Purpose Following installation of a fluoroscopic digital flat panel detector (FPD) in the neuroangiography suite, three patients who underwent repeated angiographic studies for subarachnoid hemorrhage had complications with epilation, with one patient receiving an estimated radiation dose of 12 Gray over the course of 2 months following multiple angiographic studies, including two interventional procedures. A study was undertaken to evaluate radiation dose of the new FPD unit and implement changes to reduce exposure and potential patient complications.

Methods Radiation dose measurements were made on two biplane fluoroscopic units, a FPD unit installed in 2008 and an image-intensifier unit (II) installed in 1997 with add-on filter on the frontal plane. The head of a soft tissue with internal skeleton phantom was used in conjunction with calibrated thermoluminescent dosimeters (TLD) to measure the radiation absorbed dose at ten locations on and within the phantom. For both units, independent measurements were performed for fluoroscopy and digital subtraction (DSA) modes in both the frontal and lateral planes. Measurements from the II unit were taken with and without the filter. Measurements were also taken a second time from the FPD unit after dose reduction settings were made available from the company engineers. Single sample student's t tests were run to determine if the difference was different from zero. This was done separately for the full and 50% reduced Philips settings.

Findings Initial study results demonstrated that there was a significantly higher dose rate for both fluoroscopy and DSA at each of the TLD positions when using the FPD unit at the initial factory setting compared to when using the II unit; this was true regardless of filter use in the II unit. In response to these higher radiation dosages and in collaboration with the company engineers, additional filtration was installed on the FPD unit and a palette of low dose protocols was provided: full dose, 50% and 25% reduced dose. Follow-up measurements revealed the 50% reduced dose setting to remain higher but more comparable to the II doses and with acceptable image quality for routine diagnostic angiography and the 25% reduced dose was found to be satisfactory for children and for interventions requiring prolonged fluoroscopy times or multiple repeat acquisitions.

Conclusion Flat panel detector technology for fluoroscopy is becoming increasingly popular despite a higher cost compared to conventional II systems. Due to major differences in the image acquisition chain, the image quality of FPD technology is more uniform, lacking spatial distortion and less susceptible to artifact common to image intensifiers. Nevertheless, it has become evident that the factory default settings for FPD may not have the proper balance of patient dose with regards to clinically acceptable image quality, which is a problem not only for FPD, but is a theme common to all digital x-ray technology. Neurointerventional procedures are associated with substantial higher radiation exposure due to biplanar imaging, longer procedures and need for repeated studies. Adjusting factory settings to reduce radiation output can reduce exposure without loss of image quality.

Competing interests None.