Find Tract Bounding Boxes

In [1]:

    

#f = '/media/sf_SharedFolder/Code/libraries_of_mine/github/ConWhAt/ConWhAt/atlases/volumetric/dipy_dsi_sd4_l2k8_sc33/vismap_grp_cat_rois_v2_70_norm.nii.gz'


    

In [2]:

    

%matplotlib inline

from matplotlib import pyplot

import nibabel as nib

from nilearn.plotting import plot_stat_map

from nilearn.image import index_img


    

    




/home/john/Software/anaconda/envs/_ipython2.4/lib/python2.7/site-packages/matplotlib/font_manager.py:273: UserWarning: Matplotlib is building the font cache using fc-list. This may take a moment.
  warnings.warn('Matplotlib is building the font cache using fc-list. This may take a moment.')

In [3]:

    

from mpl_toolkits.mplot3d import Axes3D
from matplotlib import pyplot as plt
import numpy as np
from itertools import product,combinations
%matplotlib inline


    

In [ ]:

    

from nilearn.plotting import cm as nl_cm


    

In [4]:

    

import os,sys


    

In [17]:

    

import pandas as pd


    

In [5]:

    

sys.path.append('/media/sf_SharedFolder/Code/libraries_of_mine/github/ConWhAt')


    

In [6]:

    

import ConWhAt


    

In [7]:

    

from ConWhAt.volumetric.utils import get_bounding_box_inds,plot_cube_from_bb,plot_vol_scatter,get_intersection


    

In [8]:

    

# These are all now in volumetric/utils.py

"""
def get_bounding_box_inds(dat):
    
  nzx,nzy,nzz = np.nonzero(dat>0)
  xmin,xmax = nzx.min(),nzx.max()
  ymin,ymax = nzy.min(),nzy.max()
  zmin,zmax = nzz.min(),nzz.max()
    
  minmaxarr = np.array([[xmin,xmax],[ymin,ymax],[zmin,xmax]])    
    
  return minmaxarr
  
    
    
def plot_cube_from_bb(bb,ax=None,c='b'):

  corners = np.array(list(product(bb[0],
                                  bb[1],
                                  bb[2])))
    
  cornerpairs = list(combinations(corners,2))

  linestoplot = [(s,e) for (s,e) in cornerpairs if ((np.abs(s-e) == 0).sum() == 2)]
    

  if not ax:
    fig = plt.figure()
    ax = fig.gca(projection='3d')
    ax.set_aspect('equal')

    
  for (s,e) in linestoplot: 
    ax.plot3D(*zip(s,e), color=c) 
        
    
    
  
def plot_vol_scatter(dat,ax=None, pointint = 50,c='b',alpha=0.2,s=20.,
                     xlim=[0,100],ylim=[0,100],zlim=[0,100],marker='o',figsize=(10,10),
                     linewidth=0.01):
    
  xs,ys,zs = np.nonzero(dat>0)    
  idx = np.arange(0,xs.shape[0],pointint)

  if not ax: 
    fig = plt.figure(figsize=figsize)
    ax = fig.gca(projection='3d')
    ax.set_aspect('equal')
    ax.set_xlim([xlim[0],xlim[1]])  
    ax.set_ylim([ylim[0],ylim[1]])
    ax.set_zlim([zlim[0],zlim[1]])

  ax.scatter3D(xs[idx],ys[idx],zs[idx],c=c,alpha=alpha,s=s, marker=marker,linewidths=linewidth)
    
    
def get_intersection(bba,bbb):
    
  (xa1,xa2),(ya1,ya2),(za1,za2) = bba
  (xb1,xb2),(yb1,yb2),(zb1,zb2) = bbb
    
  SI = max(0, min(xa2,xb2) - max(xa1,xb1)) \
     * max(0, min(ya2,yb2) - max(ya1,yb1)) \
     * max(0, min(za2,zb2) - max(za1,zb1))

  return SI
""";


    

In [64]:

    

A_xyz = [[-5,10],
         [-11,21],
         [-20,30]]


B_xyz = [[-3,15],
         [-20,2],
         [-4,12]]

fig = plt.figure(figsize=(10,10))
ax = fig.gca(projection='3d')
#ax.set_aspect('equal')

plot_cube_from_bb(A_xyz,ax=ax,c='b')
plot_cube_from_bb(B_xyz,ax=ax,c='r')


xa1,xa2 = A_xyz[0]
ya1,ya2 = A_xyz[1]
za1,za2 = A_xyz[2]

xb1,xb2 = B_xyz[0]
yb1,yb2 = B_xyz[1]
zb1,zb2 = B_xyz[2]


SI = get_intersection(A_xyz,B_xyz)
SI


    

    
Out[64]:





2704






    

In [65]:

    

A_xyz = [[-5,10],
         [-11,21],
         [-20,30]]


B_xyz = [[15,30],
         [-20,-15],
         [-4,12]]

fig = plt.figure(figsize=(10,10))
ax = fig.gca(projection='3d')
#ax.set_aspect('equal')

plot_cube_from_bb(A_xyz,ax=ax,c='b')
plot_cube_from_bb(B_xyz,ax=ax,c='r')


xa1,xa2 = A_xyz[0]
ya1,ya2 = A_xyz[1]
za1,za2 = A_xyz[2]

xb1,xb2 = B_xyz[0]
yb1,yb2 = B_xyz[1]
zb1,zb2 = B_xyz[2]

SI = get_intersection(A_xyz,B_xyz)


SI


    

    
Out[65]:





0






    

In [68]:

    

A_xyz = [[-5,10],
         [-11,21],
         [-20,30]]


B_xyz = [[15,30],
         [-20,-15],
         [40,50]]

fig = plt.figure(figsize=(10,10))
ax = fig.gca(projection='3d')
ax.view_init(0,90)
#ax.set_aspect('equal')

plot_cube_from_bb(A_xyz,ax=ax,c='b')
plot_cube_from_bb(B_xyz,ax=ax,c='r')


xa1,xa2 = A_xyz[0]
ya1,ya2 = A_xyz[1]
za1,za2 = A_xyz[2]

xb1,xb2 = B_xyz[0]
yb1,yb2 = B_xyz[1]
zb1,zb2 = B_xyz[2]



SI = get_intersection(A_xyz,B_xyz)

ax.view_init(90,90)


SI


    

    
Out[68]:





0






    

In [9]:

    

test_file = '/media/sf_SharedFolder/Code/libraries_of_mine/github/ConWhAt/ConWhAt/scratch/lesion_test_file_from_jhu.nii.gz'


    

In [10]:

    

test_img = nib.load(test_file)

tbb = get_bounding_box_inds(test_img.get_data())


    

In [11]:

    

tbb


    

    
Out[11]:





array([[47, 59],
       [46, 91],
       [25, 59]])

In [12]:

    

f = '/usr/share/fsl/data/atlases/JHU/JHU-ICBM-tracts-prob-2mm.nii.gz'


    

In [13]:

    

jhu_img = nib.load(f)
jhu_bbs = []

for j in range(jhu_img.shape[3]):
  dat = index_img(jhu_img,j).get_data()
  bb = get_bounding_box_inds(dat)
  jhu_bbs.append(bb)


    

In [14]:

    

jhu_is = []

for jbb in jhu_bbs:
  I = get_intersection(tbb,jbb)
  jhu_is.append(I)


    

In [15]:

    

jhu_is


    

    
Out[15]:





[18360,
 10800,
 5304,
 3045,
 10530,
 0,
 8160,
 0,
 792,
 15504,
 18360,
 0,
 12240,
 0,
 14688,
 0,
 17136,
 0,
 18360,
 0]

In [16]:

    

jhu_overlaps = []

for j in range(jhu_img.shape[3]):

  overlap = ((index_img(jhu_img,j).get_data()>0) * (test_img.get_data()>0)).sum()
  jhu_overlaps.append(overlap)


    

In [18]:

    

df = pd.DataFrame([jhu_is,jhu_overlaps],index=['IS', 'overlap'])


    

In [19]:

    

df


    

    
Out[19]:






  

    

      

      
0
      
1
      
2
      
3
      
4
      
5
      
6
      
7
      
8
      
9
      
10
      
11
      
12
      
13
      
14
      
15
      
16
      
17
      
18
      
19
    
  
  

    

      
IS
      
18360
      
10800
      
5304
      
3045
      
10530
      
0
      
8160
      
0
      
792
      
15504
      
18360
      
0
      
12240
      
0
      
14688
      
0
      
17136
      
0
      
18360
      
0
    
    

      
overlap
      
1258
      
0
      
10
      
0
      
0
      
0
      
0
      
0
      
0
      
202
      
402
      
0
      
0
      
0
      
42
      
0
      
282
      
0
      
1
      
0
    
  

In [26]:

    

img1 = test_img
dat1 = img1.get_data()
dat1_bb = get_bounding_box_inds(dat1)


    

In [39]:

    

# Get bounding box
volnum = 0
img2 = index_img(jhu_img,volnum)
dat2 = img2.get_data()
dat2_bb = get_bounding_box_inds(dat2)


# Plot with nilearn
display = plot_stat_map(img2,cmap= nl_cm.black_blue,alpha=0.5,colorbar=False)
display.add_overlay(img1,colorbar=False,alpha=0.5,cmap='hot')

# Plot bounding boxes
fig = plt.figure(figsize=(10,10))
ax = fig.gca(projection='3d')
ax.set_aspect('equal')
ax.set_xlim([0,100])  
ax.set_ylim([0,100])
ax.set_zlim([0,100])

plot_vol_scatter(dat1,ax=ax,c='b',alpha=1.,s=20,marker='o')
plot_vol_scatter(dat2,ax=ax,c='r',alpha=1.,s=20,marker='o')

plot_cube_from_bb(dat1_bb,ax=ax,c='b')
plot_cube_from_bb(dat2_bb,ax=ax,c='r')

plt.tight_layout()

print df[volnum]


    

    




IS         18360
overlap     1258
Name: 0, dtype: int64






    













    

In [40]:

    

# Get bounding box
volnum = 1
img2 = index_img(jhu_img,volnum)
dat2 = img2.get_data()
dat2_bb = get_bounding_box_inds(dat2)


# Plot with nilearn
display = plot_stat_map(img2,cmap= nl_cm.black_blue,alpha=0.5,colorbar=False)
display.add_overlay(img1,colorbar=False,alpha=0.5,cmap='hot')

# Plot bounding boxes
fig = plt.figure(figsize=(10,10))
ax = fig.gca(projection='3d')
ax.set_aspect('equal')
ax.set_xlim([0,100])  
ax.set_ylim([0,100])
ax.set_zlim([0,100])

plot_vol_scatter(dat1,ax=ax,c='b',alpha=1.,s=20,marker='o')
plot_vol_scatter(dat2,ax=ax,c='r',alpha=1.,s=20,marker='o')

plot_cube_from_bb(dat1_bb,ax=ax,c='b')
plot_cube_from_bb(dat2_bb,ax=ax,c='r')

plt.tight_layout()


print df[volnum]


    

    




IS         10800
overlap        0
Name: 1, dtype: int64






    













    

In [41]:

    

# Get bounding box
volnum = 2
img2 = index_img(jhu_img,volnum)
dat2 = img2.get_data()
dat2_bb = get_bounding_box_inds(dat2)


# Plot with nilearn
display = plot_stat_map(img2,cmap= nl_cm.black_blue,alpha=0.5,colorbar=False)
display.add_overlay(img1,colorbar=False,alpha=0.5,cmap='hot')

# Plot bounding boxes
fig = plt.figure(figsize=(10,10))
ax = fig.gca(projection='3d')
ax.set_aspect('equal')
ax.set_xlim([0,100])  
ax.set_ylim([0,100])
ax.set_zlim([0,100])

plot_vol_scatter(dat1,ax=ax,c='b',alpha=1.,s=20,marker='o')
plot_vol_scatter(dat2,ax=ax,c='r',alpha=1.,s=20,marker='o')

plot_cube_from_bb(dat1_bb,ax=ax,c='b')
plot_cube_from_bb(dat2_bb,ax=ax,c='r')

plt.tight_layout()

print df[volnum]


    

    




IS         5304
overlap      10
Name: 2, dtype: int64






    













    

In [31]:

    

# Get bounding box
volnum = 3
img2 = index_img(jhu_img,volnum)
dat2 = img2.get_data()
dat2_bb = get_bounding_box_inds(dat2)


# Plot with nilearn
display = plot_stat_map(img2,cmap= nl_cm.black_blue,alpha=0.5,colorbar=False)
display.add_overlay(img1,colorbar=False,alpha=0.5,cmap='hot')

# Plot bounding boxes
fig = plt.figure(figsize=(10,10))
ax = fig.gca(projection='3d')
ax.set_aspect('equal')
ax.set_xlim([0,100])  
ax.set_ylim([0,100])
ax.set_zlim([0,100])

plot_vol_scatter(dat1,ax=ax,c='b',alpha=1.,s=20,marker='o')
plot_vol_scatter(dat2,ax=ax,c='r',alpha=1.,s=20,marker='o')

plot_cube_from_bb(dat1_bb,ax=ax,c='b')
plot_cube_from_bb(dat2_bb,ax=ax,c='r')

plt.tight_layout()


    

In [42]:

    

# Get bounding box
volnum = 4
img2 = index_img(jhu_img,volnum)
dat2 = img2.get_data()
dat2_bb = get_bounding_box_inds(dat2)


# Plot with nilearn
display = plot_stat_map(img2,cmap= nl_cm.black_blue,alpha=0.5,colorbar=False)
display.add_overlay(img1,colorbar=False,alpha=0.5,cmap='hot')

# Plot bounding boxes
fig = plt.figure(figsize=(10,10))
ax = fig.gca(projection='3d')
ax.set_aspect('equal')
ax.set_xlim([0,100])  
ax.set_ylim([0,100])
ax.set_zlim([0,100])

plot_vol_scatter(dat1,ax=ax,c='b',alpha=1.,s=20,marker='o')
plot_vol_scatter(dat2,ax=ax,c='r',alpha=1.,s=20,marker='o')

plot_cube_from_bb(dat1_bb,ax=ax,c='b')
plot_cube_from_bb(dat2_bb,ax=ax,c='r')

plt.tight_layout()


print df[volnum]


    

    




IS         10530
overlap        0
Name: 4, dtype: int64






    













    

In [50]:

    

# Get bounding box
volnum = 5
img2 = index_img(jhu_img,volnum)
dat2 = img2.get_data()
dat2_bb = get_bounding_box_inds(dat2)


# Plot with nilearn
display = plot_stat_map(img2,cmap= nl_cm.black_blue,alpha=0.5,colorbar=False)
display.add_overlay(img1,colorbar=False,alpha=0.5,cmap='hot',vmax=5.)

# Plot bounding boxes
fig = plt.figure(figsize=(10,10))
ax = fig.gca(projection='3d')
ax.set_aspect('equal')
ax.set_xlim([0,100])  
ax.set_ylim([0,100])
ax.set_zlim([0,100])

plot_vol_scatter(dat1,ax=ax,c='b',alpha=1.,s=20,marker='o')
plot_vol_scatter(dat2,ax=ax,c='r',alpha=1.,s=20,marker='o')

plot_cube_from_bb(dat1_bb,ax=ax,c='b')
plot_cube_from_bb(dat2_bb,ax=ax,c='r')

plt.tight_layout()

print df[volnum]


    

    




IS         0
overlap    0
Name: 5, dtype: int64






    













    

In [43]:

    

# Get bounding box
volnum = 6
img2 = index_img(jhu_img,volnum)
dat2 = img2.get_data()
dat2_bb = get_bounding_box_inds(dat2)


# Plot with nilearn
display = plot_stat_map(img2,cmap= nl_cm.black_blue,alpha=0.5,colorbar=False)
display.add_overlay(img1,colorbar=False,alpha=0.5,cmap='hot')

# Plot bounding boxes
fig = plt.figure(figsize=(10,10))
ax = fig.gca(projection='3d')
ax.set_aspect('equal')
ax.set_xlim([0,100])  
ax.set_ylim([0,100])
ax.set_zlim([0,100])

plot_vol_scatter(dat1,ax=ax,c='b',alpha=1.,s=20,marker='o')
plot_vol_scatter(dat2,ax=ax,c='r',alpha=1.,s=20,marker='o')

plot_cube_from_bb(dat1_bb,ax=ax,c='b')
plot_cube_from_bb(dat2_bb,ax=ax,c='r')

plt.tight_layout()


print df[volnum]


    

    




IS         8160
overlap       0
Name: 6, dtype: int64






    













    

In [47]:

    

# Get bounding box
volnum = 7
img2 = index_img(jhu_img,volnum)
dat2 = img2.get_data()
dat2_bb = get_bounding_box_inds(dat2)


# Plot with nilearn
display = plot_stat_map(img2,cmap= nl_cm.black_blue,alpha=0.5,colorbar=False)
display.add_overlay(img1,colorbar=False,alpha=0.5,cmap='hot')

# Plot bounding boxes
fig = plt.figure(figsize=(10,10))
ax = fig.gca(projection='3d')
ax.set_aspect('equal')
ax.set_xlim([0,100])  
ax.set_ylim([0,100])
ax.set_zlim([0,100])

plot_vol_scatter(dat1,ax=ax,c='b',alpha=1.,s=20,marker='o')
plot_vol_scatter(dat2,ax=ax,c='r',alpha=1.,s=20,marker='o')

plot_cube_from_bb(dat1_bb,ax=ax,c='b')
plot_cube_from_bb(dat2_bb,ax=ax,c='r')

plt.tight_layout()

print df[volnum]


    

    




IS         0
overlap    0
Name: 7, dtype: int64






    













    

In [44]:

    

# Get bounding box
volnum = 8
img2 = index_img(jhu_img,volnum)
dat2 = img2.get_data()
dat2_bb = get_bounding_box_inds(dat2)


# Plot with nilearn
display = plot_stat_map(img2,cmap= nl_cm.black_blue,alpha=0.5,colorbar=False)
display.add_overlay(img1,colorbar=False,alpha=0.5,cmap='hot')

# Plot bounding boxes
fig = plt.figure(figsize=(10,10))
ax = fig.gca(projection='3d')
ax.set_aspect('equal')
ax.set_xlim([0,100])  
ax.set_ylim([0,100])
ax.set_zlim([0,100])

plot_vol_scatter(dat1,ax=ax,c='b',alpha=1.,s=20,marker='o')
plot_vol_scatter(dat2,ax=ax,c='r',alpha=1.,s=20,marker='o')

plot_cube_from_bb(dat1_bb,ax=ax,c='b')
plot_cube_from_bb(dat2_bb,ax=ax,c='r')

plt.tight_layout()


print df[volnum]


    

    




IS         792
overlap      0
Name: 8, dtype: int64






    













    

In [46]:

    

# Get bounding box
volnum = 9
img2 = index_img(jhu_img,volnum)
dat2 = img2.get_data()
dat2_bb = get_bounding_box_inds(dat2)


# Plot with nilearn
display = plot_stat_map(img2,cmap= nl_cm.black_blue,alpha=0.5,colorbar=False)
display.add_overlay(img1,colorbar=False,alpha=0.5,cmap='hot')

# Plot bounding boxes
fig = plt.figure(figsize=(10,10))
ax = fig.gca(projection='3d')
ax.set_aspect('equal')
ax.set_xlim([0,100])  
ax.set_ylim([0,100])
ax.set_zlim([0,100])

plot_vol_scatter(dat1,ax=ax,c='b',alpha=1.,s=20,marker='o')
plot_vol_scatter(dat2,ax=ax,c='r',alpha=1.,s=20,marker='o')

plot_cube_from_bb(dat1_bb,ax=ax,c='b')
plot_cube_from_bb(dat2_bb,ax=ax,c='r')

plt.tight_layout()
print df[volnum]


    

    




IS         15504
overlap      202
Name: 9, dtype: int64






    













    

In [51]:

    

# Get bounding box
volnum = 10
img2 = index_img(jhu_img,volnum)
dat2 = img2.get_data()
dat2_bb = get_bounding_box_inds(dat2)


# Plot with nilearn
display = plot_stat_map(img2,cmap= nl_cm.black_blue,alpha=0.5,colorbar=False)
display.add_overlay(img1,colorbar=False,alpha=0.5,cmap='hot')

# Plot bounding boxes
fig = plt.figure(figsize=(10,10))
ax = fig.gca(projection='3d')
ax.set_aspect('equal')
ax.set_xlim([0,100])  
ax.set_ylim([0,100])
ax.set_zlim([0,100])

plot_vol_scatter(dat1,ax=ax,c='b',alpha=1.,s=20,marker='o')
plot_vol_scatter(dat2,ax=ax,c='r',alpha=1.,s=20,marker='o')

plot_cube_from_bb(dat1_bb,ax=ax,c='b')
plot_cube_from_bb(dat2_bb,ax=ax,c='r')

plt.tight_layout()


print df[volnum]


    

    




IS         18360
overlap      402
Name: 10, dtype: int64






    













    

In [54]:

    

# Get bounding box
volnum = 11
img2 = index_img(jhu_img,volnum)
dat2 = img2.get_data()
dat2_bb = get_bounding_box_inds(dat2)


# Plot with nilearn
display = plot_stat_map(img2,cmap= nl_cm.black_blue,alpha=0.5,colorbar=False)
display.add_overlay(img1,colorbar=False,alpha=0.5,cmap='hot')

# Plot bounding boxes
fig = plt.figure(figsize=(10,10))
ax = fig.gca(projection='3d')
ax.set_aspect('equal')
ax.set_xlim([0,100])  
ax.set_ylim([0,100])
ax.set_zlim([0,100])

plot_vol_scatter(dat1,ax=ax,c='b',alpha=1.,s=20,marker='o')
plot_vol_scatter(dat2,ax=ax,c='r',alpha=1.,s=20,marker='o')

plot_cube_from_bb(dat1_bb,ax=ax,c='b')
plot_cube_from_bb(dat2_bb,ax=ax,c='r')

plt.tight_layout()


print df[volnum]


    

    




IS         0
overlap    0
Name: 11, dtype: int64






    













    

In [55]:

    

# Get bounding box
volnum = 13
img2 = index_img(jhu_img,volnum)
dat2 = img2.get_data()
dat2_bb = get_bounding_box_inds(dat2)


# Plot with nilearn
display = plot_stat_map(img2,cmap= nl_cm.black_blue,alpha=0.5,colorbar=False)
display.add_overlay(img1,colorbar=False,alpha=0.5,cmap='hot')

# Plot bounding boxes
fig = plt.figure(figsize=(10,10))
ax = fig.gca(projection='3d')
ax.set_aspect('equal')
ax.set_xlim([0,100])  
ax.set_ylim([0,100])
ax.set_zlim([0,100])

plot_vol_scatter(dat1,ax=ax,c='b',alpha=1.,s=20,marker='o')
plot_vol_scatter(dat2,ax=ax,c='r',alpha=1.,s=20,marker='o')

plot_cube_from_bb(dat1_bb,ax=ax,c='b')
plot_cube_from_bb(dat2_bb,ax=ax,c='r')

plt.tight_layout()


print df[volnum]


    

    




IS         0
overlap    0
Name: 13, dtype: int64






    













    

In [56]:

    

# Get bounding box
volnum = 13
img2 = index_img(jhu_img,volnum)
dat2 = img2.get_data()
dat2_bb = get_bounding_box_inds(dat2)


# Plot with nilearn
display = plot_stat_map(img2,cmap= nl_cm.black_blue,alpha=0.5,colorbar=False)
display.add_overlay(img1,colorbar=False,alpha=0.5,cmap='hot')

# Plot bounding boxes
fig = plt.figure(figsize=(10,10))
ax = fig.gca(projection='3d')
ax.set_aspect('equal')
ax.set_xlim([0,100])  
ax.set_ylim([0,100])
ax.set_zlim([0,100])

plot_vol_scatter(dat1,ax=ax,c='b',alpha=1.,s=20,marker='o')
plot_vol_scatter(dat2,ax=ax,c='r',alpha=1.,s=20,marker='o')

plot_cube_from_bb(dat1_bb,ax=ax,c='b')
plot_cube_from_bb(dat2_bb,ax=ax,c='r')

plt.tight_layout()


print df[volnum]


    

    




IS         0
overlap    0
Name: 13, dtype: int64






    













    

In [57]:

    

# Get bounding box
volnum = 14
img2 = index_img(jhu_img,volnum)
dat2 = img2.get_data()
dat2_bb = get_bounding_box_inds(dat2)


# Plot with nilearn
display = plot_stat_map(img2,cmap= nl_cm.black_blue,alpha=0.5,colorbar=False)
display.add_overlay(img1,colorbar=False,alpha=0.5,cmap='hot')

# Plot bounding boxes
fig = plt.figure(figsize=(10,10))
ax = fig.gca(projection='3d')
ax.set_aspect('equal')
ax.set_xlim([0,100])  
ax.set_ylim([0,100])
ax.set_zlim([0,100])

plot_vol_scatter(dat1,ax=ax,c='b',alpha=1.,s=20,marker='o')
plot_vol_scatter(dat2,ax=ax,c='r',alpha=1.,s=20,marker='o')

plot_cube_from_bb(dat1_bb,ax=ax,c='b')
plot_cube_from_bb(dat2_bb,ax=ax,c='r')

plt.tight_layout()


print df[volnum]


    

    




IS         14688
overlap       42
Name: 14, dtype: int64






    













    

In [58]:

    

# Get bounding box
volnum = 15
img2 = index_img(jhu_img,volnum)
dat2 = img2.get_data()
dat2_bb = get_bounding_box_inds(dat2)


# Plot with nilearn
display = plot_stat_map(img2,cmap= nl_cm.black_blue,alpha=0.5,colorbar=False)
display.add_overlay(img1,colorbar=False,alpha=0.5,cmap='hot')

# Plot bounding boxes
fig = plt.figure(figsize=(10,10))
ax = fig.gca(projection='3d')
ax.set_aspect('equal')
ax.set_xlim([0,100])  
ax.set_ylim([0,100])
ax.set_zlim([0,100])

plot_vol_scatter(dat1,ax=ax,c='b',alpha=1.,s=20,marker='o')
plot_vol_scatter(dat2,ax=ax,c='r',alpha=1.,s=20,marker='o')

plot_cube_from_bb(dat1_bb,ax=ax,c='b')
plot_cube_from_bb(dat2_bb,ax=ax,c='r')

plt.tight_layout()


print df[volnum]


    

    




IS         0
overlap    0
Name: 15, dtype: int64






    













    

In [59]:

    

# Get bounding box
volnum = 16
img2 = index_img(jhu_img,volnum)
dat2 = img2.get_data()
dat2_bb = get_bounding_box_inds(dat2)


# Plot with nilearn
display = plot_stat_map(img2,cmap= nl_cm.black_blue,alpha=0.5,colorbar=False)
display.add_overlay(img1,colorbar=False,alpha=0.5,cmap='hot')

# Plot bounding boxes
fig = plt.figure(figsize=(10,10))
ax = fig.gca(projection='3d')
ax.set_aspect('equal')
ax.set_xlim([0,100])  
ax.set_ylim([0,100])
ax.set_zlim([0,100])

plot_vol_scatter(dat1,ax=ax,c='b',alpha=1.,s=20,marker='o')
plot_vol_scatter(dat2,ax=ax,c='r',alpha=1.,s=20,marker='o')

plot_cube_from_bb(dat1_bb,ax=ax,c='b')
plot_cube_from_bb(dat2_bb,ax=ax,c='r')

plt.tight_layout()


print df[volnum]


    

    




IS         17136
overlap      282
Name: 16, dtype: int64






    













    

In [60]:

    

# Get bounding box
volnum = 17
img2 = index_img(jhu_img,volnum)
dat2 = img2.get_data()
dat2_bb = get_bounding_box_inds(dat2)


# Plot with nilearn
display = plot_stat_map(img2,cmap= nl_cm.black_blue,alpha=0.5,colorbar=False)
display.add_overlay(img1,colorbar=False,alpha=0.5,cmap='hot')

# Plot bounding boxes
fig = plt.figure(figsize=(10,10))
ax = fig.gca(projection='3d')
ax.set_aspect('equal')
ax.set_xlim([0,100])  
ax.set_ylim([0,100])
ax.set_zlim([0,100])

plot_vol_scatter(dat1,ax=ax,c='b',alpha=1.,s=20,marker='o')
plot_vol_scatter(dat2,ax=ax,c='r',alpha=1.,s=20,marker='o')

plot_cube_from_bb(dat1_bb,ax=ax,c='b')
plot_cube_from_bb(dat2_bb,ax=ax,c='r')

plt.tight_layout()


print df[volnum]


    

    




IS         0
overlap    0
Name: 17, dtype: int64






    













    

In [61]:

    

# Get bounding box
volnum = 18
img2 = index_img(jhu_img,volnum)
dat2 = img2.get_data()
dat2_bb = get_bounding_box_inds(dat2)


# Plot with nilearn
display = plot_stat_map(img2,cmap= nl_cm.black_blue,alpha=0.5,colorbar=False)
display.add_overlay(img1,colorbar=False,alpha=0.5,cmap='hot')

# Plot bounding boxes
fig = plt.figure(figsize=(10,10))
ax = fig.gca(projection='3d')
ax.set_aspect('equal')
ax.set_xlim([0,100])  
ax.set_ylim([0,100])
ax.set_zlim([0,100])

plot_vol_scatter(dat1,ax=ax,c='b',alpha=1.,s=20,marker='o')
plot_vol_scatter(dat2,ax=ax,c='r',alpha=1.,s=20,marker='o')

plot_cube_from_bb(dat1_bb,ax=ax,c='b')
plot_cube_from_bb(dat2_bb,ax=ax,c='r')

plt.tight_layout()


print df[volnum]


    

    




IS         18360
overlap        1
Name: 18, dtype: int64






    













    

In [62]:

    

# Get bounding box
volnum = 19
img2 = index_img(jhu_img,volnum)
dat2 = img2.get_data()
dat2_bb = get_bounding_box_inds(dat2)


# Plot with nilearn
display = plot_stat_map(img2,cmap= nl_cm.black_blue,alpha=0.5,colorbar=False)
display.add_overlay(img1,colorbar=False,alpha=0.5,cmap='hot')

# Plot bounding boxes
fig = plt.figure(figsize=(10,10))
ax = fig.gca(projection='3d')
ax.set_aspect('equal')
ax.set_xlim([0,100])  
ax.set_ylim([0,100])
ax.set_zlim([0,100])

plot_vol_scatter(dat1,ax=ax,c='b',alpha=1.,s=20,marker='o')
plot_vol_scatter(dat2,ax=ax,c='r',alpha=1.,s=20,marker='o')

plot_cube_from_bb(dat1_bb,ax=ax,c='b')
plot_cube_from_bb(dat2_bb,ax=ax,c='r')

plt.tight_layout()


print df[volnum]


    

    




IS         0
overlap    0
Name: 19, dtype: int64