Abstract Background Real‐time volumetric perfusion assessment is highly desirable during various clinical interventional procedures to guide treatment, confirm safety, and assess the success. However, the current standard of care is impeded by the lack of real‐time intra‐operative perfusion imaging. Typically, only 2‐D X‐ray angiography (XA) imaging is available, so soft tissue perfusion cannot be assessed. Purpose To develop and validate a novel end‐to‐end method, I ntra‐operative PE rfusion assessment with N o gantry rotation (IPEN), for volumetric brain perfusion assessment for ischemic stroke patients using the standard XA images. Methods We propose IPEN, which takes a different yet mathematically rigorous approach to this 3‐D tomography problem. IPEN models the volume of interest as a collection of multiple small, homogeneous regions of interest (ROIs), estimates their contrast enhancement over time (called time–enhancement curves, TECs), and computes key perfusion indices such as Tmax (time–to–maximum) and rCBF (relative cerebral blood flow) in the ROIs. We developed a five‐step algorithm around IPEN and validated it using a digital perfusion phantom for ischemic stroke. We simulated twelve stroke cases and acquired intra‐arterially contrast‐enhanced biplane XA images. The accuracy of the TEC, Tmax, and rCBF was assessed using mean error (m.e.) ± standard deviation of the errors (s.d.e.), and correlation coefficients ( r ) against the true values. Results Applying IPEN to the XA images, we successfully estimated TEC, Tmax, and rCBF for 3‐D ROIs in the brain with high accuracy. Across all twelve cases, the statistical metrics of error were small. The m.e. was –1.0 HU for TEC, –0.1 s for Tmax, and 4.4% for rCBF, while the s.d.e. was 17.5 HU, 1.2 s, and 17.7%, respectively. We also observed high r values for all estimated values (0.88 for TEC, 0.67 for Tmax, and 0.81 for rCBF). Conclusions IPEN accurately and reliably estimated brain perfusion indices on a regional basis. This work demonstrated the potential of IPEN in assessing perfusion in real‐time, which could significantly improve procedural success and patient outcomes.
Subramanian et al. (Thu,) studied this question.