Working with XNAT using PyXNAT

Download PyXNAT Module

!pip install pyxnat

Connect to server: Enter user : password

import pyxnat
import os

# connect to XNAT instance
from pyxnat import Interface
xnat = Interface(server='http://sharp.bidmc.harvard.edu:8080',  cachedir='/tmp')
xnat.select.projects().get()

Different types of datatypes supported by XNAT

xnat.inspect.datatypes()

Check number of Subject

subjects = xnat.select('/projects/FAST/subjects')
subjects.get().__len__()

Loading the project

project = xnat.select.project('FAST')
print(project)

Working with Subject Data

contraints = [('xnat:subjectData/SUBJECT_ID','LIKE','%'),
                  'AND', ('xnat:subjectData/PROJECT', '=', 'FAST')
              ]
table = xnat.select('xnat:subjectData', ['xnat:subjectData/SUBJECT_LABEL','xnat:subjectData/PROJECT','xnat:subjectData/SUBJECT_ID']).where(contraints)
table.__len__()

print(table)

Working with MRI Session Data

contraints = [('xnat:mrSessionData/ID','LIKE','%'),
                  'AND', ('xnat:subjectData/PROJECT', '=', 'FAST')
             ]
table1 = xnat.select('xnat:mrSessionData', ['xnat:mrSessionData/SUBJECT_LABEL','xnat:mrSessionData/SESSION_ID']).where(contraints)
table1.__len__()

print(table1)

Filtering using Behavioral scores

contraints = [('xnat:subjectData/SUBJECT_ID','LIKE','%'),
                  ('behavioral:scores/VocabScore', '>=', '36'),
                  'AND', ('xnat:subjectData/PROJECT', '=', 'FAST')
             ]
table1 = xnat.select('xnat:subjectData').where(contraints)
table1.__len__()

Downloading the selective data

Lets start with download data for one subject

subject = xnat.select.project('FAST').subject('0001')
experiment = subject.experiment("SHARP_E00746")
allscans = experiment.scans()
allscans.download("/tmp", type='ALL', extract=False)

Now lets write a filer to download the selective data for all the subjects

# Filer can be developed based on the data parameters
contraints = [('xnat:mrSessionData/ID','LIKE','%'),
                  'AND', ('xnat:subjectData/PROJECT', '=', 'FAST')

list_subjects = xnat.select.project('FAST').subjects().where(contraints)
for list_subject in list_subjects:
    list_experiments = list_subject.experiments().where(contraints)
    for list_experiment in list_experiments:
        print list_experiment
        scans = list_experiment.scans()
        try:
            # Number 2 is for Anatomical data. Similar types can be set for other data types
            scans.download("/tmp", type='2', extract=False)
        except:
            print "There are no scans to download"