"good", "bad" patient states in the UMAP (Figure 3 in the paper)

Question

"good", "bad" patient states in the UMAP (Figure 3 in the paper)

Mda233 opened this issue a year ago · comments

Hi,

Firstly, I want to appreciate your excellent work and the dataset release.

I was trying to reproduce the UMAP as drawn in Figure 3 by following "A good state is defined as non-intubated, not in the ICU, or the last state before discharge. A bad state is defined as intubated, in the ICU, or any state which is the last in the series and the patient is determined to have died." However, I have too many "bad" state dots shown in my figure.

Could you please tell me the exact rules you select for the "good"/"bad" states, or release the script for drawing that UMAP?

It would be a great help for my work.I am looking forward to hearing from you.

Thank you!

Joseph Paul Cohen · Answer 1 · Mon Apr 24 2023 11:30:25 GMT+0800 (China Standard Time)

Thanks for reaching out. I thought it was online but it isn't. I found a copy and I'll clean it up and put it online in the next day or so.

Joseph Paul Cohen · Answer 2 · Mon Apr 24 2023 11:50:32 GMT+0800 (China Standard Time)

It was easier than I expected! Here it is: https://github.com/mlmed/torchxrayvision/blob/master/scripts/xray_trajectory.ipynb

Mda233 · Answer 3 · Mon Apr 24 2023 12:14:52 GMT+0800 (China Standard Time)

Thank you so much!