In pamelarussell/TCIApathfinder: Client for the Cancer Imaging Archive REST API

api_key <- Sys.getenv("TCIA_API_KEY")
if (identical(api_key, "")) {
  knitr::opts_chunk$set(eval = FALSE)
  message("Note: code examples will not be evaluated because the system does not have a valid API key installed.")
}

Installation

From within R:

install.packages("TCIApathfinder")

From GitHub:

# install.packages("devtools")
devtools::install_github("pamelarussell/TCIApathfinder")

Authentication

An API key is required to access data from TCIA. To obtain and correctly store your API key:

1. Request a key from TCIA by following the instructions here.

2. Create a text file in your home directory (~/) called .Renviron.

3. Create the contents of the .Renviron file like this, making sure the last line in the file is empty. Otherwise, R will silently fail to load the file.

   ``` TCIA_API_KEY=xxx-xxx-xxx-xxx

   ```

4. Restart R. .Renviron is only processed at the beginning of an R session.

Usage

Load the package:

library(TCIApathfinder)

Get the names of all TCIA collections:

collections <- get_collection_names()
head(collections$collection_names)

Get the names of all imaging modalities

modalities <- get_modality_names()
head(modalities$modalities)

Note: a collection or body part can be specified to narrow down results.

Get the names of all body parts studied:

body_parts <- get_body_part_names()
head(body_parts$body_parts)

Note: a collection or modality can be specified to narrow down results.

Get information for all patients in a collection

patients_tcga_brca <- get_patient_info(collection = "TCGA-BRCA")
head(patients_tcga_brca$patients)

Note: if no collection is passed, patients for all collections are returned.

Get all image series based on criteria

series <- get_series_info(patient_id = "TCGA-AR-A1AQ")
head(series$series)

Note: other ways to narrow down results include

• collection
• study instance UID
• series instance UID
• modality
• body part
• manufacturer
• manufacturer model name

Get detailed information on all imaging studies for a patient

studies <- get_patient_studies(patient_id = "TCGA-AR-A1AQ")
head(studies$patient_studies)

The variables in studies$patient_studies correspond to the fields of a PatientStudy object as described in the API documentation.

Note: other ways to narrow down results include a collection or a study instance UID.

Get all imaging studies for a collection

studies_tcga_brca <- get_studies_in_collection(collection = "TCGA-BRCA")
head(studies_tcga_brca$studies)

Note: a patient ID can be provided to further narrow down results.

Get individual DICOM image IDs for an image series

sop_uids <- get_sop_instance_uids(
  series_instance_uid = "1.3.6.1.4.1.14519.5.2.1.3344.4002.298037359751562809791703106256")
head(sop_uids$sop_instance_uids)

Download a single DICOM image

im <- save_single_image(series_instance_uid = "1.3.6.1.4.1.14519.5.2.1.3344.4002.298037359751562809791703106256",
                  sop_instance_uid = "1.3.6.1.4.1.14519.5.2.1.3344.4002.113224119964450170072494597907")
im$out_file

Note: a file name can be provided to override the original file name.

Download an image series and extract it

ser <- save_image_series(series_instance_uid = "1.3.6.1.4.1.14519.5.2.1.3344.4002.298037359751562809791703106256",
                         out_file_name = "series.zip")
zip <- ser$out_file
extracted <- extract_image_series(zip)
extracted$dirs

Download, save and extract an image series, optionally to a temporary location

ser <- save_extracted_image_series(series_instance_uid = "1.3.6.1.4.1.14519.5.2.1.5382.4002.806935685832642465081499816867")
ser$dirs

Additional functions

See package documentation for further details:

• get_series_size
• get_manufacturer_names
• get_new_patients_in_collection
• get_new_studies_in_collection
• get_patients_by_modality

pamelarussell/TCIApathfinder documentation built on Feb. 11, 2021, 7:02 a.m.