Summary

Clinical, genetic, and pathological data resides in the TCGA data portal, while the radiological data is stored on The Cancer Imaging Archive (TCIA). The data utilizes the same TCGA patient identifiers in both repositories, allowing researchers to explore the correlations between tissue genotype and radiological phenotype. This collection within TCIA is related to the PRAD disease type within TCGA.

Note: Per TCGA Guidelines, formal permission requests are still required for publications using TCGA-PRAD data. More info about TCGA Publication Policies can be found on the Data Usage Policies and Restrictions page.

CIP TCGA Radiology Initiative

The CIP is supporting multiple projects within the academic community to encourage cross disciplinary research, which utilizes the data provided in these resources. More can be learned about this effort on the TCGA-PRAD Phenotype Research Group page.

How to Retrieve the Images from TCIA

If you are unsure how to download this collection, please view Searching by Collection or refer to TCIA's User's Guide for more detailed instructions on using the site.

Related Data

The TCGA data portal has extensive clinical and genomic data, which can be matched to the patient identifiers of the images here in TCIA.  Below is a snapshot of clinical data extracted on 11/4/2014.

• TCGA-PRAD Clinical Data.zip

Explanations of the clinical data can be found on the Biospecimen Core Resource Clinical Data Forms linked below:

• Prostate Case Quality Control Form
• Prostate Enrollment Form
• Prostate Follow-Up Form

A Note about TCIA and TCGA Subject Identifiers and Dates

Subject Identifiers: a subject with radiology images stored in TCIA is identified with a Patient ID that is identical to the Patient ID of the same subject with demographic, clinical, pathological, and/or genomic data stored in TCGA. For each TCGA case, the baseline TCGA imaging studies found on TCIA are pre-surgical. 

Dates: TCIA and TCGA handle dates differently, and there are no immediate plans to reconcile:

• TCIA Dates: dates (be they birth dates, imaging study dates, etc.) in the Digital Imaging and Communications in Medicine (DICOM) headers of TCIA radiology images have been offset by a random number of days. The offset is a number of days between 3 and 10 years prior to the real date that is consistent for each TCIA image-submitting site and collection, but that varies among sites and among collections from the same site. Thus, the number of days between a subject’s longitudinal imaging studies are accurately preserved when more than one study has been archived while still meeting HIPAA requirements.
• TCGA Dates: the patient demographic and clinical event dates are all the number of days from the index date, which is the actual date of pathologic diagnosis. So all the dates in the data are relative negative or positive integers, except for the “days_to_pathologic_diagnosis” value, which is 0 – the index date. The years of birth and diagnosis are maintained in the distributed clinical data file. The NCI retains a copy of the data with complete dates, but those data are not made available.With regard to other TCGA dates, if a date comes from a HIPAA “covered entity’s” medical record, it is turned into the relative day count from the index date. Dates like the date TCGA received the specimen or when the TCGA case report form was filled out are not such covered dates, and they will appear as real dates (month, day, and year).

Acknowledgements

We would like to acknowledge the individuals and institutions that have provided data for this collection:

• University of Pittsburgh/UPMC, Pittsburgh, PA - Special thanks to Margarita L Zuley MD, and Rose Jarosz.
• Washington University, Saint Louis, MO - Special thanks to Bettina F. Drake, PhD, Danielle Rancilio, and Aleksandra Klim, Department of Surgery.