-in-training,” the authors suggested.

“Our model may be especially helpful in evaluating patients with suspected neuroinflammatory disorders and multiple brain lesions on MRI,” they added. “Presence of lesions in MS-typical locations would increase the probability of MS diagnosis, and absence of lesions in such locations may trigger evaluation for the less common disorders.”


Patel J, Pires A, Derman A, et al. Development and validation of a simple and practical method for differentiating MS from other neuroinflammatory disorders based on lesion distribution on brain MRI. J Clin Neurosci. 2022;101:32-36. doi:10.1016/j.jocn.2022.04.035.