Linescan

Settings

cbv calculation parameters dialog

cbv method from linescan

Select method to use to calculate velocity from the linescan data.

Spatial correlation

Default setting. Correlation of grey level pattern along the length of the measurement line.

Hough Transform (fast flows)

Experimental. Uses Hough transform to try and detect lines in the linescan data and calculate the the velocity from the angle of the detected lines. Works best on faster velocities.

Temporal correlation

Experimental. Requires two measurement lines positioned at two points on the same vessel. First measurement line should be upstream. Plots the average grey level from each line and calculates the correlation over the time difference between the two traces.

No. of correlations to smooth

Default 4. Averages this number of correlations together before looking for peak value to reduce spurious false readings.

No. of linescan sample periods between correlations

The time interval between the two linescan samples used in the correlation. Sample period is one field by default. (i.e. 1/50 or 1/60 second). Increasing this value will lower the maximum velocity that can be measured, but will increase the resolution of the velocity result.

For very slow flows, a long gap might give poor correlation if the blood cell grey level pattern changes as it moves along the line.

Best to always use a multiple of 2 in case the fields are captured in the wrong order or when using non-interlaced video such as from the IEEE1394 or USB cameras.

Default 2.

Correlation coefficient lower acceptable limit

Correlation peak values below this setting are ignored and cbv is set to an “invalid” value so is not included in average. You will probably want to lower this value when using “subtract scans” since the correlations tend to be a bit poorer.

Default 0.2

No. of points for temporal correlations

Experimental. Period of grey level trace to use in the correlation. Needs to be short for pulsatile flows.

Default 10

Set sample rate

For very slow flows the sample rate at which linescans are recorded can be lowered. Not recommended.

Default 1

Subract scans

If the image has some fixed pattern or marks over the vessel being measured, it may give a false zero flow since it will correlate best on the fixed pattern. Setting subtract scans will remove the stationery pattern from the linescan but tends to give lower correlations. You might also want to lower the correlation limit.

Default off.