Annotated image data is valuable for assessing the performance of an image processing pipeline and as training data for machine learning methods such as deep learning. When assessing the performance of a CellProfiler pipeline, for example a pipeline that segments nuclei, the annotated image data are used as the ground truth. The performance of the pipeline can be quantified by comparing the segmentation output to the ground truth and calculating a comparison metric, such as the Jaccard Index or F1 Score. Annotated images are also essential for deep learning applications as training data, for example see the 2018 Data Science Bowl; an in-depth discussion on how the Data Science Bowl images were annotated can be found on the Kaggle forum. Continue reading
Sharing Image Data
At the Imaging Platform we frequently need to send and receive images between collaborators or forum users. Working with images can become challenging simply due to their file size. Memory and disk limitations are often irritants when analyzing images, but in some cases, such as multi-dimensional images and whole-slide scans, images’ file sizes can be large enough that even accessing images becomes a major roadblock, especially when the size starts to rival the size of desktop or laptop hard drives. Continue reading
Now that CellProfiler can handle 3D images, how do you view the results? When processing 2D images, the IdentifyPrimaryObjects module will produce a figure that displays the original image next to the segmented image to make validation fast and convenient, a side-by-side comparison. However, the extra dimension in 3D complicates this approach, because the 3D image must be transformed in some manner in order to appear on a flat surface (e.g. the monitor or phone you’re using to read this blog post!). This blog post contains advice on how to look at 3D images and the 3D segmentation image produced by CellProfiler and introduces orthoviews in FIJI. If you would like to explore the example used in this post please check out the demo posted on the CellProfiler forum. Continue reading
CellProfiler is capable of accurate and reliable segmentation of cells by utilizing a broad collection of classical image processing methods. Peruse the documentation on the IdentifyPrimaryObjects module, for example, to get a sense of these, e.g., thresholding, declumping, and watershed. However, despite the many problems CellProfiler can readily solve, certain types of images are particularly challenging. For instance, when the biologically relevant objects are defined more by texture and context than raw intensity many classical image processing techiques can be foiled; DIC images of cells are a common biological example.