def make_prior(subj_list, output_dir):
"""
Create label probability map and the mask based on the sebject ID.
"""
print 'Create a whole-fROI mask and several probabilistic map for' + \
' each fROI.'
db_dir = r'/nfs/t2/atlas/database'
subj_num = len(subj_list)
for i in range(subj_num):
sid = subj_list[i]
subj_dir = os.path.join(db_dir, sid, 'face-object')
label_file = get_label_file(subj_dir)
label_data = nib.load(label_file).get_data()
img_header = nib.load(label_file).get_header()
if not i:
data_size = label_data.shape
prob_data = np.zeros((data_size[0],
data_size[1],
data_size[2],
2))
temp = label_data.copy()
temp[temp!=1] = 0
prob_data[..., 0] += temp
temp = label_data.copy()
temp[temp!=3] = 0
temp[temp!=0] = 1
prob_data[..., 1] += temp
# calculate probability map
prob_data = prob_data / subj_num
mask_data = prob_data.sum(axis=3)
mask_data[mask_data!=0] = 1
# save to file
prob_file = os.path.join(output_dir, 'prob.nii.gz')
mybase.save2nifti(prob_data, img_header, prob_file)
mask_file = os.path.join(output_dir, 'mask.nii.gz')
mybase.save2nifti(mask_data, img_header, mask_file)
# return data
return prob_data, mask_data
def clf_gen(para_idx, subj_list, zstat_data, roi_data, working_dir, roi_header):
"""
Train a new clfer based on the subject list and specified parameters.
"""
# configuration
iter_num = 5
c_value = np.logspace(-3, 3, 10)
gamma_value = np.logspace(-3, 3, 10)
# set c and gamma
c = c_value[para_idx/10]
gamma = gamma_value[np.mod(para_idx, 10)]
cv_idx = np.array_split(subj_list, iter_num)
for i in range(iter_num):
# divide subjects into train-set and test-set
subj_train = list(cv_idx)
subj_test = subj_train.pop(i)
subj_train = np.concatenate(subj_train)
# generate ROI mask and probabilioty map
mask_data, prob_data = autoroi.make_priori(subj_train, zstat_data, roi_data)
# generate training samples
train_samples = []
train_labels = []
for subj_idx in subj_train:
print 'Subject ' + str(subj_idx)
marker, result = autoroi.segment(zstat_data[..., subj_idx])
subj_samples, subj_labels, parcel_id = autoroi.sample_extract(
zstat_data[..., subj_idx],
roi_data[..., subj_idx],
mask_data,
prob_data,
marker,
result)
train_samples += subj_samples
train_labels += subj_labels
# train classifier
scaler, svc = autoroi.roi_svc(train_samples, train_labels, kernel='rbf',
c = c, gamma=gamma)
# test test classifier performance on the test dataset
test_subj_num = len(subj_test)
auto_roi_data = np.zeros([91, 109, 91, test_subj_num])
subj_count = 0
for subj_idx in subj_test:
marker, result = autoroi.segment(zstat_data[..., subj_idx])
subj_samples, subj_labels, parcel_id = autoroi.sample_extract(
zstat_data[..., subj_idx],
roi_data[..., subj_idx],
mask_data,
prob_data,
marker,
result)
# remove all parcels except samples
all_parcel_id = np.unique(marker)
all_parcel_id = all_parcel_id.tolist()
all_parcel_id.pop(0)
for idx in all_parcel_id:
if idx not in parcel_id:
result[result==idx] = 0
subj_samples = scaler.transform(subj_samples)
predict_labels = svc.predict(subj_samples)
parcel_num = len(parcel_id)
auto_res = auto_roi_data[..., subj_count]
for idx in range(parcel_num):
auto_res[result==parcel_id[idx]] = predict_labels[idx]
subj_count += 1
# save predicted ROI
roi_header.set_data_shape(auto_roi_data.shape)
out_file_dir = os.path.join(working_dir, 'parameter_' + str(para_idx))
if not os.path.exists(out_file_dir):
os.system('mkdir ' + out_file_dir)
out_file_name = os.path.join(out_file_dir, 'test_CV_iter_' + str(i) + '.nii.gz')
mybase.save2nifti(auto_roi_data, roi_header, out_file_name)
zstat_data = nib.load(zstat_file).get_data()
roi_list = [1, 2, 3, 4]
#out_dir = os.path.join(data_dir, 'peak_mask_1')
#out_dir = os.path.join(gcss_dir, 'peak_mask_1')
#out_dir = os.path.join(ma_dir, 'peak_mask_1')
out_dir = os.path.join(group08_dir, 'predicted_files', 'peak_mask_1')
for i in range(len(sessid)):
ind_label = label_data[..., i].copy()
ind_label = np.around(ind_label)
ind_zstat = zstat_data[..., i].copy()
peak_label = np.zeros(ind_label.shape)
for roi in roi_list:
tmp = ind_label.copy()
tmp[tmp!=roi] = 0
tmp[tmp==roi] = 1
if not tmp.sum():
continue
tmp_peak = np.zeros(tmp.shape)
tmp_zstat = tmp * ind_zstat
peak_coord = np.unravel_index(tmp_zstat.argmax(), tmp_zstat.shape)
tmp_peak = imtool.cube_roi(tmp_peak, peak_coord[0],
peak_coord[1], peak_coord[2],
1, roi)
peak_label += tmp_peak * tmp
out_file = os.path.join(out_dir, sessid[i] + '_atlas.nii.gz')
mybase.save2nifti(peak_label, header, out_file)
def clf_validate(cv_idx, workingdir, zstat_data, roi_data, roi_header):
"""
Train a new clfer based on the tuned parameters.
"""
# get train- and test-dataset
subj_num = 202
subj_list = np.arange(subj_num)
cv_subj_list = np.array_split(subj_list, 5)
train_subj = list(cv_subj_list)
test_subj = train_subj.pop(cv_idx)
train_subj = np.concatenate(train_subj)
# get tuned SVC parameters
cv_dir = os.path.join(workingdir, 'Outer_CV_' + str(cv_idx))
tuned_para_idx = get_tuned_parameter(cv_dir)
# parameter configuration
c_value = np.logspace(-3, 3, 10)
gamma_value = np.logspace(-3, 3, 10)
# set c and gamma
c = c_value[tuned_para_idx/10]
gamma = gamma_value[np.mod(tuned_para_idx, 10)]
print 'Tuned parameters: ',
print 'C - ' + str(c),
print ' Gamma - ' + str(gamma)
# -- start validation
# generate ROI mask and probabilioty map
mask_data, prob_data = autoroi.make_priori(train_subj, zstat_data, roi_data)
# generate training samples
train_samples = []
train_labels = []
for subj_idx in train_subj:
print 'Subject ' + str(subj_idx)
marker, result = autoroi.segment(zstat_data[..., subj_idx])
subj_samples, subj_labels, parcel_id = autoroi.sample_extract(
zstat_data[..., subj_idx],
roi_data[..., subj_idx],
mask_data,
prob_data,
marker,
result)
train_samples += subj_samples
train_labels += subj_labels
# train classifier
scaler, svc = autoroi.roi_svc(train_samples, train_labels, kernel='rbf',
c = c, gamma=gamma)
# test test classifier performance on the test dataset
test_subj_num = len(test_subj)
auto_roi_data = np.zeros([91, 109, 91, test_subj_num])
subj_count = 0
for subj_idx in test_subj:
marker, result = autoroi.segment(zstat_data[..., subj_idx])
subj_samples, subj_labels, parcel_id = autoroi.sample_extract(
zstat_data[..., subj_idx],
roi_data[..., subj_idx],
mask_data,
prob_data,
marker,
result)
# remove all parcels except samples
all_parcel_id = np.unique(marker)
all_parcel_id = all_parcel_id.tolist()
all_parcel_id.pop(0)
for idx in all_parcel_id:
if idx not in parcel_id:
result[result==idx] = 0
subj_samples = scaler.transform(subj_samples)
predict_labels = svc.predict(subj_samples)
parcel_num = len(parcel_id)
auto_res = auto_roi_data[..., subj_count]
for idx in range(parcel_num):
auto_res[result==parcel_id[idx]] = predict_labels[idx]
subj_count += 1
# save predicted ROI
roi_header.set_data_shape(auto_roi_data.shape)
out_file_name = os.path.join(cv_dir, 'auto_test_data.nii.gz')
mybase.save2nifti(auto_roi_data, roi_header, out_file_name)
#mybase.save2nifti(pred_data, header, out_file)
#start_num += sample_num
# probability map smoothing and save to the nifti files
prob_data = np.zeros((91, 109, 91, len(clf_classes)))
for k in range(len(clf_classes)):
prob_val = pred_prob[start_num:end_num, k]
prob_data[..., k] = arlib.write2array(coords, prob_val)
mask = np.sum(prob_data, axis=3)
sm_prob_data = arlib.smooth_data(prob_data, 1)
sm_pred_data = np.argmax(sm_prob_data, axis=3)
sm_pred_data[sm_pred_data==2] = 3
sm_pred_data = sm_pred_data * mask
out_file = os.path.join(pred_dir,
test_sessid[subj_idx]+'_pred.nii.gz')
mybase.save2nifti(sm_pred_data, header, out_file)
start_num += sample_num
print 'Mean CV score is %s'%(cv_score.mean())
print 'Mean FFA Dice is %s'%(ffa_dice.mean())
print 'Mean OFA Dice is %s'%(ofa_dice.mean())
##-- effect of sample numbers (subjects number)
#out_file = os.path.join(data_dir, 'subj_size_effect.txt')
#f = open(out_file, 'wb')
#f.write('subj_size,samples,label_1,label_3,score,ofa_dice,ffa_dice\n')
## subjects size range
#subj_num = range(20, 160, 10)
## repeat number
#repeat_num = 10
#