-
Notifications
You must be signed in to change notification settings - Fork 2
/
test.py
146 lines (123 loc) · 6.24 KB
/
test.py
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
import monai
import logging
import os
import sys
import matplotlib.pyplot as plt
import torch
import numpy as np
import pandas as pd
from sklearn.metrics import confusion_matrix
from sklearn import metrics
import seaborn as sns
from sklearn.metrics import classification_report
import argparse
import torch.nn.functional as nnf
import monai
from monai.data import DataLoader, ImageDataset
from monai.transforms import (
Orientation,
Rotate90,
EnsureChannelFirst,
Compose,
ScaleIntensityRangePercentiles,
ResizeWithPadOrCrop
)
pin_memory = torch.cuda.is_available()
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
print("Cuda device: ",torch.cuda.current_device())
print(device)
logging.basicConfig(stream=sys.stdout, level=logging.INFO)
parser = argparse.ArgumentParser()
parser.add_argument("--path_data", type=str, default='./data', help="data path")
parser.add_argument("--path_checkpoint", type=str, default='./pretrained_model', help="path to checkpoint")
parser.add_argument("--path_result", type=str, default='./', help="path to save results")
parser.add_argument("--model", type=str, default='DenseNet121', help="classification model")
args = parser.parse_args()
os.makedirs(args.path_result, exist_ok=True)
########################################################################
#### Set data loader for testing ##############
########################################################################
print('##### Setting test data #####')
images_test = []
labels_test = []
images_path = args.path_data
test_classes = ['T1w_pre', 'T1w_art', 'T1w_ven', 'T1w_del', 'T2w', 'T2fs', 'DWI', 'ADC']
for item in test_classes:
vol_type_path = os.path.join(images_path, item)
vol_type_files = sorted(os.listdir(vol_type_path))
for ifileName in vol_type_files:
images_test.append(os.path.join(vol_type_path, ifileName))
labels_test.append(item)
# map target values to an integer, one-hot encode
labels_test = pd.Series(labels_test)
labels_test = labels_test.map({'T1w_pre': 0, 'T1w_art': 1, 'T1w_ven': 2, 'T1w_del': 3, 'T2w': 4, 'T2fs': 5, 'DWI': 6, 'ADC': 7}).to_numpy()
labels_test = torch.nn.functional.one_hot(torch.as_tensor(labels_test)).float()
labels_test_np = labels_test.cpu().numpy()
labels_test_df = pd.DataFrame(labels_test_np)
labels_test_df.to_csv(os.path.join(args.path_result, 'labels_test_df.csv'))
test_transforms = Compose([
ScaleIntensityRangePercentiles(1, 99, 0, 1, clip=True, relative=False),
EnsureChannelFirst(),
ResizeWithPadOrCrop((256, 256, 36), "symmetric", "constant"),
Orientation(axcodes='LAS'),
Rotate90(k=3, spatial_axes=(0, 1))
])
test_ds = ImageDataset(image_files=images_test, labels=labels_test, transform=test_transforms)
test_loader = DataLoader(test_ds, batch_size=1, num_workers=4, pin_memory=pin_memory)
print('-------- Total # of test data: %d' % len(labels_test))
########################################################################
#### Set model ##############
########################################################################
print('##### Setting 3D model: %s #####' % args.model)
if args.model == 'DenseNet121':
model = monai.networks.nets.DenseNet121(spatial_dims=3, in_channels=1, out_channels=8).to(device)
else:
raise NameError('Wrong network name.')
########################################################################
#### Test Models ##############
########################################################################
all_probs = []
for fold in range(1,6):
print('##### Test MRI_classifier (Fold %d) #####' % fold)
checkpoint_dir = os.path.join(args.path_checkpoint, 'best_metric_model_fold_%s.pth'%fold)
state_dict = torch.load(checkpoint_dir, map_location='cuda:0')
model.load_state_dict(state_dict)
model.eval()
step_test = 0
all_probabilities = torch.tensor([], dtype=torch.float32, device=device)
y_pred = torch.tensor([], dtype=torch.float32, device=device)
for test_data in test_loader:
step_test += 1
test_images, test_labels = test_data[0].to(device), test_data[1].to(device)
with torch.no_grad():
test_outputs = model(test_images)
y_pred = torch.cat([y_pred, test_outputs], dim=0)
batch_probabilities = nnf.softmax(test_outputs, dim=1)
all_probabilities = torch.cat([all_probabilities, batch_probabilities], 0)
predictions = float(test_outputs.argmax(dim=1)[0].cpu().detach().numpy())
actuals = float(test_labels.argmax(dim=1)[0].cpu().detach().numpy())
print('---%05d | label: %.1f | predict: %.1f' % (step_test, actuals, predictions))
all_probabilities_np = all_probabilities.cpu().numpy()
all_probs.append(all_probabilities_np)
########################################################################
#### Ensemble the probabilities from 5-fold models ##############
########################################################################
print('##### Results from ensembling the 5-fold models #####')
ensemble_probs = (all_probs[0] + all_probs[1] + all_probs[2] + all_probs[3] + all_probs[4])/5
ensemble_probs_df = pd.DataFrame(ensemble_probs)
ensemble_probs_df.to_csv(os.path.join(args.path_result,'ensemble_probs.csv'))
predictions = np.argmax(ensemble_probs, axis=1).tolist()
actuals = np.argmax(labels_test_np, axis=1).tolist()
labels = ['T1w-p', 'T1w-a', 'T1w-v', 'T1w-d', 'T2w', 'T2fs', 'DWI', 'ADC']
confusion_matrix_result = confusion_matrix(actuals, predictions, labels=[0, 1, 2, 3, 4, 5, 6, 7])
ax = sns.heatmap(confusion_matrix_result, annot= True, xticklabels = labels, yticklabels = labels, cmap='Blues', fmt='g')
ax.set(xlabel='Predicted', ylabel='Actual')
ax.xaxis.set_ticks_position('top')
ax.xaxis.set_label_position('top')
plt.title("Test Outcomes: Actual vs. Predicted")
plt.savefig(os.path.join(args.path_result,"ensemble_confusion_matrix.pdf"), format='pdf')
classification_report = classification_report(actuals, predictions, target_names=labels, digits=4, output_dict=True)
classification_report_df = pd.DataFrame(classification_report).transpose()
classification_report_df.to_csv(os.path.join(args.path_result, 'ensemble_classification_report.csv'))
print(classification_report_df)
print("Ensemble AUC: ",metrics.roc_auc_score(y_true=labels_test_np, y_score=ensemble_probs, average='weighted'))