医学
磁共振成像
脑炎
自身免疫性脑炎
病理
放射科
免疫学
病毒
作者
Poornima Narayanan Nambiar,K Y Manisha,Jayakumari Nandana,Ramshekhar N. Menon,S. Vinayagamani,Bejoy Thomas,Ashalatha Radhakrishnan
标识
DOI:10.1177/02841851241307330
摘要
Background The role of imaging in autoimmune encephalitis (AIE) remains unclear, and there are limited data on the utility of magnetic resonance imaging (MRI) to diagnose, treat, or prognosticate AIE. Purpose To evaluate whether MRI is a diagnostic and prognostic marker for AIE and assess its efficacy in distinguishing between various AIE subtypes. Material and Methods We analyzed data from 96 AIE patients from our prospective autoimmune registry. MRI sequences examined were FLAIR, diffusion, SWI, T2WI, ASL, and contrast enhancement. Short-term outcomes were measured using the Modified Rankin Scale (mRS) at discharge; long-term outcomes were assessed with the Functional Independence Measure (FIM) at 6 months. Results MRI confirmed AIE in cases of new-onset seizures (82.1%, P < 0.001) and dementia (100%, P = 0.02). Antibody-negative AIE exhibited significant multifocal FLAIR abnormalities compared to antibody-positive cases ( P = 0.002). LGI1 and CASPR2 encephalitis frequently involved the mesial temporal region ( P = 0.004), while ASL revealed hyperperfusion of the contralateral basal ganglia in faciobrachial dystonic seizures ( P = 0.016). GAD65 encephalitis predominantly affected the cerebellum ( P = 0.002), and NMDA encephalitis showed contrast enhancement in five cases ( P = 0.045). MRI was not useful for predicting short-term outcomes but was associated with long-term outcomes; specifically, a normal MRI was linked to a better long-term outcome in 47.8% of patients ( P = 0.035), and resolution of abnormalities correlated with a favorable FIM score (>54) in 76.7% ( P = 0.016). Conclusion MRI is valuable for early detection of seizures or dementia as initial manifestations of AIE and for differentiating AIE subtypes. Follow-up MRI is significant in predicting long-term outcomes.
科研通智能强力驱动
Strongly Powered by AbleSci AI