Skip to content

Track Metrics

guijacquemet edited this page May 23, 2024 · 5 revisions

Track available in CellTracksColab

Track Metrics in CellTracksColab can either be computed in the provided notebooks or directly imported from the tracking software. These metrics are organized into an expandable and collapsible accordion menu grouped by the categories above. Each category can be individually expanded or collapsed, and all sections are closed by default. A "Select All" checkbox is provided for each category, allowing users to select or deselect all metrics within a category quickly.

Track Metrics computed in CellTracksColab

Does not support Track splitting.

Supports 3D tracking data.

Track Metrics (computed start to end)

The start to end approach calculates metrics over the entire length of the track, providing a comprehensive overview of the track's characteristics from beginning to end. This method is useful for understanding overall trends such as directionality or average speed over the entire track.

These metrics can be computed in the TrackMate, Custom or Viever notebooks 'Track Duration', 'Mean Speed', 'Median Speed', 'Max Speed', 'Min Speed', 'Speed Standard Deviation', 'Directionality', 'Total Distance Traveled', 'Spatial Coverage', 'Tortuosity', 'Total Turning Angle'

  • Track Duration: The total duration of the track.
  • Mean Speed: The average speed of the track.
  • Max Speed: The maximum speed recorded for the track.
  • Min Speed: The minimum speed recorded for the track.
  • Speed Standard Deviation: The standard deviation of the speeds recorded for the track.
  • Total Distance Traveled: The cumulative distance traveled by the track.

Directionality is a metric that provides a measure of the overall direction of a track, from its starting point to its ending point, relative to the total path length. It offers insight into the straightness of the path taken: a value of 1 indicates a straight path between the start and end points, and values approaching 0 indicate more circuitous paths.

Tortuosity provides a measure of how convoluted or twisted a path is. For a straight path, the tortuosity is 1, while it increases for more winding paths. It gives insight into the curvature and complexity of the path taken: a value of 1 indicates a straight path between the start and end points, and values greater than 1 indicate paths with more twists and turns.

Total Turning Angle This measure provides insight into the cumulative amount of turning along the path. A value of 0 indicates a straight path with no turning, and higher values indicate paths with more turning.

Spatial coverage provides insight into the spatial extent covered by the object's movement. Higher values indicate that the object has covered a larger area or volume during its movement. The spatial coverage represents the area (in 2D) or volume (in 3D) enclosed by the convex hull formed by the points in the track.

Metrics computed using rolling windows

The rolling window approach is particularly useful when comparing tracks of different lengths, especially when the metric is not normalized over time, such as the total distance traveled. By using rolling averages, you ensure that the comparisons account for variations in track length and provide a more consistent basis for analysis.

Choosing the Window Size Window Size: The window_size parameter determines the number of data points considered in each rolling calculation. A larger window size will smooth the data more, averaging out short-term variations and focusing on long-term trends. Conversely, a smaller window size will be more sensitive to short-term changes, capturing finer details of the movement. Selection Tips: The optimal window size depends on the nature of your data and the specific analysis goals. It also depends on the length of your tracks.

These metrics can be computed in the TrackMate, Custom or Viever notebooks

    'Mean Speed Rolling', 'Median Speed Rolling', 'Max Speed Rolling',
    'Min Speed Rolling', 'Speed Standard Deviation Rolling',
    'Total Distance Traveled Rolling', 'Directionality Rolling', 'Tortuosity Rolling', 'Total Turning Angle Rolling', 'Spatial Coverage Rolling'

Mean Speed Rolling: The average speed within each rolling window. Median Speed Rolling: The median speed within each rolling window. Max Speed Rolling: The highest speed within each rolling window. Min Speed Rolling: The lowest speed within each rolling window. Speed Standard Deviation Rolling: The variability of speeds within each rolling window. Total Distance Traveled Rolling: The average distance traveled within each rolling window.

Directionality Rolling: The average directionality within each rolling window, indicating how straight the path is in that segment of the track.

Tortuosity Rolling: The average tortuosity within each rolling window, indicating how convoluted or twisted the path is in that segment of the track. Tortuosity is calculated as the ratio of the total path length to the Euclidean distance between the start and end points of each window. This metric helps in understanding the complexity of movement patterns over short segments of the track, providing insights into the movement behavior of tracked objects.

Total Turning Angle Rolling: The average total turning angle within each rolling window, indicating how much the direction of movement changes over short segments of the track. This metric helps in understanding the directional changes and maneuverability of the tracked objects over time.

Spatial Coverage Rolling: The average spatial coverage within each rolling window, representing the area (in 2D) or volume (in 3D) covered by the tracked object over short segments of the track. This metric helps in understanding the spatial extent and movement patterns of the tracked objects over time.

Morphological Metrics

In the notebook, additional morphological and intensity metrics such as the mean, median, standard deviation, minimum, and maximum values of spots parameters computed for each track when available.

The following metrics are derived from the information provided by TrackMate in the spots table and include statistical properties. Please check the TrackMate documentation for more details about these individual spot metrics:

'MEAN_INTENSITY_CH1', 'MEDIAN_INTENSITY_CH1', 'MIN_INTENSITY_CH1', 'MAX_INTENSITY_CH1',
'TOTAL_INTENSITY_CH1', 'STD_INTENSITY_CH1', 'CONTRAST_CH1', 'SNR_CH1', 'ELLIPSE_X0',
'ELLIPSE_Y0', 'ELLIPSE_MAJOR', 'ELLIPSE_MINOR', 'ELLIPSE_THETA', 'ELLIPSE_ASPECTRATIO',
'AREA', 'PERIMETER', 'CIRCULARITY', 'SOLIDITY', 'SHAPE_INDEX','MEAN_INTENSITY_CH2', 'MEDIAN_INTENSITY_CH2', 'MIN_INTENSITY_CH2', 'MAX_INTENSITY_CH2',
'TOTAL_INTENSITY_CH2', 'STD_INTENSITY_CH2', 'CONTRAST_CH2', 'SNR_CH2', 'MEAN_INTENSITY_CH3', 'MEDIAN_INTENSITY_CH3', 'MIN_INTENSITY_CH3', 'MAX_INTENSITY_CH3',
'TOTAL_INTENSITY_CH3', 'STD_INTENSITY_CH3', 'CONTRAST_CH3', 'SNR_CH3', 'MEAN_INTENSITY_CH4', 'MEDIAN_INTENSITY_CH4', 'MIN_INTENSITY_CH4', 'MAX_INTENSITY_CH4',
'TOTAL_INTENSITY_CH4', 'STD_INTENSITY_CH4', 'CONTRAST_CH4', 'SNR_CH4'

The following metrics are derived from the information provided by ilastik. Please check the ilastik documentation for more details about these individual spot metrics:

'Diameter_0',	'Euclidean_Diameter_0',	'Number_of_Holes_0',	'Center_of_the_Skeleton_0',	'Center_of_the_Skeleton_1',
'Length_of_the_Skeleton_0',	'Convexity_0',	'Number_of_Defects_0',	'Mean_Defect_Displacement_0',	'Mean_Defect_Area_0',
'Variance_of_Defect_Area_0',	'Convex_Hull_Center_0',	'Convex_Hull_Center_1', 'Object_Center_0',	'Object_Center_1',
'Object_Area_0',	'Kurtosis_of_Intensity_0',	'Maximum_intensity_0',	'Mean_Intensity_0',	'Minimum_intensity_0',
'Principal_components_of_the_object_0', 'Principal_components_of_the_object_1',	'Principal_components_of_the_object_2',
'Principal_components_of_the_object_3', 'Radii_of_the_object_0',	'Radii_of_the_object_1',	'Skewness_of_Intensity_0',
'Total_Intensity_0',	'Variance_of_Intensity_0',	'Bounding_Box_Maximum_0',	'Bounding_Box_Maximum_1',	'Bounding_Box_Minimum_0',
'Bounding_Box_Minimum_1',	'Size_in_pixels_0'

Distance to ROI Metrics

These metrics are computed in the CellTracksColab - Distance to ROI notebook (https://github.com/CellMigrationLab/CellTracksColab/wiki/Distance-to-ROI-analyses).

Calculate metrics like MaxDistance, MinDistance, MedianDistance, and more to analyze spatial behavior in relation to ROIs.

MaxDistance_{ROI_name}:

The maximum distance of the track from the ROI during the tracking period. Indicates the farthest point reached relative to the ROI. MinDistance_{ROI_name}:

The minimum distance of the track from the ROI during the tracking period. Represents the closest approach to the ROI. StartDistance_{ROI_name} and EndDistance_{ROI_name}:

Distances from the ROI at the start and end of the tracking period, respectively. Useful for understanding initial and final positioning relative to the ROI. MedianDistance_{ROI_name}:

The median of all recorded distances to the ROI. Provides a central tendency measure, less affected by outliers than the mean. StdDevDistance_{ROI_name}:

Standard deviation of the distances. Indicates the variability or consistency of the track's distance from the ROI. DirectionMovement_{ROI_name}:

Calculated as EndDistance - StartDistance. A positive value indicates moving away from the ROI over time, and a negative value suggests moving closer. AvgRateChange_{ROI_name}:

Average rate of change in distance per frame. Helps assess the speed of movement towards or away from the ROI. PercentageChange_{ROI_name}:

Percentage change in distance from the start to the end of the track. Normalizes the movement relative to the initial distance. TrendSlope_{ROI_name}:

Slope of a linear regression line fitted to the distance values over time. Indicates the general trend of movement (increasing or decreasing distance). These metrics are calculated considering the distance to the closest ROI at each time point. If the ROI moves, the metrics reflect the relative motion between the track and the ROI. The closest ROI to a track at each time point may change if multiple ROIs exist. This factor is inherently considered in the distance calculations. It is essential to consider the movement of both the track and the ROI when interpreting these metrics. For example, a decreasing distance over time could mean the track is moving towards the ROI, the ROI is moving towards the track, or both.

Track Metrics fully imported from your tracking software

Please check the documentation of your tracking software for more details