Trace Trees: Automatic tracing mode (3D)

Procedure overview

See Tracing trees in automatic mode (3D) for step-by-step instructions.

  • If you want to associate trees with a specific color channel, select a single channel using either the Channel panel on the left side of the 3D environment window or the Image Adjustment tool (on the Image and Workspace ribbons) in the 2D window.
  • To trace using the default settings, click the Trace button.
  • Delete automatically traced trees by clicking Clear Tracing.
  • For a more accurate or faster tracing, adjust the settings first (Show Settings), then click the Trace button.
  • To see the full panel, click Show Settings.

Automatic tree tracing settings

Tracing Mode

The tracing methods are based on algorithms that work by tracing along the trees one point at a time and producing three measurements at each point: (X,Y,Z) coordinate, thickness, estimated position of the next point. Choose one of the methods:

  • Directional Kernels: With this method, four directional kernels are matched to the image data. For a given point within the tree, the algorithm identifies the best positions and orientations for the top, bottom, left, and right kernels surrounding the point. The positions and orientations results are combined to estimate the next point to trace. Points are estimated until a set of stopping criteria is met.

    For details on the algorithm, see Rapid automated three-dimensional tracing of neurons from confocal image stacks (Al-Kofahi, Lasek, Szarowski, Pace, Nagy, Turner, and Roysam, 2002).

  • Rayburst Crawl: This method performs rayburst sampling measurements to obtain the diameter and centroid of the cross-section of the tree. Successive measurements are used to position the nodes that define its centerline. Points are estimated until a set of stopping criteria is met.

    For details on the rayburst sampling algorithm see Rayburst sampling, an algorithm for automated three-dimensional shape analysis from laser scanning microscopy images (Rodriguez, Ehlenberger, Hof, & Wearne, 2006).

  • Voxel Scooping: This algorithm generates clusters of voxels iteratively along the tree. These clusters are then used to position the nodes that define the centerline of the tree. As in the other methods, points are estimated until a set of stopping criteria is met.

    For details on the algorithm, see Three-Dimensional Neuron Tracing by Voxel Scooping (Rodriguez, Ehlenberger, Hof, & Wearne, 2009).

Detection Settings

Click Show Settings to view the following tools for setting and evaluating automatic tree tracing settings. The tools are arranged in 3 tabs:

Display Seeds tab

Display/Hide Seeds toggles between visible and invisible "seeds" that represent where Neurolucida 360 would automatically trace a tree. Clicking the display seeds button offers a preview of how the sensitivity and density settings affect results.

  • To change the seed color, use the color picker pull-down next to the Display/Hide Seeds button.

Sensitivity: Use the slider to adjust the sensitivity (%) for dim and low-contrast structures.

  • A value that is too high may result in seeds placed in the background where no tree exists.
  • A value that is too low may result in missing trees.

Typical process width: This setting helps guide the algorithm during tracing and is especially helpful when there is a large range in tree sizes. In this case, enter the typical width of the larger trees.

Density: An invisible grid is applied to sample the image uniformly. Choose one of the following:

  • Coarse: Fewer grid lines and less sampling performed; this can result in missing trees.
  • Medium: Optimizes density with speed, seed coverage, and final reconstruction accuracy.
  • Dense: More grid lines and more sampling performed. Dense sampling typically leads to longer processing times to produce more seeds and increase the tracing coverage. Accuracy may not improve with this setting; it may result in seeds placed in the background where no tree exists.

Refine Seeds tab

Validate Seeds button: Click to enable refinement of seeds. Neurolucida 360 software examines the possible seed-points' secondary characteristics to determine whether they are likely to belong to a tree.

Refine filter: The value indicates the quality of the match between seeds and trees. To change it, use the slider or type into the box.

Edit Seeds: Choose to add or remove seeds with your mouse:

  • Add seeds: When selected, clicking in the image adds a seed.

    • Use the color picker to change the color of manually added seeds.

    • Remove All Added: Click to remove all seeds that were added manually.

  • Remove seeds within circular cursor: When selected, you can remove seeds.

    1. Hover the mouse over a tree to see the circular cursor.
    2. Press and hold the CTRL key and scroll the mouse wheel to resize the cursor.
    3. Click in the image to remove seeds within the cursor radius OR hold down CTRL and drag to remove contiguous seeds.

Trace tab

Sensitivity: Use the slider or type a value to adjust sensitivity to dim and low-contrast structures.

  • A higher value is more sensitive and can lead to tracings in the background where there is no tree.

Gap tolerance: Use the slider to indicate the desired maximum distance between two segments to make a connection. If the distance is greater than the gap value, the segments are not connected.

Remove traces shorter than __µm: Type a value into the box; segments shorter than the value specified are not traced.

tree : Check the box to connect trees according to the Branch Connections criteria set in the Advanced Settings.

Advanced settings button: See Advanced settings for automatic tree tracing (3D)