naamiinepal / xrayto3D-benchmark

GNU General Public License v3.0
9 stars 3 forks source link

Visualization: Compare best,median,worse cases for all benchmarked methods #21

Open msrepo opened 1 year ago

msrepo commented 1 year ago
msrepo commented 1 year ago

pandas has a weird API

print(df.nlargest(1, ["DSC"], "first")["subject-id"].values[0])
print(df.nsmallest(1, ["DSC"], "first")["subject-id"].values[0])
print(df[df.DSC == df.median(numeric_only=True)["DSC"]]["subject-id"].values[0])
msrepo commented 1 year ago

Median Results median_coronal median_coronal median_coronal median_coronal

msrepo commented 1 year ago

pick random samples instead of median, worse, best. currently, the results are stored as follows: image keep the same configuration, but we need a way to identify random samples for tiling later.

def save_montage(ANATOMY, subject_type):

will also need to be modified. either track the subject-id or find a way to consistently track the random samples by consistent naming convention.

msrepo commented 1 year ago

Useful comparative visualization image

msrepo commented 1 year ago

Median Results median_coronal median_coronal median_coronal median_coronal

The prediction resolution resampling and metadata issue ,see #1, has risen its ugly head here. I must have changed the parameters when evaluating attentionunet results for vertebra.

msrepo commented 1 year ago

Median Results median_coronal median_coronal median_coronal median_coronal

The prediction resolution resampling and metadata issue ,see #1, has risen its ugly head here. I must have changed the parameters when evaluating attentionunet results for vertebra.

Without resampling, metadata copied from groundtruth image without resampling, without metadata copied from groundtruth image

both of these are wrong. 1) either set resample=True when you copy metada from groundtruth image The preview happened correctly even if the data and metadata did not match in this case because we reslice when previewing. image

2) or set resample=False and set the correct metadata manually.

msrepo commented 1 year ago

fixed visualization issues: see #28 before and after median_coronal

median_coronal