Intracerebral Hematoma Volume Measurement Using ITK-SNAP Software: Initial Experience (P7.145)

Objective: To explore the feasibility of using a computerized software method (ITK-SNAP) to measure intracerebral hematoma volume via measurement of inter-observer reliability and comparison with the ABC/2 method of measurement. Background: Intracerebral hemorrhage (ICH) is a neurologic emergency with potentially high mortality. Initial hematoma volume has been shown to correlate with clinical outcome and is used in guiding acute treatment and in prognostication. Volumes greater than 60ml and less than 30ml have been associated with 30-day mortality rates of 90% and 17%, respectively. Due to this difference in outcomes and its impact on care, accurate and reliable measurement of the hematoma volume is critical. ABC/2 is a commonly used manual measurement method which has been reported to have an inter-observer reliability of 0.99. Manual measurement methods can be susceptible to human error. Method: Three non-contrasted head CT’s obtained within 24hrs of spontaneous ICH were analyzed. Four independent observers measured hematoma volumes using both ITK-SNAP and ABC/2 methods. Mean volumes were compared using Student’s t-tests. The inter-rater reliability coefficient was calculated for both methods after applying the Spearman-Brown correction. Analyses were done with SPSS software. Results: There was a 2.9% absolute difference in inter-observer reliability between the two measurement methods. Inter-observer reliability was higher in the ITK-SNAP group (0.998, p<0.001, 95% CI=0.99-1.0) compared to the ABC/2 group (0.969, p<0.001, 95% CI=0.79-0.99). There was no statistically significant difference in volumes between the two methods. Conclusion: Computerized ICH volume measurement using ITK-SNAP had improved reliability compared to manual ABC/2 method when assessed using inter-observer reliability statistics. Measurement methods such as ITK-SNAP should be considered for ICH volume measurement in order to minimize the potential for human error and improve measurement reliability. Reliability of volume measurement could also impact the ICH score which is commonly used to prognosticate and stratify risk in intracerebral hemorrhage. Disclosure: Dr. Mishra has nothing to disclose. Dr. Shukla has nothing to disclose. Dr. Bodhit has nothing to disclose. Dr. Asaithambi has nothing to disclose. Dr. Velez has nothing to disclose. Dr. Rosado has nothing to disclose. Dr. Mishra has nothing to disclose. Dr. Saravanapavan has nothing to disclose. Dr. Khan has nothing to disclose. Dr. Reddy has nothing to disclose. Dr. Hedna has nothing to disclose.