Common functional imaging approaches such as cerebral blood flow-weighted arterial spin labeling and cerebrovascular reactivity-weighted blood oxygenation level-dependent MRI are susceptible to quantification errors when applied to patients with significant arterial steno-occlusive disease, due to artifacts that result from delayed blood arrival and arteriolar rigidity. Recently it was suggested that the artifacts from standard quantitation approaches can be exploited together with machine learning algorithms to localize regions of hemodynamic impairment as defined by gold standard catheter angiography. Here, we investigate whether similar algorithms can be applied to identify spatial regions that progress to infarction in a longitudinal study.
This abstract and the presentation materials are available to members only; a login is required.