ABSTRACT
PURPOSE: This paper reports the first results obtained using a novel technology called eye controlled region of interest (ECR) that substantially reduces both staff and patient irradiation during an interventional fluoroscopy procedure without interfering with workflow. Its collimator includes a partially x-ray attenuating plate with a nonattenuating aperture. An eye tracker follows the operator's gaze to automatically position the aperture to the clinical region of interest (CROI) anywhere in the image in real-time. METHODS: Experiments were performed in a swine model using a mobile fluoroscope with a 30 cm image intensifier and manual control of fluoroscopic factors. The factory collimator and image display monitor were replaced with different components for this study. The full 30 cm field-of-view (FOV) of the image intensifier was irradiated at normal levels, and served as a baseline, when ECR was disengaged. With ECR engaged, most of the 30 cm FOV was irradiated to less than 20% of normal levels while the CROI was normally irradiated. Animal irradiation was determined by physical KAP (kerma area product) measurements. Operator irradiation was characterized by air kerma and air kerma rate measurements near the operator. Data were collected from three pairs of interventions in each of five swine models. RESULTS: When ECR was engaged, KAP was reduced to 0.22 (p < 0.001) of baseline and operator irradiation to 0.27 (p < 0.001) of baseline. Overall procedure time had a borderline increase (p = 0.07) but fluoroscopy time was unchanged (p = 0.36) (Wilcoxon signed rank). Measured staff and patient radiation reductions are consistent with this collimator's design. Subjective impressions of imaging improvements are consistent with less scatter reaching the CROI. Engaging ECR reduced irradiation without subjectively or objectively increasing operator workload. CONCLUSIONS: The first in vivo evaluation of ECR demonstrated that this technology has objectively reduced KAP and operator irradiation by approximately 75% without interfering with the performance of fluoroscopically guided interventional procedures. In addition, reduced scatter production subjectively improved device visualization. These findings indicate the practicability of achieving better radiation optimization.
