Dice index to judge whether the model is good or bad

M3DV / FracNet

[eBioMedicine] Deep-learning-assisted detection and segmentation of rib fractures from CT scans: Development and validation of FracNet

https://m3dv.github.io/FracNet/

Apache License 2.0

51 stars 25 forks source link

Dice index to judge whether the model is good or bad #15

Open BelieferQAQ opened 2 years ago

BelieferQAQ commented 2 years ago

The final indicator of the ribfrac competition is the FROC value. Your dice is 71.5%, so what is the FROC value obtained in the final evaluation. I want to roughly judge the quality of my model according to dice index. Thank you very much.