Introduction:
In this study, we sought to develop a collateral flow mapping method based on MR or CT perfusion imaging and compare this method with the digital subtraction angiography (DSA) method.
Methods:
Ischemic tissue supplied by collateral flow is likely to have delayed tracer arrival but relatively normal flow. In this study, abnormal Tmax (> 6 sec), MTT (4 sec or greater than unaffected hemisphere) and CBF (<40% of unaffected hemisphere) were labeled with green, red and blue, respectively (Fig. A). Any perfusion patterns can then be represented by a combination of this integrated RGB maps (Integrated Collateral flow Maps, iCMaps). 24 patients were included in this study. DSA images were obtained from all patients, whereas MR and CT perfusion images were acquired from 15 and 9 patients, respectively, within 30 hours after MCA occlusion. iCMaps collateral flow was scored independently as follows. iCMaps without any perfusion abnormality was assigned “1”. iCMaps with white as the dominant perfusion pattern was assigned a “5”; while those with a dominant green and yellow pattern were assigned a “2” or a “3”, depending on the relative green and yellow volume. Finally, iCMaps with a white region surrounded with yellow and green was assigned a “4”. Examples of iCMaps scoring are shown in Fig. B. Using a DSA scoring method in the literature5, a score of 1-5 was given independently based on the degree of retrograde collateralization (5: poor collateralization).
Results:
Good correlation was found between the MR perfusion iCMaps and DSA (DSA=0.70*MR iCMaps+0.54, R=0.75). Moderate correlation was obtained between the CT perfusion iCMaps and DSA (DSA=0.51*CT iCMaps+1.9, R=0.56).
Conclusions:
Based on the different characteristics of Tmax, MTT and CBF maps, the proposed MR or CT perfusion based iCMaps method can provide DSA comparable collateral flow information.
