Collections.rst

Collections

Introduction

In addition to the Series level, there are collection classes Study, Patient and Cohort for the case when multiple Series will be handled.

Each of the collection classes will take a source, and sort the images into appropriate instances. In each case, the input order of each Series will be auto-detected. There is no way to set input order explicitly.

Note: At present, this works for DICOM data only.

Reading a Study of Multiple Series

The Study class can be used to sort DICOM files according to SeriesInstanceUID. The input order of each Series is auto-detected.

from imagedata import Study

vibe, dce = None
study = Study('data/dicom')
for uid in study:
    series = study[uid]
    if series.seriesDescription == 'vibe':
        vibe = series
    ...
If not (vibe and dce):
    raise ValueError('Some series not found in study.')

Study Attributes

Study property name	DICOM Attribute Name	Usage
studyDate	StudyDate	datetime.datetime
studyTime	StudyTime	datetime.datetime
studyDescription	StudyDescription	str
studyID	StudyID	str
studyInstanceUID	StudyInstanceUID	str
referringPhysiciansName	ReferringPhysiciansName	str
generalEquipment	Instance of GeneralEquipment class	GeneralEquipment

GeneralEquipment Attributes

Study property name	DICOM Attribute Name	Usage
manufacturer	Manufacturer	str
manufacturersModelName	ManufacturerModelName	str
stationName	StationName	str
deviceSerialNumber	DeviceSerialNumber	str
softwareVersions	SoftwareVersions	str

Reading a Patient with Multiple Study Instances

from imagedata import Patient

patient = Patient('data/dicom')
for uid in patient:
    study = patient[uid]
    print(study.studyDate, study.studyTime)

Patient Attributes

Patient property name	DICOM Attribute Name	Usage
patientName	PatientName	str
patientID	PatientID	str
patientBirthDate	PatientBirthDate	str
patientSex	PatientSex	str
patientAge	PatientAge	str
patientSize	PatientSize	float
patientWeight	PatientWeight	float
qualityControlSubject	QualityControlSubject	str
clinicalTrialSubject	Instance of ClinicalTrialSubject class	ClinicalTrialSubject
patientIdentityRemoved	PatientIdentityRemoved	str
deidentificationMethod	DeidentificationMethod	str

ClinicalTrialSubject Attributes

Patient property name	DICOM Attribute Name	Usage
sponsorName	ClinicalTrialSponsorName
protocolID	ClinicalTrialProtocolID
protocolName	ClinicalTrialProtocolName
siteID	ClinicalTrialSiteID
siteName	ClinicalTrialSiteName
subjectID	ClinicalTrialSubjectID
subjectReadingID	ClinicalTrialSubjectReadingID
protocolEthicsCommitteeName	ClinicalTrialProtocolEthicsCommitteeName
protocolEthicsCommitteeApprovalNumber	ClinicalTrialProtocolEthicsCommitteeApprovalNumber

Reading a Cohort of Multiple Patient Instances

from imagedata import Cohort

cohort = Cohort('data/dicom')
for id in cohort:
    patient = cohort[id]
    print(patient.patientName, patient.patientID)

Cohort Attributes

At present no Cohort attributes are implemented.

Cohort property name	DICOM Attribute Name	Usage
N/A	N/A	Not used