Background: We aimed to describe the use and findings of cranial computerized tomography (CT-head), spine and brain magnetic resonance imaging (MRI-spine/MRI-brain) in Lyme neuroborreliose (LNB). 

Methods: Patients with LNB were identified using a nationwide, population-based prospective cohort of all adults treated for neuroinfections at departments of infectious diseases in Denmark from 2015 to 2019. Multivariate logistic regression analyses assessed associations between clinical characteristics and MRI-findings consistent with LNB. 

Results: We included 368 patients (272 definite LNB and 96 probable LNB), 280 scans were performed in 198 patients. Neuroimaging was associated with older age (59 vs. 57, p = 0.03), suspicion of other diseases (77% vs. 37%, p < 0.0001), no history of tick bites (58% vs. 43%, p = 0.01), physical/cognitive deficits prior to admission (15% vs 5%, p = 0.006), peripheral palsy (10% vs. 2%, p = 0.0008), encephalitis (8% vs. 1%, p = 0.0007) and cognitive impairment (8% vs. 2%, p = 0.03) compared with those without neuroimaging. Normal or incidental findings were common (93/98 CT-head and 154/182 MRI). 1/98 CT-head, 19/131 MRI-brain and 6/51 MRI-spine had findings consistent with LNB. Symptoms ≥45 days was associated with MRI-findings consistent with LNB (adjusted odds ratio (aOR) 4.2, 95%confidence interval 1.2–14.4, p = 0.02). 

Conclusion: In this Danish cohort including 368 LNB-patients, use of neuroimaging was common and often performed in older comorbid patients without previous tick-bite intended to investigate alternative diagnoses. The results were in general without pathology and neuroimaging cannot exclude LNB or replace lumbar puncture. MRI is of value when investigating alternative neurological diseases and may support suspicion of LNB in cases with meningeal/leptomeningeal/neural enhancement.