วีดีโอ

Reply to Laura's 'From Zero to QuPath Hero' video on cell detection th-cam.com/video/dekcI8cjAeQ/w-d-xo.html Explaining why you can't interact with the image when setting cell intensity classifications (even though it's quite annoying, and you can interact when training a cell classifier)

Reply to Laura's 'From Zero to QuPath Hero' video on cell detection th-cam.com/video/9P8NsAVfrZE/w-d-xo.html Includes: * 5 ways to delete cells after their 'parent' annotation was already removed. * Some detection measurement table tricks * How it's possible to double-count cells if you start messing with annotations... and how to avoid it

Part one of the 'Introduction to Bioimage Analysis' course based on bioimagebook.github.io 0:00 Intro 0:29 Trustworthy images 3:56 A simple model of image formation 9:16 Brightness, contrast and lookup tables (colormaps) 16:38 Image histograms 19:47 Making measurements with ImageJ 22:04 Regions of interest (ROIs) 24:31 Line ROIs & profile plots 25:18 Point ROIs & counting 26:31 Multiple ROIs wi...

An introduction to an introduction... exploring image analysis with bioimagebook.github.io 0:00 Intro 0:16 Bioimage Book 2:02 Purpose of the videos 2:54 Basic ideas of bioimage anlaysis

Link to the script: gist.github.com/petebankhead/486690d13912f6a95bb0489458df959e For any questions, feedback or discussion: forum.image.sc/tag/qupath (I'm bad at checking TH-cam comments, so the forum is better...)

Demo of how annotations can be transferred between images in QuPath v0.4.0. Download QuPath and find links to the docs at qupath.github.io There's no audio, but hopefully it's easy enough to see what's going on. - I create annotations by thresholding one (huge) image - I copy the annotations to the system clipboard (this uses GeoJSON) - I paste the annotations to a new (huge) image - I interact...

QuPath is versatile bioimage analysis software, created by an enthusiastic user of ImageJ and Fiji. It was originally designed to address problems that couldn't easily be solved using other open-source tools, particularly for whole slide image analysis and digital pathology. However, QuPath can be used for many other tasks - and becomes even more powerful when combined with other open tools. Th...

Really just some words. Not least because the number of videos in the playlist should be a power of two.

Creating a script for reproducibility. Warning! I could remove the *highly* smoothed features (FWHM 100 µm) because they weren't used for the classifier. If any of them had been, this would have caused the classifier to fail - but not necessarily in an obvious way, and without reporting an error.

Overcoming a common script problem: requiring that annotations are selected before running cell detection. Note that the annotations remain selected, which is why there was no need for another 'selectAnnotations' call before calculating the smoothed features.

How scripts enable batch processing. Note that in QuPath v0.1.2 doesn't make exporting annotation measurements as easy as it might. There are some extra scripts to help at petebankhead.github.io/qupath/scripting/2018/03/04/script-annotation-export.html and petebankhead.github.io/qupath/scripting/2018/03/05/script-annotation-results-merge.html

Ending where we might have started... detecting the tissue. (Identifying regions of interest to analyze can be tricky in itself, and another potential source of error. I tend to think it's best to at least start with manual annotating regions of interest...)

Finishing up the analysis of one image.

Assigning intensity classifications to tumor cells while training the classifier. There is more information about setting positive classifications at petebankhead.github.io/qupath/tips/2018/03/22/setting-positive.html

25 - Revisiting past mistakes

24 - Toggle displaying classes

22 - Training a cell classifier

23 - More classifier suggestions

21 - Smoothed measurements

19 - Measurement maps

20 - Displaying cells

17 - A click too far

15 - Recovering parameters from the Workflow history

18 - Learning lessons or not

16 - Local measurements and the object hierarchy

14 - Other cell detection parameters

13 - Cell expansion

12 - Avoiding non nuclear staining

11 - Split by shape