This GitHub gist is for converting DICOM metadata into CSV file. You can see the DICOM dataset on Kaggle following the link: https://www.kaggle.com/datasets/dmisky/dlwptvolumetricdicomlung

dicom2csv.py

# Implement the required package
# If you don't have PyDicom, uncomment the next line of code
# !pip install pydicom
import pydicom
import pandas as pd
import glob

def dicom2csv(extract = [],
              move_on = [(0x7FE0,0x0008), (0x7FE0,0x0009), (0x7FE0,0x0010)],
              folder_path = str(),
              csv_file_name = "metadata.csv",
              return_dataframe = False):
  
  '''
  Extract specific DICOM metadata from multiple DICOM files collected in a
  specified folder.

  ::Params
      - extract (list):
          The keywords of the DICOM attributes you want to extract.

      - move_on (list):
          The tags of the DICOM attributes you don't want to move on.
          By default,it contains a list of the tags of pixel data:
          [(0x7FE0,0x0008), (0x7FE0,0x0009), (0x7FE0,0x0010)]
          If you want to move on onto some attributes, it's recommended
          to append its unique tags in addition to pixel data tags.

      - folder_path (string):
          Path of the folder that contains the DICOM files.

      - csv_file_name (string):
          The name of the CSV file.

      - return_dataframe (bool):
          if True, returns a pandas dataframe of the extracted data for
          direct use.
  
  :: Returns
      - CSV file contains DICOM metadata specified the parameter extracted.
      - Pandas dataframe when return_dataframe is set to True.
  
  :: Example:
      dicom2csv(extract = ['StudyDate'],
                folder_path = 'content/dicomfolder',
                csv_file_name = "Study_Dates.csv",
                return_dataframe = True)
  '''
  
  # Initialize the meta dictionary that will have the specified attributes
  meta = {keyword:[] for keyword in extract}

  # List the files' names that we want to extract data from
  dicom_files = glob.glob(folder_path+'/*.dcm')

  # Iterate over each DICOM file in the folder and read it using dcmread()
  for file_path in dicom_files:
    # Read the DICOM file from the specified path
    dcm = pydicom.dcmread(file_path)

    # Iterate over the DICOM attributes in the current DICOM file "dcm"
    for elem in dcm.iterall():
      # Ensure that the attribute is not a pixel data and it's one of the
      # required attributes
      if (elem.tag not in move_on) and (elem.keyword in extract):
        # Append the value of the current attribute
        meta[elem.keyword].append(elem.value)
  
  # Create a pandas dataframe for better use of the data
  df = pd.DataFrame(data=meta, columns = extract)

  # Create the CSV file with the specified name
  df.to_csv(csv_file_name, index=False)

  # Return the extracted dataframe for direct use
  if return_dataframe:
    return df