请将LiTS2017数据集的图像和标签处理成2D的png格式，并且标签需要拆分成灰度图，背景灰度值0，肝脏灰度值1，肝脏肿瘤灰度值2，给出完整代码，我的LiTS2017数据集路径为C:\Users\Administrator\Desktop\LiTS2017，里面包含Training Batch 1，其中里面有图像，图像格式为volume-0.nii。和Training Batch 2，其中里面有标签，标签格式为segmentation-0.nii，保存路径为C:\Users\Administrator\Desktop\2D-LiTS2017，里面有image和mask两个文件夹，请将转换好的保存到对应的文件夹

以下是将LiTS2017数据集的图像和标签处理成2D的png格式，并且标签拆分成灰度图的完整代码：

import os
import numpy as np
import nibabel as nib
from PIL import Image

# 读取nii文件
def read_nii_file(file_path):
    nii_file = nib.load(file_path)
    nii_data = nii_file.get_fdata()
    return nii_data

# 将3D图像转换成2D图像并保存为png格式
def convert_image_to_png(image_array, save_path):
    if not os.path.exists(save_path):
        os.makedirs(save_path)
    for i in range(image_array.shape[2]):
        image = image_array[:, :, i]
        image = Image.fromarray(image).convert('L')
        image.save(os.path.join(save_path, f'{i+1:03d}.png'))

# 将3D标签拆分成3个二值灰度图，并保存为png格式
def convert_label_to_png(label_array, save_path):
    if not os.path.exists(save_path):
        os.makedirs(save_path)
    for i in range(label_array.shape[2]):
        label = label_array[:, :, i]
        liver = np.zeros(label.shape)
        tumor = np.zeros(label.shape)
        liver[label == 1] = 1
        tumor[label == 2] = 1
        combined_label = np.stack([liver, tumor], axis=-1)
        combined_label = (combined_label * 255).astype(np.uint8)
        Image.fromarray(combined_label).save(os.path.join(save_path, f'{i+1:03d}.png'))

# LiTS2017数据集路径
data_path = r'C:\Users\Administrator\Desktop\LiTS2017'

# Training Batch 1的图像路径
image_path = os.path.join(data_path, 'Training Batch 1', 'volume-0.nii')
image_array = read_nii_file(image_path)
image_save_path = r'C:\Users\Administrator\Desktop\2D-LiTS2017\image'
convert_image_to_png(image_array, image_save_path)

# Training Batch 2的标签路径
label_path = os.path.join(data_path, 'Training Batch 2', 'segmentation-0.nii')
label_array = read_nii_file(label_path)
label_save_path = r'C:\Users\Administrator\Desktop\2D-LiTS2017\mask'
convert_label_to_png(label_array, label_save_path)

执行上述代码后，将会在C:\Users\Administrator\Desktop\2D-LiTS2017目录下生成image和mask两个文件夹，分别包含转换好的2D图像和标签。其中，图像的文件名为001.png、002.png、003.png，依次类推；标签的文件名也是一样的。