迭代 DICOM 文件

Question

Im using pydicom to read dicom files (the patient name), but all my dicom files are name as i34, i67 and so on (no specific order or sequence for these names). My problem: i have a list of patients names in a txt file and i want search every patient name in this txt file with all the dicom files, looking to their PatientName in the dicom file (for those who are not familiar, every dicom file has a number of informations of a patient, one of them being the PatientName)

Exemple:

txt file: John George Ringo Paul

Dicom files: i54 i98 i64 i12

What I want: Take John and search i54, i98, i64, i12 for the PatientName match. And do it for every name in the txt file. Hope I made my self clear. Thks in advance!

EDIT

i try this code but didnt work

import dicom
import os

dir = "/path/to/dir_where_dicom_file_are"
txt = open('/path/to/txt_with_patientsnames.txt','r')
for line in txt:
    for sub, dirs, files in os.walk(dir):
        names = line.strip()
        for dcm in files:
           ds = dicom.read_file(dir+dcm) #necessary for read dicom files                            
           patient = ds.PatientName
           if names in patient:
           #do something

I took a sample of 20 dicom images and a sample of 20 names. I knew that only one dicom file of this sample belong to one patient name so my ouput should be only one match righ? but im getting 20 match results out of this. Whats wrong?

Answer 1

The problem was logic (more specifically the lack of it)

dir = "/path/where/dcm/files/are"
arq = open('path/where/file with pacient names are/patient.txt','r')
for linha in arq:
     c = linha.strip()
     for sub,p,a in os.walk(dir):
        for dcm in a:
          ds = dicom.read_file(dir+dcm)
          pctname = ds.PatientName           
          if c in pctname:
          #do something

To be honest i dont know what was wrong in the first place.

Answer 2

You could also consider approaching this problem from a different angle. You can organize your DICOM files into a directory hierarchy that includes PatientName at one of the levels. This will simplify locating relevant files quickly. One script that can help you with this task is dicomsort .

Using that tool, the command below will sort the DICOM files in the input directory into the PatientName > StudyDate > SeriesDescription-InstanceNumber named directory structure. I often use such sorting as a prerequisite for subsequent processing. Of course, if you are working with a very large dataset, and only need to access small subset of it, this will be a suboptimal solution, but I thought it may be handy for your situation.

dicomsort data sorted/%PatientName/%StudyDate/%SeriesDescription-%InstanceNumber.dcm