CImageProcessor/new_training_evaluating_setup.py at master · chiellini/CImageProcessor · 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
import os.path
import shutil

import numpy as np
from skimage.transform import resize, rescale
import glob

from utils.data_io import nib_load, nib_save
from utils.obj_visulization import combine_objs, rename_objs
from utils.utils import get_boundary


def resize_the_segcell_niigz():
    embryo_names = ['221017plc1p2']
    root_path = r'C:\Users\zelinli6\OneDrive - City University of Hong Kong - Student\MembraneProjectData\Embryo pre segmented\CMap'
    target_path = r'C:\Users\zelinli6\OneDrive - City University of Hong Kong - Student\MembraneProjectData\Embryo pre segmented\CMap_whole'
    target_shape_scale = 2
    for embryo_name in embryo_names:
        seg_file_paths = glob.glob(os.path.join(root_path, embryo_name, 'SegCell','*_segCell.nii.gz'))
        for seg_file_path in seg_file_paths:
            seg_array = nib_load(seg_file_path)
            resize_seg_array = rescale(seg_array, scale=target_shape_scale,preserve_range=True, mode='constant',order=0)
            name_tp=os.path.basename(seg_file_path)
            save_path = os.path.join(target_path, embryo_name,'SegCell', name_tp)
            nib_save(resize_seg_array, save_path)
            print(name_tp,seg_array.shape,'_->>>',resize_seg_array.shape)


def set_up_new_training_and_evaluation_data():
    # =================================================
    #  Set new training evaluation dataset
    # =================================================
    embryo_name_to_save='170704plc1p1'
    root_path=r'C:\Users\zelinli6\Downloads\Sample02\SegCell'
    target_path=r'C:\Users\zelinli6\Downloads\Sample02\FilppedZSegCell'
    niigz_paths=glob.glob(os.path.join(root_path,'*.nii.gz'))
    for niigz_path in niigz_paths:
        # print(niigz_path,os.path.basename(niigz_path))
        embryo_name,tp=os.path.basename(niigz_path).split('.')[0].split('_')[:2]
        arraythis=nib_load(niigz_path)
        # nib_save(os.path.join(target_path,os.path.basename(niigz)),array)

        arraythis=np.flip(arraythis,2)
        #
        img_stack = resize(image=arraythis, output_shape=(256,356,168), preserve_range=True, mode='constant',cval=0,order=0,anti_aliasing=False).astype(np.int16)

        nib_save(img_stack,os.path.join(target_path, "{}_{}_segCell.nii.gz".format(embryo_name_to_save, tp)))

    #-----------generate seg memb-----------------------------------------
    trainin_seg_folder = r"C:\Users\zelinli6\Downloads\Sample02\FilppedZSegCell"
    traingi_memb_dst_folder = r'C:\Users\zelinli6\Downloads\Sample02\FilppedZSegCell'
    # embryo_name = "170704plc1p1"

    seg_cell_file_paths = glob.glob(os.path.join(trainin_seg_folder,"*segCell.nii.gz"))

    for seg_file_path in seg_cell_file_paths:
        seg = nib_load(seg_file_path)
        memb = get_boundary(seg).astype(np.int16)
        embryo_name,tp=os.path.basename(seg_file_path).split('.')[0].split('_')[:2]
        # save_name_cell = os.path.join(dst_folder, "{}_{}_segCell.nii.gz".format(embryo_name, tp))
        save_seg_memb_path = os.path.join(traingi_memb_dst_folder, "{}_{}_segMemb.nii.gz".format(embryo_name_to_save, tp))
        # nib_save(save_name_cell, seg)
        nib_save(memb,save_seg_memb_path)

def resize_the_training_data():
    new_raw_shape256356160_path=r'C:\Users\zelinli6\Downloads\tem packed membrane nucleus'
    root_path = r'C:\Users\zelinli6\OneDrive - City University of Hong Kong - Student\MembraneProjectData\DataSource\TrainingandEvaluation\training'
    rawmemb_niigz_paths = glob.glob(os.path.join(root_path, 'RawMemb', '*rawMemb.nii.gz'))
    segmemb_niigz_paths = glob.glob(os.path.join(root_path, 'SegMemb', '*segMemb.nii.gz'))

    rawnuc_niigz_paths = glob.glob(os.path.join(root_path, 'RawNuc', '*rawNuc.nii.gz'))

    segcell_niigz_paths = glob.glob(os.path.join(root_path, 'SegCell', '*segCell.nii.gz'))

    for niigz in rawmemb_niigz_paths:
        embryo_name, tp = os.path.basename(niigz).split('.')[0].split('_')[:2]
        print(embryo_name, tp)
        raw_memb_shape = nib_load(niigz).shape
        # print()
        if raw_memb_shape[-1]<160:
            print('replacing with larger resolution')
            saving_rawmemb_path = os.path.join(root_path, 'RawMemb', '{}_{}_rawMemb.nii.gz'.format(embryo_name, tp))
            new_shape_rawmemb_path = os.path.join(new_raw_shape256356160_path, embryo_name, 'RawMemb',
                                                 '{}_{}_rawMemb.nii.gz'.format(embryo_name, tp))
            shutil.copy2(new_shape_rawmemb_path, saving_rawmemb_path)

            saving_rawnuc_path = os.path.join(root_path, 'RawNuc', '{}_{}_rawNuc.nii.gz'.format(embryo_name, tp))
            new_shape_rawnuc_path=os.path.join(new_raw_shape256356160_path,embryo_name,'RawNuc', '{}_{}_rawNuc.nii.gz'.format(embryo_name, tp))
            shutil.copy2(new_shape_rawnuc_path,saving_rawnuc_path)

            saving_seg_nucpath = os.path.join(root_path, 'SegNuc', '{}_{}_segNuc.nii.gz'.format(embryo_name, tp))
            new_shape_segnuc_path=os.path.join(new_raw_shape256356160_path,embryo_name,'SegNuc', '{}_{}_segNuc.nii.gz'.format(embryo_name, tp))
            shutil.copy2(new_shape_segnuc_path,saving_seg_nucpath)


            seg_cell_path = os.path.join(root_path, 'SegCell', '{}_{}_segCell.nii.gz'.format(embryo_name, tp))
            seg_cell_arr = nib_load(seg_cell_path)
            seg_cell_arr=resize(image=seg_cell_arr, output_shape=(256, 356, 168), preserve_range=True, mode='constant', cval=0, order=0,
                   anti_aliasing=False).astype(np.int16)
            nib_save(seg_cell_arr,seg_cell_path)

        seg_memb_path = os.path.join(root_path, 'SegMemb', '{}_{}_segMemb.nii.gz'.format(embryo_name, tp))
        seg_memb_arr = nib_load(seg_memb_path)
        if seg_memb_arr.shape[-1] < 160:
            seg_memb_arr = resize(image=seg_memb_arr, output_shape=(256, 356, 168), preserve_range=True,
                                  mode='constant', cval=0, order=0,
                                  anti_aliasing=False).astype(np.int16)
            nib_save(seg_memb_arr, seg_memb_path)


def check_new_training_evaluating_data():
    # ==============================================
    # check new training data and validation data
    # ==============================================
    root_path = r'C:\Users\zelinli6\OneDrive - City University of Hong Kong - Student\MembraneProjectData\DataSource\TrainingandEvaluation\training'
    rawmemb_niigz_paths = glob.glob(os.path.join(root_path, 'RawMemb', '*rawMemb.nii.gz'))
    segmemb_niigz_paths = glob.glob(os.path.join(root_path, 'SegMemb', '*segMemb.nii.gz'))

    rawnuc_niigz_paths = glob.glob(os.path.join(root_path, 'RawNuc', '*rawNuc.nii.gz'))

    segcell_niigz_paths = glob.glob(os.path.join(root_path, 'SegCell', '*segCell.nii.gz'))
    print(len(rawmemb_niigz_paths),len(segmemb_niigz_paths),len(segcell_niigz_paths),len(rawnuc_niigz_paths))
    assert len(rawnuc_niigz_paths)==len(segmemb_niigz_paths)
    assert len(segmemb_niigz_paths)==len(segcell_niigz_paths)
    assert len(segcell_niigz_paths)==len(rawnuc_niigz_paths)

    for niigz in rawmemb_niigz_paths:
        embryo_name, tp = os.path.basename(niigz).split('.')[0].split('_')[:2]
        print(embryo_name, tp)
        raw_memb_shape = nib_load(niigz).shape
        nuc_path = os.path.join(root_path, 'RawNuc', '{}_{}_rawNuc.nii.gz'.format(embryo_name, tp))
        raw_nuc_shape = nib_load(nuc_path).shape
        seg_cell_path = os.path.join(root_path, 'SegCell', '{}_{}_segCell.nii.gz'.format(embryo_name, tp))
        seg_cell_shape = nib_load(seg_cell_path).shape
        seg_memb_path = os.path.join(root_path, 'SegMemb', '{}_{}_segMemb.nii.gz'.format(embryo_name, tp))
        seg_memb = nib_load(seg_memb_path)
        print(embryo_name,tp, 'cell membrane pixel',(seg_memb!=0).sum(),(seg_memb==0).sum(),(seg_memb!=0).sum()/(seg_memb==0).sum())
        seg_memb_shape=seg_memb.shape


        seg_nuc_path = os.path.join(root_path, 'SegNuc', '{}_{}_segNuc.nii.gz'.format(embryo_name, tp))
        seg_nuc_shape = nib_load(seg_nuc_path).shape

        assert raw_memb_shape == raw_nuc_shape
        assert raw_nuc_shape == seg_nuc_shape
        assert seg_cell_shape == seg_memb_shape

        assert seg_cell_shape == seg_nuc_shape
        print(seg_nuc_shape)
        assert len(np.unique((nib_load(seg_nuc_path)))) > 2


if __name__ == "__main__":
    # resize_the_training_data()
    check_new_training_evaluating_data()