Intracerebral haemorrhage (ICH) accounts for approximately 15% of all strokes in Denmark and remains associated with high mortality and morbidity. It is challenging to distinguish neoplastic from non-neoplastic causes of ICH in the acute setting, and CT findings that may aid early differentiation have not been fully characterised. Existing ICH-classification systems (SMASH-U, H-ATOMIC and CLAS-ICH) have not been directly compared for diagnostic accuracy in this setting. Identifying radiological and clinical factors associated with underlying aetiology may support faster diagnosis, reduce time to workup related to potential underlying cancer and facilitate early targeted treatment of the underlying cause of ICH.

This study is a retrospective observational cohort including all patients admitted with acute ICH to the Department of Neurology, University Hospital of Southern Denmark, Aabenraa between January 2014 and December 2024 (estimated approximately n=610). Medical records and initial non-enhanced CT scans will be reviewed. Two neurologists and two radiologists, blinded to final diagnosis, will independently extract clinical presentation, topographical and volumetric haemorrhage characteristics, and classify each case using the abovementioned ICH-classification systems. Primary analyses will assess associations between clinical and radiological features and underlying neoplastic vs non-neoplastic aetiology. Secondary analyses will compare diagnostic performance of classification systems using sensitivity, specificity and receiver operating characteristic curves. Multivariate logistic regression models will be applied with Holm correction for multiple comparisons.

The study has been submitted to the National Danish Research Ethics Committee and the Danish Data Protection Agency. As data derive from completed disease courses, no patient contact is expected. Results will be disseminated through peer-reviewed journals, conferences and scientific presentations.