在已知DICOM和三維模型對(duì)應(yīng)掩膜的情況下，計(jì)算三維模型的體積。

思路：

1、計(jì)算每個(gè)體素的體積。每個(gè)體素為長(zhǎng)方體，x，y為PixelSpacing，z為層間距

使用pydicom.read_file讀取DICOM文件，dcm_tag.PixelSpacing獲取像素間距，dcm_tag.SliceLocation 獲取層間距

2、計(jì)算體素的個(gè)數(shù)

代碼如下：

from PIL import Imageimport numpy as npimport pydicomimport os def get_pixels_No(bmp_data_dir):  pixels_No = 0  bmp_files = os.listdir(bmp_data_dir)  for bmp in bmp_files:    bmp_file = os.path.join(bmp_data_dir,bmp)    img = Image.open(bmp_file)    img_array = np.array(img)    # img_array.dtype為布爾類型，需要轉(zhuǎn)換為Int類型，其累加和恰好為體素總和    img_array_int = img_array.astype(int)    pixels_No = pixels_No+img_array_int.sum()  return pixels_No def get_pixel_info(dcm_data_dir):   pixel_infos = []  dcm_files = os.listdir(dcm_data_dir)   dcm_file_1 = os.path.join(dcm_data_dir,dcm_files[0])  dcm_tag_1 = pydicom.read_file(dcm_file_1)  # 獲取像素間距.  spacex, spacey = dcm_tag_1.PixelSpacing  # 獲取層間距  # 有些 dcm圖像并不是按照InstanceNumber進(jìn)行排序的，不能直接用最后一張的slicelocation減去第一張，再除以張數(shù)  SliceLocations = []  ImagePositon_z = []  for dcm in dcm_files:    dcm_file = os.path.join(dcm_data_dir, dcm)    dcm_tag = pydicom.read_file(dcm_file)    SliceLocations.append(dcm_tag.SliceLocation)    ImagePositon_z.append(dcm_tag.ImagePositionPatient[2])  SliceLocations_max =max(SliceLocations)  SliceLocations_min =min(SliceLocations)  ImagePositon_z_max = max(ImagePositon_z)  ImagePositon_z_min = min(ImagePositon_z)  print(SliceLocations_max)  print(SliceLocations_min)  print(ImagePositon_z_max)  print(ImagePositon_z_min)  if SliceLocations_max - SliceLocations_min < 1e-10:    spacez = abs(ImagePositon_z_max - ImagePositon_z_min)/(len(dcm_files)-1)  else:    spacez = abs(SliceLocations_max - SliceLocations_min)/(len(dcm_files)-1)  pixel_infos = [spacex, spacey, spacez]   return pixel_infos def get_volume(dcm_data_dir,bmp_data_dir):  pixel_infos = get_pixel_info(dcm_data_dir)  pixels_No = get_pixels_No(bmp_data_dir)  volume=pixel_infos[0]*pixel_infos[1]*pixel_infos[2]*pixels_No/1000  return volume # dcm = pydicom.read_file(r"E:/20181210090945_LENG HONGYING F-44Y/Venous/0000.dcm")# print(dcm)# print(dcm.ImagePositionPatient[2])# print(dcm[0x0020, 0x0032].keyword,dcm[0x0020, 0x0032].value) volume=get_volume(r"E:/20181210090945_LENG HONGYING F-44Y/Venous",r"E:/20181210090945_LENG HONGYING F-44Y/Results/LungL")print("體積為%.1f"%volume)

以上就是本文的全部?jī)?nèi)容，希望對(duì)大家的學(xué)習(xí)有所幫助，也希望大家多多支持武林站長(zhǎng)站。