First page Back Continue Last page Overview Graphics
Conclusion
A general power curve fit can be applied to most data of noise against resolution
A better correlation is sometimes found when certain algorithms are disregarded
The mean value of the MTF 50% and 10% gives consistently high correlation for all scanners
The theoretical 1.5 power does not apply to empirical data on modern scanners when looking at dependence with algorithm
Power values in the range 2.4 – 2.8 are obtained for head and body scanning on four and sixteen slice scanners
When utilising the generalised relationship of noise and resolution deviation of specific algorithms must be considered