Reliable Identification and Quantification of Neural Cells in Microscopic Images of Neurospheres

This the release marks the timestamp and commit presented in the proposed published works called "Reliable Identification and Quantification of Neural Cells in Microscopic Images of Neurospheres".

See below training history, metrics and example data of the best proposed models.

Example Data

The esm9_d02 png-files uploaded with this release, mark an excerpt of the data used for the published works. The raw tag is the originally captured stained image. The adjusted tag is the manually color adjusted image.

• Red: Neurite Staining
• Green: Oligodendrocyte Staining
• Blue: Nucleus Staining

The .zip archives contain the corresponding annotations from two ROIs of the .png image.
One zip for oligodendrocytes, one for neurons.
When you unzip these files, you will see:

• overview.csv: This csv contains an ID for every nucleus in the ROIs, as well as their corresponding x and y coordinates (in the image) as well as the manually determined label.
• ESM9_D02.h5: This file contains the 64x64 pixel tile for every nucleus in the image ROIs. The ID matches the one in the csv.
• label.h5: This file contains the labels for the image IDs from the CSV.