Encrypted login | home

Program Information

Comparison and Evaluation of Commercially Available CMOS X-Ray Detectors for Neurological Endovascular Image Guided Interventions (neuro-EIGI's)

B Loughran

B Loughran*, A Jain, V Singh, S Setlur Nagesh, C Ionita, D Bednarek, S Rudin, Univ. at Buffalo (SUNY) School of Med., Buffalo, NY

TU-A-116-7 Tuesday 8:00AM - 9:55AM Room: 116

Purpose: To evaluate three commercially available CMOS x-ray detector systems as potential dual fluoroscopic and angiographic, high-resolution, ROI detectors for neuro-endovascular image-guided interventions (neuro-EIGIs) and to compare them to other experimentally evaluated or simulated neuro-EIGI detectors.

Methods: Each CMOS detector was exposed to a standard RQA5 x-ray spectrum obtained by adding 21 mm of aluminum to a 70 kVp x-ray beam. Flat-field images were acquired by each detector and exposure per frame was determined using an ion chamber. From these images and exposures, sensitivity in digital numbers (DN/μR) and instrumentation noise-equivalent exposure (INEE) was found for each detector. These values were used to compare the three commercial detectors to other neuro-EIGI detectors previously built or simulated.

Results: Detector A has a 12-bit ADC, a pixel size of 100 μm, and a full well capacity (FWC) of 2.2
million electrons (Me-). The sensitivity was found to be 5.8 DN/μR and the INEE was found to be 7.58 μR. Detector B has a 14-bit ADC, a pixel size of 75 μm, and a FWC of 0.36 Me- for the high sensitivity mode and 1.4 Me- for the high saturation mode. The sensitivity was found to be 28.9 DN/μR and the INEE was found to be 2.04 μR in the high sensitivity mode. Detector C has a 14-bit ADC, a pixel size of 100 μm and a FWC of 1.5 Me-. The sensitivity was found to be 15.1 DN/μR and the INEE was found to be 3.04 μR. The previous neuro-EIGI detectors had similar sensitivities of 10s-100s DN/μR, but much lower INEEs of <0.2 μR.

Conclusion: While the commercial CMOS x-ray detectors are improving, it remains a desirable goal to further reduce the commercial detector pixel size and instrumentation noise to approach the values seen with the previously built and simulated neuro-EIGI detectors.

Funding Support, Disclosures, and Conflict of Interest: Supported by NIH Grant 2R01EB002873 and an equipment grant from Toshiba Corporation.

Contact Email: