How to read a DICOM image in Python

The below should give you an idea on how the Pydicom package works. This should be more than enough to extract the pixel data for post-processing.

Hope this helps!

1. import pydicom as dicom 
2.  
3. # Dataset is the main dicom object and is derived from python dictionary data type (i.e. key:value pairs) 
4. # key: is the DICOM (group,element) tag (as a Tag object) 
5. # value: is a DataElement instance. It stores: 
6. # tag - a DICOM tag 
7. # VR – DICOM value representation (see http://dicom.nema.org/dicom/2013/output/chtml/part05/sect_6.2.html) 
8. # VM – value multiplicity 
9. # value – the actual value. 
10. ds = dicom.dcmread("00000001.dcm") 
11.  
12. # Check for the existence of a particular Attributes 
13. print("PatientName" in ds) 
14.  
15. # Access DataElement (recommended methods) 
16. patient_name = ds.data_element("PatientName") 
17. patient_name_tag = patient_name.tag 
18. patient_name_VR = patient_name.VR 
19. patient_name_VM = patient_name.VM 
20. patient_name_value = patient_name.value 
21.  
22. # Access the raw pixel data 
23. pixel_bytes = ds.PixelData 
24.  
25. # Access pixel data in more intelligent way for uncompressed images (recommended) 
26. pixel_array_numpy = ds.pixel_array # returns a NumPy array for uncompressed images 
27. dimensions = ds.pixel_array.shape 
28.  
29. # NumPy can be used to modify the pixel_array, but they must be written back to the PixelData to be saved 
30. # example: set pixels with values < 300 to zero 
31. for n, val in enumerate(ds.pixel_array.flat): 
32. if val < 300: 
33. ds.pixel_array.flat[n] = 0 
34.  
35. # construct Python bytes containing the raw data bytes in the array 
36. ds.PixelData = ds.pixel_array.tostring() 
37. ds.save_as("00000001_new.dcm")