Acute Imaging in Stroke— Correlating Imaging Data to Clinical Observation
(funded by NIH)

Intracerebral hemorrhage, while accounting for only 10-15% of stroke in the U.S., has high morbidity and mortality rates, and has no approved therapies. Our studies aim to understand the relationship between CT images obtained (see digitally enhanced stroke area) in the ER with bedside clinical assessment. Towards this aim, we have examined 282 consecutive head CTs obtained emergently in our level 1 trauma center, from patients with primary intracerebral hemorrhage, and are now examining how we can relate features of stroke size and location to clinical outcome.

Informatics Needs

• Software infrastructure to efficiently store and retrieve radiological images.
• Software to effectively quantify CT image analysis tools and compare with other outcome-predictive data obtained from patient history and physical examination.
• Grid enabled image and metadata aggregation and query.
• Collaborative tools to enable social networking among heterogeneous researchers (physicians, nurses, physical therapists), open and secure access to information.
• Systems support for data query, visualization and analysis.