def test_read_atlas_preaks():
# Load a correct atlas
atlasreader.read_atlas_peak('aal', [10, 10, 10])
# Load a list of atlases
with pytest.raises(ValueError):
atlasreader.read_atlas_peak(2 * ['aal'], [10, 10, 10])
# Load 'all' atlas
with pytest.raises(ValueError):
atlasreader.read_atlas_peak('all', [10, 10, 10])
def queryatlas_main():
"""
The primary entrypoint for querying atlases via the command line
All parameters are read via argparse, so this should only be called from
the command line!
"""
opts = _queryatlas_parser()
print('{0:<25} {1:<25}\n{2:<25} {2:<25}'.format('Atlas', 'Label',
'=' * 10))
for atlas in check_atlases(opts.atlas):
label = read_atlas_peak(atlas, opts.coordinate, opts.prob_thresh)
if isinstance(label, list):
label = '\n{}'.format(' ' * 26).join(
['{:>2}% {}'.format(*e) for e in label])
print('{:<25} {}'.format(atlas.atlas, label))
def get_peaks(img_file, output_dir):
stats = fsl.utils.ImageStats()
stats.inputs.in_file = img_file
stats.inputs.op_string = '-r'
min = stats.run()
minval = min.outputs.out_stat[0]
out_fn = op.join(output_dir, '{0}_clusterinfo.tsv'.format(op.basename(img_file).split('.')[0]))
# use NiPype's FSL wrapper for "cluster" to generate peaks
cl = fsl.model.Cluster()
cl.inputs.in_file = img_file
cl.inputs.threshold = minval
cl.inputs.connectivity = 26
cl.inputs.no_table = True
cl.inputs.peak_distance = 15
cl.inputs.use_mm = True
cl.inputs.out_localmax_txt_file = out_fn
cl.run()
#Use pandas to clean up file
df_mm = pd.read_csv(out_fn, sep = "\t", index_col = False)
df_mm = df_mm.drop([df_mm.columns[5]], axis=1)
for i, row in df_mm.iterrows():
tmplabel = atlasreader.read_atlas_peak('aal', [row['x'], row['y'], row['z']])
if i == 0:
if tmplabel.split('_')[-1] in ['L', 'R']:
hemis = [tmplabel.split('_')[-1]]
labels = [' '.join(tmplabel.split('_')[:-1])]
else:
hemis = ['']
labels = [' '.join(tmplabel.split('_'))]
else:
if tmplabel.split('_')[-1] in ['L', 'R']:
hemis.append(tmplabel.split('_')[-1])
labels.append(' '.join(tmplabel.split('_')[:-1]))
else:
hemis.append('')
labels.append(' '.join(tmplabel.split('_')))
df_mm['Hemisphere'] = hemis
df_mm['Label'] = labels
df_mm.to_csv(out_fn, sep='\t', index=False)
return df_mm