File(s) under embargo
Reason: We are currently publishing a paper and applying a U.S. patent with Technology Transfer Office with the use of this dataset. It would be great if we could keep it embargo during the application process. We will open-source the dataset online too after filing the patent.
10
month(s)9
day(s)until file(s) become available
Supporting Data for "Advancing Biophysical Cytometry with Deep-learning (ABCD) for New-Generation Cell-based Diagnostics"
There are in total of 7 types of lung cancer cells (i.e. H69, H358, H520, H526, H1975, H2170 and HCC827). All the data were collected on 7 days using multi-ATOM setup, giving 3 batches per cell lines. Both single-cell brightfield and quantitative phase images (QPI) were collected.
For training and testing the beGAN model, the data were separated into "Train", "Valid" and "Test" set. They are subsampled and contain 500 cells respectively as a demonstration in this repository (Folder Dataset). Data was uploaded in .mat format with brightfield images in _BF.mat and QPI in _QPI.mat. The images are stored in format of ImageHeight * ImageWidth * NoOfCells with a field of view of 45μm.
