dcm转jpg，遇到的问题

最新推荐文章于 2023-06-07 11:45:50 发布

Kaia_H

最新推荐文章于 2023-06-07 11:45:50 发布

阅读量461

点赞数

文章标签：深度学习人工智能 python 计算机视觉

本文链接：https://blog.csdn.net/qq_44168134/article/details/125507353

版权

RIAS START软件使用

**问题：**序列名中含有冒号，不能进行拆分，拆分时会出现错误
**解决办法：**使用如下代码，选出我所要的序列，并重命名该序列，变成没有冒号的名字，并转成nii.gz格式

import os
import numpy as np
import SimpleITK as sitk

path = "D:/BaiduNetdiskDownload/2018_15T"
file_name_list = os.listdir(path)
size = len(file_name_list)
for i in range (size) :

    dir = file_name_list[i]
    # dir_path = "data/out_test/" + dir
    dir_path = path + "/" + dir
    out = "new"
    reader = sitk.ImageSeriesReader()
    series_IDs = reader.GetGDCMSeriesIDs(dir_path)
    file_reader = sitk.ImageFileReader()
    if series_IDs:
        for series in series_IDs:
            series_file_names =  reader.GetGDCMSeriesFileNames(dir_path, series)
                # 根据一个单张的dcm文件，读取这个series的metedata，即可以获取这个序列的描述符
            file_reader.SetFileName(series_file_names[0])
            file_reader.ReadImageInformation()
            series_description = file_reader.GetMetaData("0008|103e")
            dicom_itk = sitk.ReadImage(series_file_names)
            if series_description == "Water: Knee OSag PD FSEfw BYSY" :
                series_description = "Water Knee OSag PD FSEfw BYSY"
                print("正在保存序列：%s" % series_description)
                save_path = os.path.join(out, file_name_list[i] + ".nii.gz")
                sitk.WriteImage(dicom_itk, save_path)