Program Information
Evaluation of Optimum Technical Factors in Abdominal Digital Radiography
A Scott*, Y Zhou , J Nute , C Lee , Cedars-Sinai Medical Center, Los Angeles, CA
Presentations
SU-H2-GePD-I-6 (Sunday, July 30, 2017) 3:30 PM - 4:00 PM Room: Imaging ePoster Lounge
Purpose: As part of optimizing manual techniques for digital radiography, we investigated which kVp values in abdominal imaging would maximize a metric balancing image quality and dose. If there were no kVp clearly maximized this metric, then we would recommend a mAs for each kVp resulting in constant CNR.
Methods: Our experimental setup used a Carestream Evolution x-ray unit with Carestream cesium iodide detector in the table bucky. An ACR RF accreditation plate was placed below either 5” or 8” of Lucite blocks. This plate has nine holes of increasing depth milled in aluminum to serve as contrast targets. For both phantoms, phototimed exposures were acquired from 70 to 100 kVp, in 5 kVp increments. The contrast and noise for each target were measured using ImageJ by placing an ROI over each target to measure signal and drawing an annulus ROI of similar area around the target to measure background. Our quality metric Q was the CNR of each target divided by the square root of the dose, normalized to the maximum value so that Q ranged from 0 to 1. We then looked for a maximum in Q. We also calculated a mAs-adjustment-factor for each kVp which was the square of the normalized CNR.
Results: We observed that our quality metric varied by less than 15% across all kVps. Both dose and CNR decreased from 70 to 100 kVp at approximately the same rate. The mAs-adjustment-factors ranged from 1.0 to 2.75.
Conclusion: Based on our findings, our institution has no reason to prefer a particular kVp setting in abdominal imaging, in regards to balancing CNR and dose. When applying mAs scaling factors to achieve consistent CNR at various kVps, the detector exposure for 80 kVp best matches our institutional target EI. Thus, 80 kVp is our recommendation.
Contact Email: