Skip to content

Commit

Permalink
fixed axis minor error
Browse files Browse the repository at this point in the history 
fixed axis minor error caused by skimage.measure:

scikit-image/scikit-image#6630
  • Loading branch information
@RoccoDAnt
RoccoDAnt committed Aug 12, 2024
1 parent 9d1a2ad commit 4842ae5
Showing 1 changed file with 131 additions and 63 deletions.
194 changes: 131 additions & 63 deletions napari_zelda/napari_zelda.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -4,6 +4,7 @@
from napari_plugin_engine import napari_hook_implementation
from qtpy.QtWidgets import QWidget, QHBoxLayout, QPushButton, QGridLayout, QGroupBox
from napari.layers import Image, Labels, Layer, Points
from napari.utils import progress
from magicgui import magicgui, magic_factory
import napari
from napari import Viewer
Expand Down Expand Up @@ -209,43 +210,75 @@ def watershed_children_pop(viewer: 'napari.Viewer', label, DistanceMap: Image, b
)
def measure_one_pop( label, labels: Image, original: Image, save_log, save_to):
voxel_size=np.prod(original.scale)
properties=measure.regionprops_table(labels.data, original.data, properties= ['label','area', 'intensity_mean','intensity_min','intensity_max','equivalent_diameter','axis_major_length','axis_minor_length','centroid','centroid_weighted','extent','solidity'])
prop={'Label': properties['label'], 'Area': properties['area']*original.scale[-1]*original.scale[-2], 'Volume': properties['area']*voxel_size,'Equivalent_diameter': properties['equivalent_diameter']*original.scale[-1],'MFI': properties['intensity_mean'],
'Min_Intensity': properties['intensity_min'], 'Max_Intensity': properties['intensity_max'],'MajorAxis_Length': properties['axis_major_length']*original.scale[-1],
'MinorAxis_Length': properties['axis_minor_length']*original.scale[-1],
'Extent': properties['extent'],
'Solidity': properties['solidity']
}
if len(original.data.shape)==2:
prop['Centroid_X']= properties['centroid-1']*original.scale[-1]
prop['Centroid_Y']= properties['centroid-0']*original.scale[-2]

prop['Weighted_Centroid_X']= properties['centroid_weighted-1']*original.scale[-1]
prop['Weighted_Centroid_Y']= properties['centroid_weighted-0']*original.scale[-2]

additional_properties_2D=measure.regionprops_table(labels.data, original.data, properties= ['orientation','perimeter'])
prop['Orientation']= additional_properties_2D['orientation']
prop['Perimeter']= additional_properties_2D['perimeter']*original.scale[-1]

if len(original.data.shape)==3:
prop['Centroid_X']= properties['centroid-2']*original.scale[-1]
prop['Centroid_Y']= properties['centroid-1']*original.scale[-2]
prop['Centroid_Z']= properties['centroid-0']*original.scale[-3]

prop['Weighted_Centroid_X']= properties['centroid_weighted-2']*original.scale[-1]
prop['Weighted_Centroid_Y']= properties['centroid_weighted-1']*original.scale[-2]
prop['Weighted_Centroid_Z']= properties['centroid_weighted-0']*original.scale[-3]

prop_df=pd.DataFrame(prop)
single_properties=pd.DataFrame()
for i in progress(range(1,np.max(labels.data)+1)):
mask=np.isin(labels.data, i)
single_labels=np.array(labels.data*np.array(mask))
try:
properties=measure.regionprops_table(single_labels, original.data, properties= ['label','area', 'intensity_mean','intensity_min','intensity_max','equivalent_diameter','axis_major_length','axis_minor_length','centroid','centroid_weighted','extent','solidity'])
prop={'Label': properties['label'], 'Area': properties['area']*original.scale[-1]*original.scale[-2], 'Volume': properties['area']*voxel_size,'Equivalent_diameter': properties['equivalent_diameter']*original.scale[-1],'MFI': properties['intensity_mean'],
'Min_Intensity': properties['intensity_min'], 'Max_Intensity': properties['intensity_max'],'MajorAxis_Length': properties['axis_major_length']*original.scale[-1],
'MinorAxis_Length': properties['axis_minor_length']*original.scale[-1],
'Extent': properties['extent'],
'Solidity': properties['solidity']
}
if len(original.data.shape)==2:
prop['Centroid_X']= properties['centroid-1']*original.scale[-1]
prop['Centroid_Y']= properties['centroid-0']*original.scale[-2]

prop['Weighted_Centroid_X']= properties['centroid_weighted-1']*original.scale[-1]
prop['Weighted_Centroid_Y']= properties['centroid_weighted-0']*original.scale[-2]

additional_properties_2D=measure.regionprops_table(single_labels.data, original.data, properties= ['orientation','perimeter'])
prop['Orientation']= additional_properties_2D['orientation']
prop['Perimeter']= additional_properties_2D['perimeter']*original.scale[-1]

if len(original.data.shape)==3:
prop['Centroid_X']= properties['centroid-2']*original.scale[-1]
prop['Centroid_Y']= properties['centroid-1']*original.scale[-2]
prop['Centroid_Z']= properties['centroid-0']*original.scale[-3]

prop['Weighted_Centroid_X']= properties['centroid_weighted-2']*original.scale[-1]
prop['Weighted_Centroid_Y']= properties['centroid_weighted-1']*original.scale[-2]
prop['Weighted_Centroid_Z']= properties['centroid_weighted-0']*original.scale[-3]
single_properties= pd.concat([single_properties, pd.DataFrame(prop)])
except:
properties=measure.regionprops_table(single_labels, original.data, properties= ['label','area', 'intensity_mean','intensity_min','intensity_max','equivalent_diameter','axis_major_length','centroid','centroid_weighted','extent','solidity'])
prop={'Label': properties['label'], 'Area': properties['area']*original.scale[-1]*original.scale[-2], 'Volume': properties['area']*voxel_size,'Equivalent_diameter': properties['equivalent_diameter']*original.scale[-1],'MFI': properties['intensity_mean'],
'Min_Intensity': properties['intensity_min'], 'Max_Intensity': properties['intensity_max'],'MajorAxis_Length': properties['axis_major_length']*original.scale[-1],
'MinorAxis_Length': 0.0*original.scale[-1], #regionprops axis_minor_length error for small objects
'Extent': properties['extent'],
'Solidity': properties['solidity']
}
if len(original.data.shape)==2:
prop['Centroid_X']= properties['centroid-1']*original.scale[-1]
prop['Centroid_Y']= properties['centroid-0']*original.scale[-2]

prop['Weighted_Centroid_X']= properties['centroid_weighted-1']*original.scale[-1]
prop['Weighted_Centroid_Y']= properties['centroid_weighted-0']*original.scale[-2]

additional_properties_2D=measure.regionprops_table(single_labels.data, original.data, properties= ['orientation','perimeter'])
prop['Orientation']= additional_properties_2D['orientation']
prop['Perimeter']= additional_properties_2D['perimeter']*original.scale[-1]

if len(original.data.shape)==3:
prop['Centroid_X']= properties['centroid-2']*original.scale[-1]
prop['Centroid_Y']= properties['centroid-1']*original.scale[-2]
prop['Centroid_Z']= properties['centroid-0']*original.scale[-3]

prop['Weighted_Centroid_X']= properties['centroid_weighted-2']*original.scale[-1]
prop['Weighted_Centroid_Y']= properties['centroid_weighted-1']*original.scale[-2]
prop['Weighted_Centroid_Z']= properties['centroid_weighted-0']*original.scale[-3]
single_properties= pd.concat([single_properties, pd.DataFrame(prop)])

prop_df=pd.DataFrame(single_properties)
#prop_df=dt.Frame(prop) #datatable instead of pandas
prop_df.to_csv(str(save_to)+'\Results.csv')

log=Label(name='Log', tooltip=None)
log.value="-> GB: sigma="+str(gaussian_blur_one_pop.sigma.value)+"-> Th="+str(threshold_one_pop.threshold.value)+"-> DistMap"
#log.value=log.value+"-> Maxima: min_dist=" + str(show_seeds_one_pop.min_dist.value) + " -> Found n="+str(prop_df.nrows)+ " objects" #if using datatable
log.value=log.value+"-> Maxima: min_dist=" + str(show_seeds_one_pop.min_dist.value) + " -> Found n="+str(len(prop_df))+ " objects"
#measure_one_pop.insert(4,log)


if save_log == True:
Log_file = open(str(save_to)+'\Log_ZELDA_single_population.txt','w')
Expand All @@ -257,42 +290,78 @@ def measure_one_pop( label, labels: Image, original: Image, save_log, save_to):
persist=True
)
def relate_and_measure(viewer: 'napari.Viewer', label, Parents_labels: Image, Children_labels: Image, Original_to_measure: Image, save_to_path):
properties=measure.regionprops_table(Children_labels.data, Original_to_measure.data, properties= ['label','area','intensity_mean','intensity_min','intensity_max','equivalent_diameter','axis_major_length','axis_minor_length','centroid','centroid_weighted','extent','solidity']
)
binary_ch=Children_labels.data>0
corresponding_parents=Parents_labels.data*binary_ch
viewer.add_image(corresponding_parents, scale=Parents_labels.scale, rgb=False, name='Labelled children objects by parent', opacity=0.6, rendering='mip', blending='additive', colormap='inferno')
voxel_size=np.prod(Original_to_measure.scale)
properties_CorrespondingParent=measure.regionprops_table(Children_labels.data, Parents_labels.data, properties=['intensity_max'])
prop={'Parent_label': properties_CorrespondingParent['intensity_max'].astype(float), 'Label': properties['label'], 'Area': properties['area']*Original_to_measure.scale[-1]*Original_to_measure.scale[-2],
'Volume': properties['area']*voxel_size,
'Equivalent_diameter': properties['equivalent_diameter']*Original_to_measure.scale[-1],'MFI': properties['intensity_mean'],'Min_Intensity': properties['intensity_min'], 'Max_Intensity': properties['intensity_max'],
'MajorAxis_Length': properties['axis_major_length']*Original_to_measure.scale[-1],
'MinorAxis_Length': properties['axis_minor_length']*Original_to_measure.scale[-1],
'Extent': properties['extent'],
'Solidity': properties['solidity']
}
if len(Original_to_measure.data.shape)==2:
prop['Centroid_X']= properties['centroid-1']*Original_to_measure.scale[-1]
prop['Centroid_Y']= properties['centroid-0']*Original_to_measure.scale[-2]

prop['Weighted_Centroid_X']= properties['centroid_weighted-1']*Original_to_measure.scale[-1]
prop['Weighted_Centroid_Y']= properties['centroid_weighted-0']*Original_to_measure.scale[-2]

additional_properties_2D_twoPop=measure.regionprops_table(Children_labels.data, Original_to_measure.data, properties= ['orientation','perimeter'])
prop['Orientation']= additional_properties_2D_twoPop['orientation']
prop['Perimeter']= additional_properties_2D_twoPop['perimeter']*Original_to_measure.scale[-1]

if len(Original_to_measure.data.shape)==3:
prop['Centroid_X']= properties['centroid-2']*Original_to_measure.scale[-1]
prop['Centroid_Y']= properties['centroid-1']*Original_to_measure.scale[-2]
prop['Centroid_Z']= properties['centroid-0']*Original_to_measure.scale[-3]

prop['Weighted_Centroid_X']= properties['centroid_weighted-2']*Original_to_measure.scale[-1]
prop['Weighted_Centroid_Y']= properties['centroid_weighted-1']*Original_to_measure.scale[-2]
prop['Weighted_Centroid_Z']= properties['centroid_weighted-0']*Original_to_measure.scale[-3]

prop_df=pd.DataFrame(prop)
single_child_properties=pd.DataFrame()
for i in progress(range(1,np.max(Children_labels.data)+1)):
mask=np.isin(Children_labels.data, i)
single_child_labels=np.array(Children_labels.data*np.array(mask))
try:
properties=measure.regionprops_table(single_child_labels, Original_to_measure.data, properties= ['label','area','intensity_mean','intensity_min','intensity_max','equivalent_diameter','axis_major_length','axis_minor_length','centroid','centroid_weighted','extent','solidity'])
properties_CorrespondingParent=measure.regionprops_table(single_child_labels, Parents_labels.data, properties=['intensity_max'])
prop={'Parent_label': properties_CorrespondingParent['intensity_max'].astype(float), 'Label': properties['label'], 'Area': properties['area']*Original_to_measure.scale[-1]*Original_to_measure.scale[-2],
'Volume': properties['area']*voxel_size,
'Equivalent_diameter': properties['equivalent_diameter']*Original_to_measure.scale[-1],'MFI': properties['intensity_mean'],'Min_Intensity': properties['intensity_min'], 'Max_Intensity': properties['intensity_max'],
'MajorAxis_Length': properties['axis_major_length']*Original_to_measure.scale[-1],
'MinorAxis_Length': properties['axis_minor_length']*Original_to_measure.scale[-1],
'Extent': properties['extent'],
'Solidity': properties['solidity']
}
if len(Original_to_measure.data.shape)==2:
prop['Centroid_X']= properties['centroid-1']*Original_to_measure.scale[-1]
prop['Centroid_Y']= properties['centroid-0']*Original_to_measure.scale[-2]

prop['Weighted_Centroid_X']= properties['centroid_weighted-1']*Original_to_measure.scale[-1]
prop['Weighted_Centroid_Y']= properties['centroid_weighted-0']*Original_to_measure.scale[-2]

additional_properties_2D_twoPop=measure.regionprops_table(single_child_labels, Original_to_measure.data, properties= ['orientation','perimeter'])
prop['Orientation']= additional_properties_2D_twoPop['orientation']
prop['Perimeter']= additional_properties_2D_twoPop['perimeter']*Original_to_measure.scale[-1]

if len(Original_to_measure.data.shape)==3:
prop['Centroid_X']= properties['centroid-2']*Original_to_measure.scale[-1]
prop['Centroid_Y']= properties['centroid-1']*Original_to_measure.scale[-2]
prop['Centroid_Z']= properties['centroid-0']*Original_to_measure.scale[-3]

prop['Weighted_Centroid_X']= properties['centroid_weighted-2']*Original_to_measure.scale[-1]
prop['Weighted_Centroid_Y']= properties['centroid_weighted-1']*Original_to_measure.scale[-2]
prop['Weighted_Centroid_Z']= properties['centroid_weighted-0']*Original_to_measure.scale[-3]
single_child_properties= pd.concat([single_child_properties, pd.DataFrame(prop)])
except:
properties=measure.regionprops_table(single_child_labels, Original_to_measure.data, properties= ['label','area','intensity_mean','intensity_min','intensity_max','equivalent_diameter','axis_major_length','centroid','centroid_weighted','extent','solidity'])
properties_CorrespondingParent=measure.regionprops_table(single_child_labels, Parents_labels.data, properties=['intensity_max'])
prop={'Parent_label': properties_CorrespondingParent['intensity_max'].astype(float), 'Label': properties['label'], 'Area': properties['area']*Original_to_measure.scale[-1]*Original_to_measure.scale[-2],
'Volume': properties['area']*voxel_size,
'Equivalent_diameter': properties['equivalent_diameter']*Original_to_measure.scale[-1],'MFI': properties['intensity_mean'],'Min_Intensity': properties['intensity_min'], 'Max_Intensity': properties['intensity_max'],
'MajorAxis_Length': properties['axis_major_length']*Original_to_measure.scale[-1],
'MinorAxis_Length': 0.0*Original_to_measure.scale[-1], #regionprops axis_minor_length error for small objects
'Extent': properties['extent'],
'Solidity': properties['solidity']
}
if len(Original_to_measure.data.shape)==2:
prop['Centroid_X']= properties['centroid-1']*Original_to_measure.scale[-1]
prop['Centroid_Y']= properties['centroid-0']*Original_to_measure.scale[-2]

prop['Weighted_Centroid_X']= properties['centroid_weighted-1']*Original_to_measure.scale[-1]
prop['Weighted_Centroid_Y']= properties['centroid_weighted-0']*Original_to_measure.scale[-2]

additional_properties_2D_twoPop=measure.regionprops_table(single_child_labels, Original_to_measure.data, properties= ['orientation','perimeter'])
prop['Orientation']= additional_properties_2D_twoPop['orientation']
prop['Perimeter']= additional_properties_2D_twoPop['perimeter']*Original_to_measure.scale[-1]

if len(Original_to_measure.data.shape)==3:
prop['Centroid_X']= properties['centroid-2']*Original_to_measure.scale[-1]
prop['Centroid_Y']= properties['centroid-1']*Original_to_measure.scale[-2]
prop['Centroid_Z']= properties['centroid-0']*Original_to_measure.scale[-3]

prop['Weighted_Centroid_X']= properties['centroid_weighted-2']*Original_to_measure.scale[-1]
prop['Weighted_Centroid_Y']= properties['centroid_weighted-1']*Original_to_measure.scale[-2]
prop['Weighted_Centroid_Z']= properties['centroid_weighted-0']*Original_to_measure.scale[-3]
single_child_properties= pd.concat([single_child_properties, pd.DataFrame(prop)])

prop_df=pd.DataFrame(single_child_properties)
#prop_df=dt.Frame(prop) #datatable instead of pandas
prop_df.to_csv(str(save_to_path)+'\Results_parents-children.csv')

Expand All @@ -302,7 +371,6 @@ def relate_and_measure(viewer: 'napari.Viewer', label, Parents_labels: Image, Ch
log.value=log.value+"\n-> GB: sigma="+str(gaussian_blur_children_pop.sigma.value)+"-> Th_children="+str(threshold_children.threshold.value)+"-> DistMap"
#log.value=log.value+"-> Maxima: min_dist=" + str(show_seeds_children_pop.min_dist.value) + " -> Found n="+str(prop_df.nrows)+ " objects" #if using datatable
log.value=log.value+"-> Maxima: min_dist=" + str(show_seeds_children_pop.min_dist.value) + " -> Found n="+str(len(prop_df))+ " objects"
measure_one_pop.insert(4,log)

#if save_log == True:
Log_file = open(r''+str(save_to_path)+'\Log_ZELDA_Parents_Children.txt','w')
Expand Down

0 comments on commit 4842ae5

Please sign in to comment.