Xiyue-Wang / RetCCL

GNU General Public License v3.0
93 stars 8 forks source link

training code #5

Open juingzhou opened 10 months ago

juingzhou commented 10 months ago

Hi, I've read your paper in MIA and I have an interest to your works. I would like to consult when the training code will be available?

Xiyue-Wang commented 10 months ago

Thank you very much for your attention, these three search methods, you can refer to here(https://github.com/jacobluber/PathologySearchComparison),the paper is here https://arxiv.org/abs/2306.17019. "We found that RetCCL was the most reliable in tissue retrieval by correctly identifying within the first five retrieved slides the correct tissue of diagnosis for all three cases. In contrast, both Yottixel and SISH failed to retrieve the correct tissue of origin at all top five ranked positions for two out of the three cases, which was unexpected given prior work demonstrating the high performance of these models." " Similarly, we also found that RetCCL outperformed both SISH and Yotixxel in retrieving patches containing normal alveolar tissue."

juingzhou commented 9 months ago

I perused the repository and found that no code exists to train on the retccl model.