Article Text
PDF
Abstract
Background Quantifiable parameters that can capture the hemodynamic profile of an AVM are lacking, as current methods of AVM assessment are descriptive. Recently, we have established a novel computational method to analyze vascular pulse waves (depicted by peak contrast bolus) from cerebral angiograms paired with the cardiac signal obtained by a pulse oximeter, showing normal arteriovenous coupling in the human brain. Latency reflecting propagation of the pressure waves pushing the contrast was observed between arterial and venous sides. We sought to explore cerebral AVMs and characterize disturbed arteriovenous (AV) coherence based on phase-shift between the peak frequency seen on arterial peak and the venous side.
Materials and Methods Computational methods (figure 1) permitting the capture and analysis of cardiac frequency phenomena from angiograms in 4 patients with AVMs, before and after embolization were employed. Phase shift was calculated between the feeding artery relative to draining vein and remote non-AVM vein.
Results Average phase shift between feeding artery and draining vein was significantly smaller compared to feeding artery and remote vein (122 vs 64 degrees, p<0.05). After embolization, two cured AVMs showed significantly decreased AV coherence between feeding artery and draining vein (85 degrees pre-embolization vs 140 degrees post-embolization, figure 2) while other two where partial embolization of high risk features only was performed didn’t improve significantly and maintained unchanged, high AV coherence.
Discussion This study describes the method and potential role of wavelet angiography in the hemodynamic profiling of the AVMs. By calculating AV coherence, quantifying AV phase shift before and after embolization was explored and can pave the way for numerical assessment of the impact of intervention on AVM functional angioarchitecture. Future studies are needed to create AV coherence weighted-maps to explore hemodynamic compartmentalization, assessment of risk related to level of coherence and the impact of partial embolization on AVM hemodynamics.
Disclosures O. Doron: None. J. Vranic: None. C. Stapleton: None. R. Regenhardt: None. A. Patel: None. W. Butler: None.