Background: Intraventricular hemorrhage (IVH), a predictor of a poor prognosis and high mortality, typically requires external ventricular drains (EVD). However, there is no method to measure the rate of ventricular blood clearance. Blood in the ventricles increases cerebral spinal uid (CSF) viscosity and decreases the diffusion coe cient. We hypothesize that change in DTI metrics, fractional anisotropy (FA), and mean diffusivity (MD), can be used as a marker for ventricular blood clearance.Method: 26 ICH patients with IVH were imaged on an MRI at 1, 14, 28, and 42 days. Ventricular blood volume (VBV), FA, and MD were calculated. Contralateral ventricular CSF served as control. Clinical assessment (NIHSS), and baseline Glasgow coma score (GCS) were associated with FA, MD, and VBV.Generalized linear mixed models assessed changes.Results: Average hematoma volume (HV) and NIHSS at day 1 was 14.6±16.7 cc and 16±8. HV and VBV resolution rates per day were 2.1% and 1.3%, respectively. Ipsilesional ventricular FA (vFA) signi cantly decreased (0.24 to 0.15, 1.3% per day, PP>99%), and MD increased (MD=1540*10 -6 to 2516*10 -6 mm 2 /sec, 1.5% per day, PP>99%) over 6 weeks. Patients with EVD exhibited a faster decline in ipsilesional vFA (1.5% vs. 1.1% per day) and an increase in MD (1.8% vs. 1.5% per day) compared with non-EVD patients. Ipsilesional vFA associated with VBV; a 1.00 cc increase of blood resulted in a 5.2% decrease in ventricular CSF diffusion coe cient. VBV was strongly associated with NIHSS (PP= 97-99%).A larger drained CSF volume was associated with a greater decrease (PP=83.4%) in FA; a smaller drained volume showed greater increases (PP=94.8%) in MD. Slower restoration of vFA and MD suggests obstruction.
Conclusion:We developed a non-invasive marker that shows ventricular FA and MD are strongly associated with HV, VBV, and can be used to quantify the rate of ventricular blood clearance.
