Pre-processing

Here you specify pre-processing settings related to the perfusion analysis: 

Noise threshold

• Auto-detect: when checked, the noise level is determined automatically. The detection is based on histogram analysis of the pixel intensity distribution in the images.
• Use noise level threshold: If checked image pixels below the noise threshold are excluded from the analysis. The selected noise level is shown in red on the input images. Pixels where the pixel intensity in the first image in the dynamic series is below the specified value will be excluded from further analysis.
• The noise level can be modified using the slider

Motion correction: Applies motion correction of the raw dynamic data. When the motion correction is completed, a graph is displayed showing the detected translation and rotation of the dynamic data. The data is automatically realigned according to the detected motion. Note: Since the dynamic signal change due to the contrast bolus can sometimes be hard to distinguish from motion, the input data should be investigated carefully after motion correction to ensure that the motion correction procedure has not introduced erroneous changes in the dynamic contrast agent response or other unexpected changes to the input data.

Spatial smoothing: applies Gaussian spatial smoothing with the specified full width half height (FWHH) kernel. This reduces image noise but also the effective spatial resolution.

Temporal smoothing: The dynamic time signal is smoothed (low-pass filtered), to reduce effects of noise and spikes in the dynamic signal response. This smoothing does not affect spatial resolution but may reduce the ability to detect rapid signal changes. If temporal smoothing is selected the degree of smoothing can be varied using the slider. The mean curve (or ROI curve if selected) will be updated to reflect the temporal smoothing factor selected.