Complex, unstable inflow jet has been linked to aneurysm growth and rupture. However, methodologies to characterize this inflow jet have not been well established. Our previous works (Le et al., J. Biomech. Engr., 2010 and Le et al., Annals Biomedical Eng., 2013) have shown a possible transition from the stable mode (cavity) to the unstable mode (vortex ring) of this jet. We have proposed the use of a non-dimensional index called Aneurysm Number to characterize this transition (Le et al., 2013). However, the quantification of such a transition is lacking. Currently, there have no efforts in quantifying unstable flows in intracranial aneurysms, which is essential in stratifying rupture risks. In this work, the aneurysmal geometries from three patients at Sanford Health, North Dakota are reconstructed from Magnetic Resonance Angiogram and Digital Subtraction Angiogram data. Using our in-house CFD code (Virtual Flow Simulator), high-resolution flow data is obtained via numerical simulation. We perform modal analysis of blood flow dynamics for these cases using Proper Orthogonal Decomposition. Our results show that there are up to five dominant modes in the flow arising from the interaction of the incoming jet and the aneurysm dome. The spatial distribution of these modes reflect the characteristics of the inflow jet and can be used to quantify flow unsteadiness. Future works will be needed to apply the same procedure for a larger population of patients to examine its relevance in clinical practice.