abdominal_registration/ImageTensorboard.py at main · MedicalImageAnalysisTutorials/abdominal_registration · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
# -*- coding: utf-8 -*-
"""
Created on Fri Jan 31 15:36:01 2020

@author: T_ESTIENNE
"""
import matplotlib.pyplot as plt
import abdominal_registration.utils as utils
import numpy as np

def plot_registration_results(gt_sample, pred_sample, batch, args,
                              moving_patient, reference_patient,
                              loss=None, deformed_mask=None):

    if args.deep_supervision:
        deformed_img = utils.to_numpy(args, pred_sample[0][2][-1])
        (deformable_grid, integrated_grid) = (utils.to_numpy(args, i) for i in pred_sample[0][:2])
    else:
        (deformable_grid, integrated_grid, deformed_img) = (utils.to_numpy(args, i) for i in pred_sample[0])

    loss = utils.to_numpy(args, loss)

    moving, reference = (utils.to_numpy(args, gt_sample[key]) for key in ['moving_ct', 'reference_ct'])

    if deformed_mask is None:
        reference_mask, moving_mask = None, None
    else:
        reference_mask, moving_mask = (utils.to_numpy(args, gt_sample[key]) for
                                       key in ['reference_mask', 'moving_mask'])
        deformed_mask = utils.to_numpy(args, deformed_mask)


    fig_registration = plot_registration(moving, reference,
                                         deformed_img, integrated_grid, batch,
                                         moving_patient, reference_patient,
                                         loss, moving_mask, reference_mask,
                                         deformed_mask)

    return fig_registration

def plot_registration(moving, reference,
                      deformed_img, grid, batch, moving_patient, reference_patient,
                      loss=None, moving_mask=None, reference_mask=None, deformed_mask=None):

    kwargs = {'cmap': 'gray'}

    titles = ['Source', 'Target', 'Deformed']

    if grid is not None:
        titles.append('Grid')

    if loss is not None:
        titles.append('Loss')

    if deformed_mask is not None:
        titles += ['Source Mask', 'Target Mask', 'Deformed_mask']

    nb_column = len(titles)

    fig, ax = plt.subplots(3, nb_column, gridspec_kw={'wspace': 0, 'hspace': 0.02,
                                              'top': 0.93, 'bottom': 0.01,
                                              'left': 0.01, 'right': 0.99},
                           figsize=(9.6, 7.2))

    x_slice = int(moving.shape[2] // 2)
    y_slice = int(moving.shape[3] // 2)
    z_slice = int(moving.shape[4] // 2)

    modality = 0

    ax[0, 0].imshow(moving[batch, modality, x_slice, :, :], **kwargs)
    ax[1, 0].imshow(moving[batch, modality, :, y_slice, :], **kwargs)
    ax[2, 0].imshow(moving[batch, modality, :, :, z_slice], **kwargs)

    ax[0, 1].imshow(reference[batch, modality, x_slice, :, :], **kwargs)
    ax[1, 1].imshow(reference[batch, modality, :, y_slice, :], **kwargs)
    ax[2, 1].imshow(reference[batch, modality, :, :, z_slice], **kwargs)

    ax[0, 2].imshow(deformed_img[batch, modality, x_slice, :, :], **kwargs)
    ax[1, 2].imshow(deformed_img[batch, modality, :, y_slice, :], **kwargs)
    ax[2, 2].imshow(deformed_img[batch, modality, :, :, z_slice], **kwargs)

    column = 3

    if grid is not None:
        dz, dy, dx = (grid[batch, 0, :, :, :],
                      grid[batch, 1, :, :, :],
                      grid[batch, 2, :, :, :])

        ax[0, column].contour(dy[x_slice, ::-1, :], 50, alpha=0.90, linewidths=0.5)
        ax[0, column].contour(dz[x_slice, ::-1, :], 50, alpha=0.90, linewidths=0.5)

        ax[1, column].contour(dx[:, y_slice, :], 50, alpha=0.90, linewidths=0.5)
        ax[1, column].contour(dz[:, y_slice, :], 50, alpha=0.90, linewidths=0.5)

        ax[2, column].contour(dx[:, :, z_slice], 50, alpha=0.90, linewidths=0.5)
        ax[2, column].contour(dy[:, :, z_slice], 50, alpha=0.90, linewidths=0.5)

        column += 1

    if loss is not None:
        ax[0, column].imshow(loss[batch, 0, x_slice, :, :])
        ax[1, column].imshow(loss[batch, 0, :, y_slice, :])
        ax[2, column].imshow(loss[batch, 0, :, :, z_slice])
        column += 1

    if deformed_mask is not None:

        vmax = moving_mask.shape[1] # Number of channels
        mask_kwargs = {'vmin':0, 'vmax':vmax}
        moving_mask = np.argmax(moving_mask, axis=1)
        reference_mask = np.argmax(reference_mask, axis=1)
        deformed_mask = np.argmax(deformed_mask, axis=1)

        ax[0, column].imshow(moving_mask[batch, x_slice, :, :], **mask_kwargs)
        ax[1, column].imshow(moving_mask[batch, :, y_slice, :], **mask_kwargs)
        ax[2, column].imshow(moving_mask[batch, :, :, z_slice], **mask_kwargs)
        column += 1

        ax[0, column].imshow(reference_mask[batch, x_slice, :, :], **mask_kwargs)
        ax[1, column].imshow(reference_mask[batch, :, y_slice, :], **mask_kwargs)
        ax[2, column].imshow(reference_mask[batch, :, :, z_slice], **mask_kwargs)
        column += 1

        ax[0, column].imshow(deformed_mask[batch, x_slice, :, :], **mask_kwargs)
        ax[1, column].imshow(deformed_mask[batch, :, y_slice, :], **mask_kwargs)
        ax[2, column].imshow(deformed_mask[batch, :, :, z_slice], **mask_kwargs)

    for j in range(nb_column):
        ax[0, j].set_title(titles[j])
        for i in range(3):
            ax[i, j].grid(False)
            ax[i, j].axis('off')
            ax[i, j].set_xticks([])
            ax[i, j].set_yticks([])


    fig.suptitle('Moving : {:8} Reference : {:8}'.format(moving_patient[batch],
                                                     reference_patient[batch]))
    fig.canvas.draw()

    return fig


def plot_segmentation_ct(ct, mask, mask_pred, batch, patients, args):

    fig, ax = plt.subplots(3, 3, gridspec_kw={'wspace': 0, 'hspace': 0.02,
                                              'top': 0.86, 'bottom': 0.01,
                                              'left': 0.01, 'right': 0.99})
    (ct, mask, mask_pred) = (utils.to_numpy(args, x) for x in (ct, mask,
                                                               mask_pred))
    kwargs = {'cmap': 'gray'}
    vmax = mask.shape[1] # Number of channels
    mask_kwargs = {'vmin':0, 'vmax':vmax}

    modality = 0

    mask = np.argmax(mask, axis=1)
    x, y, z = np.where(mask[batch, ...] > 0)
    mask_pred = np.argmax(mask_pred, axis=1)

    if len(x) == 0:
        x_slice = int(ct.shape[2] // 2)
        y_slice = int(ct.shape[3] // 2)
        z_slice = int(ct.shape[4] // 2)
    else:
        x_slice = (np.min(x) + np.max(x)) // 2
        y_slice = (np.min(y) + np.max(y)) // 2
        z_slice = (np.min(z) + np.max(z)) // 2

    ax[0, 0].imshow(ct[batch, modality, x_slice, :, :], **kwargs)
    ax[1, 0].imshow(ct[batch, modality, :, y_slice, :], **kwargs)
    ax[2, 0].imshow(ct[batch, modality, :, :, z_slice], **kwargs)

    ax[0, 1].imshow(mask[batch, x_slice, :, :], **mask_kwargs)
    ax[1, 1].imshow(mask[batch, :, y_slice, :], **mask_kwargs)
    ax[2, 1].imshow(mask[batch, :, :, z_slice], **mask_kwargs)

    ax[0, 2].imshow(mask_pred[batch, x_slice, :, :], **mask_kwargs)
    ax[1, 2].imshow(mask_pred[batch, :, y_slice, :], **mask_kwargs)
    ax[2, 2].imshow(mask_pred[batch, :, :, z_slice], **mask_kwargs)


    titles = ['CT', 'Mask GT', 'Moving Pred']

    for j in range(3):
        ax[0, j].set_title(titles[j])
        for i in range(3):
            ax[i, j].grid(False)
            ax[i, j].axis('off')
            ax[i, j].set_xticks([])
            ax[i, j].set_yticks([])

    fig.suptitle('Patient : {}'.format(patients[batch]))
    fig.canvas.draw()

    return fig