#尊敬的用户，由于网络监管政策的限制，部分内容暂时无法在本网站直接浏览。我们已经为您准备了相关原始数据和链接，感谢您的理解与支持。
<iframe width="760" height="400" src="http://www.youtube.com/embed/ApOwcOFMUcg?si=azuelTcu87ixq8hy" title="YouTube video player"></iframe>

If video quality is poor, try watching directly from Youtube at http://www.youtube.com/watch?v=ApOwcOFMUcg

0:00 - 1:30 Install ilastik by downloading it from ilastik.org/download and running the installer.

1:30 - 3:00 Generate training images by running SynBot on a set of images so that any preprocessing steps will be the same as a SynBot run.

3:00 - 4:00 Copy two-color training images to a training images folder

4:00 - 4:20 Make empty folder for holding red images and green images

4:20 - 4:32 Run extract_channels.ijm macro from the extra code folder on the SynBot GitHub repository

4:32 - 5:22 Select merged images folder, then red images folder, then green images folder

5:22 - 5:39 Open ilastik and start a new Pixel Classification project

5:39 - 6:01 Save the ilastik project file to a safe place where you can use it to threshold with SynBot

6:01 - 6:20 Click add new image and select the red training image (or images)

6:20 - 6:31 Click the "2. Feature Selection" drop down then "select features" and use Ctrl + A to select all features.

6:31 - 6:34 Click the "3. Training" drop down to advance

6:34 - 6:41 Zoom in image using Ctrl + +

6:41 - 7:05 Use the Label 1 to mark several examples of foreground puncta. Only a handful are needed to get started.

7:05 - 7:21 Click Label 2 and use this to label some examples of background regions.

7:21 - 7:37 Click Live Update to see a rendering of the ilastik model that will update when new markings are added. Use Ctrl + - to zoom out and check how well this model is working.

7:37 - 7:49 Save the ilastik project. This saved model will be used by SynBot and you do not need to change anything in the steps 4 or 5 in the ilastik pipeline.

7:55 - 9:37 Repeat steps 8-17 for the green channel image

9:37 - 9:51 Open and run SynBot.ijm

9:51 - 10:14 Enter your analysis parameters, including the ilastik thresholding method.

10:14 - 10:29 Select ilastik.exe application . On Windows this will typically be in a path like C:/Program Files/ilastik1.4.0/ilastik.exe

10:29 - 10:34 Select the ilastik project file for thresholding the red channel.

10:34 - 10:40 Select the ilastik project file for thresholding the green channel.

10:40 - 10:42 Enter the threshold of ilastik confidence to use for the analysis. ilastik will give a confidence score from 0 to 1 of how likely the given pixel is part of the foreground. A good threshold is something between 0.5 and 0.9.

10:42 - 12:14 View results