Data/modify_label_3_parts.py

import os
import numpy as np
import SimpleITK as sitk
from pathlib import Path

def remap_labels(input_array):
    """
    重新映射标签值
    1-10 -> 1
    11-20 -> 2
    21-30 -> 3
    其他值保持不变
    """
    output_array = input_array.copy()
    
    # 1-10 -> 1
    mask_1_10 = (input_array >= 1) & (input_array <= 10)
    output_array[mask_1_10] = 1
    
    # 11-20 -> 2
    mask_11_20 = (input_array >= 11) & (input_array <= 20)
    output_array[mask_11_20] = 2
    
    # 21-30 -> 3
    mask_21_30 = (input_array >= 21) & (input_array <= 30)
    output_array[mask_21_30] = 3
    
    return output_array

def process_single_file(input_file_path, output_file_path):
    """处理单个mha文件"""
    try:
        print(f"正在处理: {input_file_path.name}")
        
        # 读取mha文件
        image = sitk.ReadImage(str(input_file_path))
        
        # 获取像素数据
        array = sitk.GetArrayFromImage(image)
        
        # 重新映射标签
        remapped_array = remap_labels(array)
        
        # 创建新的ITK图像，保持所有原始元数据
        new_image = sitk.GetImageFromArray(remapped_array)
        new_image.SetSpacing(image.GetSpacing())
        new_image.SetOrigin(image.GetOrigin())
        new_image.SetDirection(image.GetDirection())
        
        # 保存文件
        sitk.WriteImage(new_image, str(output_file_path))
        print(f"  ✅ 保存成功")
        
        return True
        
    except Exception as e:
        print(f"  ❌ 处理失败: {str(e)}")
        return False

def main():
    # 直接设置输入和输出路径
    input_folder = "/research/phd_y3/pelvic_project/Data/labels"  # 修改为你的文件夹A路径
    output_folder = "/research/phd_y3/pelvic_project/Data/3_part_labels"  # 修改为你的输出路径
    
    input_path = Path(input_folder)
    output_path = Path(output_folder)
    
    # 创建输出文件夹
    output_path.mkdir(parents=True, exist_ok=True)
    
    # 获取所有mha文件
    mha_files = list(input_path.glob("*.mha"))
    
    print(f"找到 {len(mha_files)} 个mha文件")
    print(f"输入路径: {input_folder}")
    print(f"输出路径: {output_folder}")
    print("="*50)
    
    # 批量处理
    success_count = 0
    for i, input_file in enumerate(mha_files, 1):
        print(f"[{i}/{len(mha_files)}]", end=" ")
        
        # 构建输出文件路径
        output_file = output_path / input_file.name
        
        # 处理文件
        if process_single_file(input_file, output_file):
            success_count += 1
    
    print("="*50)
    print(f"处理完成! 成功处理 {success_count}/{len(mha_files)} 个文件")

if __name__ == "__main__":
    main()