PerfusionTools workflow
Step 1 - Draw cardiac borders
The first step is to draw the epicardial and endocardial borders on a short axis time frame that exhibits good contrast between blood and myocardium. It is necessary to draw these contours on both a rest and a stress image.
Step 2 - Correct motion
An automatic motion correction algorithm is provided to remove registration errors brought about by respiratory motion. This algorithm attempts to align the LV myocardium throughout the rest and stress image series.
Step 3 - Check motion
The third step allows the user to review the results of the automatic motion correction algorithm and to correct any inaccuracies that remain. It should be noted that only in-plane registration errors can be compensated for. If through-plane motion is evident, individual fames of the image series can be disabled.
Step 4 - Delineate blood pool region
The fourth step allows the user to delineate a circular region of the blood pool that does not contain papillary muscles or trabeculations. The average signal intensity of this region is used to characterise the flow of the blood into the ventricle.
Step 5 - Segment the myocardium
In order to provide a regional assessment of the left ventricle, the orientation of the ventricle within the short axis images must be identified. This is performed by manually identifying the LV/RV junction points.
Step 6 - Specify timing
PerfusionTools allows the user to observe the timing parameters for the rest and stress acquisitions. To ensure compatibility with a wide range of research systems, timing information can be imported from a text based file.
Step 7 - Assess segments
PerfusionTools provides a user interface for observing the signal characteristics of the myocardium and the blood pool. The user can click on an individual segment to observe the traces for a particular region of myocardium. At this step, it is possible to further improve the delineation of the cardiac borders by identifying abnormal variations in the signal traces.
Step 8 - Perfusion Analysis 
PerfusionTools uses Fermi Deconvolution to fit a smooth curve through the myocardium signal traces. From this curve, the software is able to derive a range of parameters that characterise relative myocardial perfusion including myocardial perfusion reserve index.