Infographics / Image Quality

Quantum mottle, radiographic noise & SNR vs dose

A rising, concave curve of signal-to-noise ratio against dose (proportional to photon fluence N). Because signal is proportional to N and quantum (Poisson) noise is proportional to the square root of N, SNR is proportional to the square root of N, so it rises with the square root of dose. Two marked doses, d and 2d, show that doubling the dose multiplies SNR by the square root of two, not by two. A side panel gives the SNR and contrast-to-noise relationships: CNR equals contrast divided by noise, so low-contrast tasks are noise-limited while high-contrast tasks are less limited by noise.

Signal-to-noise ratio rises as the square root of dose A rising, concave curve of signal-to-noise ratio against dose (proportional to photon fluence N). Because signal is proportional to N and quantum (Poisson) noise is proportional to the square root of N, SNR is proportional to the square root of N, so it rises with the square root of dose. Two marked doses, d and 2d, show that doubling the dose multiplies SNR by the square root of two, not by two. A side panel gives the SNR and contrast-to-noise relationships: CNR equals contrast divided by noise, so low-contrast tasks are noise-limited while high-contrast tasks are less limited by noise. Signal-to-noise ratio rises as √(dose) signal-to-noise ratio (SNR) dose ∝ photon fluence N (∝ mAs) SNR ∝ √N SNR ×√2 SNR d 2d why √N, not linear signal ∝ N (detected photons) noise σ ∝ √N (Poisson) SNR = N / √N = √N ↑ dose → ↑ SNR, less mottle double dose → SNR ×√2, not ×2 contrast-to-noise CNR = ΔS / σ low contrast → noise-limited high contrast → less noise-limited

Practice the Physics paper

FRCRX is a Part 1 Physics question bank: true/false items with explanations, timed mocks and topic analytics.