Author: Sagine Berry-Tony, Lasya Daggumati, James R Duncan, Melak Senay, Allan Thomas 👨🔬
Affiliation: University of Missouri-Kansas City School of Medicine, Mallinckrodt Institute of Radiology, Washington University School of Medicine, Washington University in St. Louis 🌍
Purpose: Most acquired images in FGIs are not permanently archived. In the context of modern computational prowess, novel improvements to FGI practice likely sit just under the surface of large-scale clinical datasets. The main objective of this work is to outline a methodology to collect, archive, process, and analyze images from all irradiation events in FGIs. A specific test application of computing event-by-event patient dosimetry estimates is also presented.
Methods: Image frames were parsed from video recordings of the live fluoroscopy monitor and then classified as an irradiation event (yes/no) and event type if yes. Analyzed image frames were aligned with RDSR data for each event, with representative image frames archived. Peak skin dose, organ doses, and effective dose were computed with NCIRF (v3.0) after aligning each irradiation event with the anatomy in a computational phantom. Data from eight clinical FGIs were analyzed: two TIPS procedures, a hepatic artery embolization, and five uterine artery embolizations. Dosimetry results were compared between both standard and patient-size-specific phantoms.
Results: Matching video image frames with recorded events in the RDSR is consistent and achievable with high accuracy (>99%). Aligning irradiated fields to NCIRF phantom anatomy is time-consuming and at times uncertain, with moderate C-arm angulation (±30 degrees) causing particular difficulty. The clinical FGIs yielded different numbers of clinically relevant segments, varying distributions of radiation event types (fluoroscopy vs stationary), and contributions to overall radiation usage and patient doses.
Conclusion: The strategy outlined offers analysis and visualization of all FGI events with respect to anatomy exposed to radiation, x-ray acquisition parameters, procedure segmentation, and the resulting dosimetry estimates. Hence, there is value in archiving all the images in FGIs. This test application shows the feasibility of large volume data acquisition and analysis, which should initiate many opportunities to improve FGI practice via large-scale, data-driven approaches.