def infer(start: list): """ :param start: Initial point :return: Moving position, the index of maximum confidence direction, Current termination probability """ max_z = re_spacing_img.shape[0] max_x = re_spacing_img.shape[1] max_y = re_spacing_img.shape[2] cut_size = 9 spacing_x = spacing[0] spacing_y = spacing[1] spacing_z = spacing[2] center_x_pixel = get_spacing_res2(start[0], spacing_x, resize_factor[1]) center_y_pixel = get_spacing_res2(start[1], spacing_y, resize_factor[2]) center_z_pixel = get_spacing_res2(start[2], spacing_z, resize_factor[0]) left_x = center_x_pixel - cut_size right_x = center_x_pixel + cut_size left_y = center_y_pixel - cut_size right_y = center_y_pixel + cut_size left_z = center_z_pixel - cut_size right_z = center_z_pixel + cut_size new_patch = np.zeros( (cut_size * 2 + 1, cut_size * 2 + 1, cut_size * 2 + 1)) if not (left_x < 0 or right_x < 0 or left_y < 0 or right_y < 0 or left_z < 0 or right_z < 0 or left_x >= max_x or right_x >= max_x or left_y >= max_y or right_y >= max_y or left_z >= max_z or right_z >= max_z): for ind in range(left_z, right_z + 1): src_temp = re_spacing_img[ind].copy() new_patch[ind - left_z] = src_temp[left_y:right_y + 1, left_x:right_x + 1] input_data = data_preprocess(new_patch) inputs = input_data.to(device) outputs = infer_model(inputs.float()) outputs = outputs.view((len(input_data), max_points + 1)) outputs_1 = outputs[:, :len(outputs[0]) - 1] outputs_2 = outputs[:, -1] outputs_1 = torch.nn.functional.softmax(outputs_1, 1) indexs = np.argsort(outputs_1.cpu().detach().numpy()[0])[::-1] curr_prob = prob_terminates(outputs_1, max_points).cpu().detach().numpy()[0] curr_r = outputs_2.cpu().detach().numpy()[0] sx, sy, sz = get_shell(max_points, curr_r) return [sx, sy, sz], indexs, curr_r, curr_prob else: return None
def creat_data(max_points, path_name, spacing_path, gap_size, save_num): spacing_info = np.loadtxt(spacing_path, delimiter=",", dtype=np.float32) pre_ind_list = [] next_ind_list = [] radials_list = [] patch_name = [] i = save_num print("processing dataset %d" % i) image_pre_fix = path_name + '0' + str(i) + '/' + 'image' + '0' + str(i) file_name = image_pre_fix + '.nii.gz' src_array = sitk.GetArrayFromImage( sitk.ReadImage(file_name, sitk.sitkFloat32)) spacing_x = spacing_info[i][0] spacing_y = spacing_info[i][1] spacing_z = spacing_info[i][2] re_spacing_img, curr_spacing, resize_factor = resample( src_array, np.array([spacing_z, spacing_x, spacing_y]), np.array([0.5, 0.5, 0.5])) for v in range(4): print("processing vessel %d" % v) reference_path = path_name + '0' + str(i) + '/' + 'vessel' + str( v) + '/' + 'reference.txt' txt_data = np.loadtxt(reference_path, dtype=np.float32) center = txt_data[..., 0:3] radials_data = txt_data[..., 3] start_ind = get_start_ind(center, radials_data) end_ind = get_end_ind(center, radials_data) print("start ind:", start_ind) print("end ind:", end_ind) counter = 0 last_center_x_pixel = -1 last_center_y_pixel = -1 last_center_z_pixel = -1 for j in range(start_ind, end_ind + 1): # for j in range(start_ind, start_ind + 1): if j % gap_size == 0: print('j:', j) center_x = center[j][0] center_y = center[j][1] center_z = center[j][2] org_x_pixel = get_spacing_res2(center_x, spacing_x, resize_factor[1]) org_y_pixel = get_spacing_res2(center_y, spacing_y, resize_factor[2]) org_z_pixel = get_spacing_res2(center_z, spacing_z, resize_factor[0]) if org_x_pixel != last_center_x_pixel or org_y_pixel != last_center_y_pixel or org_z_pixel != last_center_z_pixel: print("last:", [ last_center_x_pixel, last_center_y_pixel, last_center_z_pixel ]) print("curr:", [org_x_pixel, org_y_pixel, org_z_pixel]) last_center_x_pixel = org_x_pixel last_center_y_pixel = org_y_pixel last_center_z_pixel = org_z_pixel radial = radials_data[j] pre_ind, next_ind = get_pre_next_point_ind( center, radials_data, j) if pre_ind != -1 and next_ind != -1: pre_x = center[pre_ind][0] pre_y = center[pre_ind][1] pre_z = center[pre_ind][2] next_x = center[next_ind][0] next_y = center[next_ind][1] next_z = center[next_ind][2] sx, sy, sz = get_shell(max_points, radial) shell_arr = np.zeros((len(sx), 3)) for s_ind in range(len(sx)): shell_arr[s_ind][0] = sx[s_ind] shell_arr[s_ind][1] = sy[s_ind] shell_arr[s_ind][2] = sz[s_ind] center_x_pixel = get_spacing_res2( center_x, spacing_x, resize_factor[1]) center_y_pixel = get_spacing_res2( center_y, spacing_y, resize_factor[2]) center_z_pixel = get_spacing_res2( center_z, spacing_z, resize_factor[0]) curr_c = [center_x, center_y, center_z] p = [pre_x, pre_y, pre_z] pre_sim = find_closer_point_angle(shell_arr, p, curr_c) p = [next_x, next_y, next_z] next_sim = find_closer_point_angle( shell_arr, p, curr_c) pre_ind_list.append(pre_sim) next_ind_list.append(next_sim) radials_list.append(radial) cut_size = 9 left_x = center_x_pixel - cut_size right_x = center_x_pixel + cut_size left_y = center_y_pixel - cut_size right_y = center_y_pixel + cut_size left_z = center_z_pixel - cut_size right_z = center_z_pixel + cut_size new_src_arr = np.zeros( (cut_size * 2 + 1, cut_size * 2 + 1, cut_size * 2 + 1)) for ind in range(left_z, right_z + 1): src_temp = re_spacing_img[ind].copy() new_src_arr[ind - left_z] = src_temp[left_y:right_y + 1, left_x:right_x + 1] folder_path = './patch_data/centerline_patch/no_offset/point_' + str( max_points) + '_gp_' + str( gap_size) + '/' + 'd' + str(i) if not os.path.exists(folder_path): os.makedirs(folder_path) record_name = 'centerline_patch/no_offset/point_' + str( max_points ) + '_gp_' + str(gap_size) + '/' + 'd' + str( i) + '/' + 'd_' + str(i) + '_' + 'v_' + str( v) + '_' + 'patch_%d' % counter + '.nii.gz' org_name = './patch_data/' + record_name out = sitk.GetImageFromArray(new_src_arr) sitk.WriteImage(out, org_name) patch_name.append(record_name) counter += 1 return pre_ind_list, next_ind_list, radials_list, patch_name
def creat_data(max_points, path_name, spacing_path, gap_size, save_num): ''' :param max_points: :param path_name: :param spacing_path: :param gap_size: :param save_num: :return: ''' spacing_info = np.loadtxt(spacing_path, delimiter=",", dtype=np.float32) pre_ind_list = [] next_ind_list = [] radials_list = [] patch_name = [] i = save_num print("processing dataset %d" % i) image_pre_fix = path_name + '0' + str(i) + '/' + 'image' + '0' + str(i) file_name = image_pre_fix + '.nii.gz' src_array = sitk.GetArrayFromImage( sitk.ReadImage(file_name, sitk.sitkFloat32)) spacing_x = spacing_info[i][0] spacing_y = spacing_info[i][1] spacing_z = spacing_info[i][2] re_spacing_img, curr_spacing, resize_factor = resample( src_array, np.array([spacing_z, spacing_x, spacing_y]), np.array([0.5, 0.5, 0.5])) curr_mean = np.array([0, 0, 0]) rotate_prob = 0.3 for v in range(4): # for v in range(1): print("processing vessel %d" % v) reference_path = path_name + '0' + str(i) + '/' + 'vessel' + str( v) + '/' + 'reference.txt' txt_data = np.loadtxt(reference_path, dtype=np.float32) center = txt_data[..., 0:3] radials_data = txt_data[..., 3] start_ind = get_start_ind(center, radials_data) end_ind = get_end_ind(center, radials_data) print("start ind:", start_ind) print("end ind:", end_ind) counter = 0 last_center_x_pixel = -1 last_center_y_pixel = -1 last_center_z_pixel = -1 # for j in range(start_ind, start_ind+1): for j in range(start_ind, end_ind + 1): if j % gap_size == 0: print('j:', j) center_x = center[j][0] center_y = center[j][1] center_z = center[j][2] org_x_pixel = get_spacing_res2(center_x, spacing_x, resize_factor[1]) org_y_pixel = get_spacing_res2(center_y, spacing_y, resize_factor[2]) org_z_pixel = get_spacing_res2(center_z, spacing_z, resize_factor[0]) if org_x_pixel != last_center_x_pixel or org_y_pixel != last_center_y_pixel or org_z_pixel != last_center_z_pixel: print("last:", [ last_center_x_pixel, last_center_y_pixel, last_center_z_pixel ]) print("curr:", [org_x_pixel, org_y_pixel, org_z_pixel]) last_center_x_pixel = org_x_pixel last_center_y_pixel = org_y_pixel last_center_z_pixel = org_z_pixel radial = radials_data[j] record_set = set() curr_conv = np.array([[radial * 0.25, 0.0, 0.0], [0.0, radial * 0.25, 0.0], [0.0, 0.0, radial * 0.25]]) # To then obtain an off-centerline sample, point x is translated using a random shift sampled from a 3D normal distribution with μ = 0.0, σ = 0.25r for k in range(10): off_center_x, off_center_y, off_center_z = np.random.multivariate_normal( mean=curr_mean, cov=curr_conv, size=1).T center_x_new = center_x + off_center_x[0] center_y_new = center_y + off_center_y[0] center_z_new = center_z + off_center_z[0] center_x_pixel = get_spacing_res2( center_x_new, spacing_x, resize_factor[1]) center_y_pixel = get_spacing_res2( center_y_new, spacing_y, resize_factor[2]) center_z_pixel = get_spacing_res2( center_z_new, spacing_z, resize_factor[0]) while True: if (center_x_pixel != org_x_pixel or center_y_pixel != org_y_pixel or center_z_pixel != org_z_pixel) and ( center_x_pixel, center_y_pixel, center_z_pixel) not in record_set: record_set.add((center_x_pixel, center_y_pixel, center_z_pixel)) break else: off_center_x, off_center_y, off_center_z = np.random.multivariate_normal( mean=curr_mean, cov=curr_conv, size=1).T center_x_new = center_x + off_center_x[0] center_y_new = center_y + off_center_y[0] center_z_new = center_z + off_center_z[0] center_x_pixel = get_spacing_res2( center_x_new, spacing_x, resize_factor[1]) center_y_pixel = get_spacing_res2( center_y_new, spacing_y, resize_factor[2]) center_z_pixel = get_spacing_res2( center_z_new, spacing_z, resize_factor[0]) new_radial_ind = get_new_radial_ind( center, [center_x_new, center_y_new, center_z_new]) new_radial = radials_data[new_radial_ind] sx, sy, sz = get_shell(max_points, new_radial) shell_arr = np.zeros((len(sx), 3)) for s_ind in range(len(sx)): shell_arr[s_ind][0] = sx[s_ind] shell_arr[s_ind][1] = sy[s_ind] shell_arr[s_ind][2] = sz[s_ind] pre_ind, next_ind = get_pre_next_point_ind( center, radials_data, new_radial_ind) # 只有找到了前一个点和后一个点才进入切割流程 if pre_ind != -1 and next_ind != -1: cut_size = 9 left_x = center_x_pixel - cut_size right_x = center_x_pixel + cut_size left_y = center_y_pixel - cut_size right_y = center_y_pixel + cut_size left_z = center_z_pixel - cut_size right_z = center_z_pixel + cut_size new_src_arr = np.zeros( (cut_size * 2 + 1, cut_size * 2 + 1, cut_size * 2 + 1)) for ind in range(left_z, right_z + 1): src_temp = re_spacing_img[ind].copy() new_src_arr[ind - left_z] = src_temp[left_y:right_y + 1, left_x:right_x + 1] if np.random.uniform() <= rotate_prob: curr_c = [ center_x_new, center_y_new, center_z_new ] new_src_arr, new_pre_cood, new_next_cood = rotate_augmentation( new_src_arr, pre_ind, next_ind, curr_c, center, angle_x=(-60. / 360 * 2. * np.pi, 60. / 360 * 2. * np.pi), angle_y=(-60. / 360 * 2. * np.pi, 60. / 360 * 2. * np.pi), angle_z=(-60. / 360 * 2. * np.pi, 60. / 360 * 2. * np.pi)) p = [ new_pre_cood[0], new_pre_cood[1], new_pre_cood[2] ] pre_sim = find_closer_point_angle( shell_arr, p, curr_c) p = [ new_next_cood[0], new_next_cood[1], new_next_cood[2] ] next_sim = find_closer_point_angle( shell_arr, p, curr_c) pre_ind_list.append(pre_sim) next_ind_list.append(next_sim) radials_list.append(new_radial) else: pre_x = center[pre_ind][0] pre_y = center[pre_ind][1] pre_z = center[pre_ind][2] next_x = center[next_ind][0] next_y = center[next_ind][1] next_z = center[next_ind][2] curr_c = [ center_x_new, center_y_new, center_z_new ] p = [pre_x, pre_y, pre_z] pre_sim = find_closer_point_angle( shell_arr, p, curr_c) p = [next_x, next_y, next_z] next_sim = find_closer_point_angle( shell_arr, p, curr_c) pre_ind_list.append(pre_sim) next_ind_list.append(next_sim) radials_list.append(new_radial) folder_path = './patch_data/centerline_patch/offset/point_' + str( max_points) + '_gp_' + str( gap_size) + '/' + 'd' + str(i) if not os.path.exists(folder_path): os.makedirs(folder_path) record_name = 'centerline_patch/offset/point_' + str( max_points ) + '_gp_' + str(gap_size) + '/' + 'd' + str( i) + '/' + 'd_' + str(i) + '_' + 'v_' + str( v) + '_' + 'patch_%d_' % counter + str( k) + '.nii.gz' org_name = './patch_data/' + record_name out = sitk.GetImageFromArray(new_src_arr) sitk.WriteImage(out, org_name) patch_name.append(record_name) counter += 1 return pre_ind_list, next_ind_list, radials_list, patch_name