search

OldaKodym / BUT_autoimplant_public

4 stars 1 forks source link

question about skullbreak #1

Open zhangxinkang opened 2 years ago

zhangxinkang commented 2 years ago

Dear Mr.Kodym: I am a graduate student from FuDan university in China. My research is focused on image segmentation and restoration. I have recently read your paper "Skull shape reconstruction using cascaded convolutional networks ". I am very interested in this article and have studied it carefully, and i have questions about the dataset :Skullbreak. From the paper, i see your team use a metric named " average symmetric unsigned distance" ,is this equivalent to the ADS(average symmetric surface distance)? Besides, i am wondering if you could kindly send me the source code of the evaluation/metric part of the program(of course i will be more appreciated if you can share the whole source program) and necessary information.I promise they will be used only for research purposes.

    Thank you very much for your kind consideration and i am looking forward to your early reply.

Sincerely,yours

Xinkang ZhangSkullbreak

OldaKodym commented 2 years ago

Hi,

Thanks for reaching out! In the paper you referenced, we used a distance measure only on the outer surface of the skulls which required some manual surface definitions. If you want to compare your results with state of the art, I invite you to checkout the AutoImplant challenge (https://autoimplant2021.grand-challenge.org/) which also uses the SkullBreak dataset as one of the tracks. You can find the implementation of the metrics used in the challenge evaluation here: https://github.com/OldaKodym/evaluation_metrics/blob/master/metrics.py Hope that answers your questions.

Regards Oldřich

zhangxinkang commented 2 years ago

Hi,

Thanks for reaching out! In the paper you referenced, we used a distance measure only on the outer surface of the skulls which required some manual surface definitions. If you want to compare your results with state of the art, I invite you to checkout the AutoImplant challenge (https://autoimplant2021.grand-challenge.org/) which also uses the SkullBreak dataset as one of the tracks. You can find the implementation of the metrics used in the challenge evaluation here: https://github.com/OldaKodym/evaluation_metrics/blob/master/metrics.py Hope that answers your questions.

Regards Oldřich

Thanks for your reply. I didn't find your reply in time because I was busy these two days. I apologize for this. Glad to get your reply and advice,i will try to use it.

regards xinkang