RAPID ASPECTS automatically generates a standardized score to assist clinicians who are assessing early signs of brain ischemia. RAPID ASPECTS tilt-corrects the images and identifies hyperdense regions. A classification system, based on validated machine learning algorithms, identifies regions of asymmetry. Hounsfield unit data for each region can be compared in the table adjacent to the images.
Using the slice up and down, zoom and windowing features, the clinician can evaluate regions of interest. Windowing is performed by clicking the WL/WW circles, which provide 2 default options (40/80 and 35/35). Alternatively, any desired window setting can be achieved by using the mouse (31/39 is shown above).
Manually adding or subtracting regions can adjust the score. This can be accomplished by either clicking directly on the brain regions, or the respective region on the table. You can return to the fully automated score by clicking the “restore ROIs” button, which appears if any manual adjustments are made.
With the assistance of the confidence indicator, the physician can scrutinize suspicious regions and determine a final score. For example, in this case the automated RAPID ASPECTS classifier identified 3 regions as potentially abnormal (caudate, internal capsule and lentiform). As shown above, moving the confidence slider to the left identified 2 additional regions that may potentially be involved (insula and M6). Moving the slider to the right removes regions with more subtle differences.
RAPID ASPECTS CE Marked 2017