Voxel Spcing

Hi ! Dear MELA organizers, I notice that the voxel spacings of all data are 1mm. Is that correct or could you please release the spacing of each data? It's maybe helpful for the pre-processing. Thanks

Hello! Thanks for your reminder. The previous dataset lacks the real origin and spacing information of CT scans, and the spacing is automatically set to 1mm. However, the annotation file we gave is based on pixel coordinates, so using the previous dataset will not affect the detection. We have released MELA dataset v3.0, in which CT scans contain the real origin and spacing information. A csv file containing the origin and spacing information of each CT image is also provided. For details, please refer to Dataset page.

Hi Organizers,

I have downloaded the new uploaded sapcing .csv file and find something strange.

It seem like the spacing of the training dataset (No.1 - No.770) is A, A, B (such as 0.7617, 0.7617, 0.6999), whilst the spacing of the validation dataset (after No.771) is A, B, B (such as 0.7304, 1.0, 1.0) .

Cold you please check the spacing file again?

Best.

Hello.

There are some mistakes in the spacing.csv in v3.0 on Zenodo. Sorry for our mistake. The new version of csv file have been updated as v4.0 on Zenodo. Please refer to https://doi.org/10.5281/zenodo.6597131 Thank you very much for your reminder.

Best.