Velocity Analysis

Velocity analysis estimates blood flow velocity from line scan kymographs using a Radon-transform-based method.

Input data

Expects a line-scan kymograph and a rectangular ROI covering the region used to estimate flow. Physical X/Y calibration must be correct (see AcqImage physical units).

Programmatic use

from acqstore.acq_image import AcqImage
from acqstore.acq_image.analysis import RadonVelocityAnalysis
from acqstore.sample_data import ensure_sample_file

acq = AcqImage(str(ensure_sample_file('kymograph-flow')))

channel = acq.images.channel_indices[0]
roi_ids = acq.rois.get_roi_ids()
roi_id = roi_ids[0] if roi_ids else acq.rois.create_rect_roi(name='velocity').roi_id

velocity = acq.analysis_set.create_and_run(
    RadonVelocityAnalysis,
    channel=channel,
    roi_id=roi_id,
    detection_params={'window_width': 64},
    replace_existing=True,
    execution_options={'use_multiprocessing': False},
)
print(velocity.result.summary)
acq.save()

Notebooks / macOS

Pass execution_options={'use_multiprocessing': False} inside Jupyter so Radon velocity runs serially.

Detection parameters

The primary parameter is the width of each Radon analysis window.

name	display_name	type	default	choices	unit	editable	visible	methods	description
window_width	Window Width	int	64	(16, 64, 128)		True	True		Number of time samples per Radon analysis window.

Results

For a source file my_file.tif:

my_file.tif.json
my_file.tif.radon_velocity.csv

Typical summary fields: velocity_mean, velocity_median, velocity_cv, num_windows. The CSV stores per-window tabular velocity results.

See the Velocity Analysis API and the Velocity Analysis notebook.