FocusPath Full Dataset

Content

This is the full FocusPath dataset, containing 8640 patches of 1024 x 1024 images. These images are extracted from nine different stained slides.

The original Whole Slide Images are scanned by Huron TissueScope LE1.2. It uses a 40X optics lens at 0.25µm/pixel resolution. 

There are 14 absolute z-level scores assigned to these patches which correspond to the ground-truth class for the focus level. 

More Information

For more information, please refer to the following paper. Please cite this paper when using the dataset.

@article{8672094,
author={M. S. {Hosseini} and Y. {Zhang} and K. N. {Plataniotis}},
journal={IEEE Transactions on Image Processing},
title={Encoding Visual Sensitivity by MaxPol Convolution Filters for Image Sharpness Assessment},
year={2019},
volume={28},
number={9},
pages={4510-4525},
doi={10.1109/TIP.2019.2906582},
ISSN={1057-7149},
month={Sep.},
}

For information on how FocusPath can be used in a deep learning pipeline for training CNN models, please visit the repository at: https://github.com/icbcbicc/FocusLiteNN/

Contact

For questions, please contact:

Mahdi Hosseini

mahdi.hosseini@utoronto.ca

http://orcid.org/0000-0002-9147-0731