Introduction:
Intracerebral hemorrhages (ICHs) accounts for approximately 15% of all strokes but carry high rates of morbidity and mortality. The location and volume of hematoma are strongly associated with outcomes. As novel treatments become established, early detection and proper volume measurement are becoming increasingly important. We aim to evaluate an artificial intelligence-based algorithm (Viz-ICH® v1.4) for ICH detection, volume measurement, and its differentiation from intraventricular hemorrhage (IVH).
Methods:
We performed single center retrospective analysis of non-contrast CTs (NCCTs), randomly picked from prospective cohort of acute stroke patients, with and without parenchymal ICHs, admitted to our stroke center from 02/14-03/17. Experienced stroke neurologists graded NCCTs with a semi-automated tool (OsiriX MD v.9.0.1) for presence and volume of ICH, and also presence of intraventricular hemorrhage (IVH). AI- and human-based readings were compared.
Results:
A total of 211 NCCTs were evaluated including 163 ICHs and 48 controls. The ICH location was basal ganglionic in 55.8%, Lobar in 23.3% and posterior fossa in 12% of the cases and 51.5 % of patients had associated IVH with mean volume 41.94 cc. The AI algorithm demonstrated high accuracy for ICH detection and volumetric measurement (Table). The maximal running time of the algorithm was under 15s.
Conclusion:
The presence and volume of ICHs can be accurately detected by AI Viz-ICH Algorithm, with good differentiation from IVH which will help in early triage of these patients.