Kirsten L Boedeker MS, Presenter: Nothing to Disclose

Virgil N Cooper PhD, Abstract Co-Author: Nothing to Disclose

Michael F. McNitt-Gray PhD, Abstract Co-Author: Nothing to Disclose

Noise equivalent quanta (NEQ), as defined by Wagner et al (Med Phys 1979), reflects the effective number of photons reaching the detector during CT image acquisition and is independent of reconstruction filter and other post-processing. The purpose of this study is to show the utility of NEQ in comparing quantum noise associated with a given CT protocol.

Under constant mAs (quanta) conditions, a 16cm diameter water phantom was scanned on a multislice CT scanner (Sensation 16, Siemens Medical Systems). Baseline images were acquired with 33 eff mAs, 120 kVp, 16 x 0.75 mm collimation, pitch of 1, 2mm thickness, and B30f filter. The stdev of a central ROI was calculated for reference. The raw data was reconstructed using each filter (B10f to B80f). Noise Power Spectra (NPS) and stdev were calculated for each filter. NEQ were obtained for each filter per Wagner et al as the slope of the low-frequency, linear portion of the NPS. Images with constant stdev across all filters were then created, varying the mAs for each filter to obtain the reference stdev. The required mAs was compared to that predicted based on the reference standard deviation. The NEQ was obtained for each new filter and mAs combination.

For the constant mAs (quanta) conditions, NEQ measurements yielded the same value, regardless of filter, unless aliasing was present in the NPS. For the constant stdev condition, NEQ varied in a linear relationship with changes in mAs. However, the stdev did not vary linearly with the square root of the mAs for the sharper filters.

NEQ is independent of filter, reflects solely on the quantum noise associated with image acquisition and varies directly with mAs (i.e. dose). Stdev does not change in a predictable fashion; likely due to post process smoothing transparent to the user. This can have strong implications for protocol comparability between models of scanners, allowing for more accurate consideration of tradeoffs in dose and image quality. The use of NEQ in conjunction with stdev (or NPS) is a more accurate method of comparing image noise and image quality between scanners than stdev alone.

Boedeker, K, Cooper, V, McNitt-Gray, M, Advantages of the Use of NEQ in Comparing CT Imaging Protocols.  Radiological Society of North America 2004 Scientific Assembly and Annual Meeting, November 28 - December 3, 2004 ,Chicago IL. http://archive.rsna.org/2004/4414860.html