Learning Cell Nuclei Segmentation Using Labels Generated with Classical Image Analysis Methods

• Creating manual annotations in a large number of images is a tedious bottleneck that limits deep learning use inmany applications. Here, we present a study in which we used the output of a classical image analysis pipeline aslabels when training a convolutional neural network (CNN). This may not only reduce the time experts spendannotating images but it may also lead to an improvement of results when compared to the output from the classicalpipeline used in training. In our application, i.e., cell nuclei segmentation, we generated the annotations usingCellProfiler (a tool for developing classical image analysis pipelines for biomedical applications) and trained onthem a U-Net-based CNN model. The best model achieved a 0.96 dice-coefficient of the segmented Nuclei and a0.84 object-wise Jaccard index which was better than the classical method used for generating the annotations by0.02 and 0.34, respectively. Our experimental results show that in this application, not only such training is feasiblebut also that the deep learning segmentations are a clear improvement compared to the output from the classicalpipeline used for generating the annotations.

Subject headings

NATURVETENSKAP  -- Data- och informationsvetenskap -- Annan data- och informationsvetenskap (hsv//swe)

NATURAL SCIENCES  -- Computer and Information Sciences -- Other Computer and Information Science (hsv//eng)

Keyword

Deep learning

U-Net

CellProfiler

Data annotation

Microscopy

Computerized Image Processing

Datoriserad bildbehandling

Publication and Content Type

ref (subject category)

kon (subject category)