[Python知识库]python 处理NIFTI文件去颅骨 ， 无损读写 ，并且重新名保存

这样操作能保存头文件 保存的文件能够放入 itksnap

def img2strip(imgpath,strippath,savepath):
    img = nib.load(imgpath)
    imgdata = img.get_fdata()
    stripdata = nib.load(strippath).get_fdata()

    newimgdata = imgdata * stripdata
    newimg = nib.Nifti1Image(newimgdata,img.affine)
    nib.save(newimg,savepath)```

rawpath = r'C:\Users\tangx\Desktop\raw'
sixpath = r'C:\Users\tangx\Desktop\six'
#文件夹名称
dirs = os.listdir(rawpath)
# 文件夹绝对路径
for i,dir in enumerate(dirs):
    dirpath = os.path.join(rawpath,dir)
    filenames = os.listdir(dirpath)
    imgpath = os.path.join(dirpath,filenames[0])
    strippath = os.path.join(dirpath,filenames[2])
    if i<10:
        strname = f'00{i}{filenames[0]}'
    elif i<100:
        strname = f'0{i}{filenames[0]}'
    else:
        strname = f'{i}{filenames[0]}'
    savepath = os.path.join(sixpath,strname)
    img2strip(imgpath,strippath,savepath)

发现一个问题 用get_fdata会导致数据类型从int16变成float
导致图像intensity变化
用get data虽然会警告 但是 存储的intensity不会变化
同时还有一个错误就是我的strip用错了 用成了segmentation

更新的代码

def img2strip(imgpath,strippath,savepath):
    img = nib.load(imgpath)
    imgdata = img.get_data()
    print(imgdata.dtype)
    strip = nib.load(strippath)
    stripdata = strip.get_data()
    print(stripdata.dtype)

    newimgdata = imgdata * stripdata
    print(newimgdata.dtype)
    newimg = nib.Nifti1Image(newimgdata,img.affine)
    nib.save(newimg,savepath)