A Stitch in Time Saves 1.9 Million: How Advanced Imaging Solutions in Neuroradiology can Improve Outcomes for Stroke Patients

Abstract

Stroke is the leading cause of disability in Canada and the third leading cause of death. Every year there are over 50,000 new strokes, with nearly 14,000 of cases resulting in death. For every minute delay in treating a stroke, the patient loses 1.9 million brain cells, and for each hour in which treatment does not occur, the brain loses as many neurons as it does in almost 3.6 years of normal aging. Thus, it is vital that the patient receives revascularization therapy as soon as possible. Despite this knowledge, there is still an inefficient use of time between stroke onset to revascularization. Advanced image solutions are urgently needed to shorten the door to treatment time, save more brain cells and improve patient outcomes. Currently, the stroke team relies on CT to detect blood, ischemia, occlusion or stenosis and assess brain tissue at risk. This step in the workflow can delay revascularization therapy by up to 60 minutes. Our multidisciplinary research team is working with a vendor engineering team to 1) improve the image quality of the Philips Allura c-arm cone-beam CT (XperCT) and 2) develop specialized stroke imaging software. The goal is to improve the imaging capabilities of the angiography suite so that it can provide the same essential imaging as standard CT, allowing patients to forgo a visit to CT and go straight to the angio-suite for treatment, shortening the door to puncture time by up to 60 minutes. Thus far, we have improved the image quality of cone-beam CT by utilizing a specialized aluminum filter and reconstruction algorithms designed to reduce bone-beam hardening effects. Our current stroke imaging protocol consists of the standard non-enhanced XperCT ‘mask’ image, followed by a unique dual-phase enhanced XperCT. The dual-phase scan consists of two modified XperCT acquisitions with the c-arm performing a propeller scan 220 degrees in each direction around the patient’s head during IV injection of a contrast bolus (80 cc). The scans are 10 seconds apart, and result in volumetric images of early and late brain perfusion phases. The specialized stroke analysis software then performs a subtraction of these two phases to highlight any brain regions of delayed filling. Using a quality assessment questionnaire, the unenhanced and perfusion XperCT scans acquired in the angio-suite were evaluated by an experienced radiologist and compared to unenhanced and perfusion images acquired in CT. Results demonstrated that the image quality improvements and new stroke imaging software can provide the necessary diagnostic information, in the angio-suite, immediately before revascularization therapy. This includes the detection or exclusion of blood, ischemia, occlusion or stenosis and tissue at risk. Limitations with this process do exist. For example, patient motion between the scans can interfere with the subtraction process. Additionally, although the cone-beam CT images are sufficient for diagnosis, the XperCT quality still has inferior contrast and spatial resolution compared to a standard CT perfusion. Currently, stroke patients must first visit the CT department for imaging to confirm a large vessel occlusion before going to the angiography suite for mechanical thromectomy treatment. Preliminary results from our efforts to improve cone-beam CT image quality and develop specialized stroke image analysis software demonstrate that diagnostic-quality imaging can be acquired using a c-arm in the angio-suite, allowing patients to bypass CT and come straight to the angio-suite for imaging and immediate treatment. Although image quality improvement work continues, this one-stop-stroke-shop approach promises to reduce diagnosis time, save more brain cells and improve patient outcomes.