Unlocking the PACS DICOM Domain for its Use in Clinical Research Data Warehouses

Abstract

Clinical Data Warehouses (DWHs) are used to provide researchers with simplified access to pseudonymized and homogenized clinical routine data from multiple primary systems. Experience with the integration of imaging and metadata from picture archiving and communication systems (PACS), however, is rare. Our goal was therefore to analyze the viability of integrating a production PACS with a research DWH to enable DWH queries combining clinical and medical imaging metadata and to enable the DWH to display and download images ad hoc. We developed an application interface that enables to query the production PACS of a large hospital from a clinical research DWH containing pseudonymized data. We evaluated the performance of bulk extracting metadata from the PACS to the DWH and the performance of retrieving images ad hoc from the PACS for display and download within the DWH. We integrated the system into the query interface of our DWH and used it successfully in four use cases. The bulk extraction of imaging metadata required a median (quartiles) time of 0.09 (0.03-2.25) to 12.52 (4.11-37.30) seconds for a median (quartiles) number of 10 (3-29) to 103 (8-693) images per patient, depending on the extraction approach. The ad hoc image retrieval from the PACS required a median (quartiles) of 2.57 (2.57-2.79) seconds per image for the download, but 5.55 (4.91-6.06) seconds to display the first and 40. 77 (38.60-41.63) seconds to display all images using the pure web-based viewer. A full integration of a production PACS with a research DWH is viable and enables various use cases in research. While the extraction of basic metadata from all images can be done with reasonable effort, the extraction of all metadata seems to be more appropriate for subgroups.

How to Optimize Connection Between PACS and Clinical Data Warehouse: A Web Service Approach Based on Full Metadata Integration

“…The integration of fine-grained data allowed the advanced query of imaging data with all the other data gathered in eHOP and leverages coded Dicom attributes (using LOINC or SNOMED). This metadata extraction method goes beyond in terms of data indexing than the CFIND based approach implemented in the study by Kaspar et al [8]. We plan to keep the plugin approach to implement a set of services such as classification in the RadLex ontology [10].…”

“…This plugin could map attributes with the same meaning on a unique new DCMEHOP code. Like in the study by Kaspar et al [8], our PACS does not implement the WADO and QIDO protocols [11], which seems still underused by vendors [12]. Theoretically, WADO might improve the ETL process performance because it does not require images to extract metadata.…”

“…Other solutions exist to carry out imaging studies, but they are fully imaging oriented [5,6], or extract data from the PACS for a given predefined cohort and allow crossing imaging data with clinical data a posteriori [7]. Kaspar et al [8] described the implementation of a technical component to integrate imaging metadata from the clinical PACS into a CDW. They proved the feasibility of routine feeding via basic metadata queries (CFIND), or queries on the first image of a series (CMOVE) to retrieve all metadata for the identified patient subsets.…”

Querying a Clinical Data Warehouse for Combinations of Clinical and Imaging Data

“…We added metadata from DICOM files to the CDW. This is relatively simple, but can support many used cases [14]. A major goal remaining is to also integrate variables of features extracted from the imaging raw data, i.e., the pixel data.…”

“…We previously showed the overall feasibility of a comprehensive integration architecture of a production PACS (including identified data) to a research CDW (including pseudonymized data) using ad hoc pseudonymization [14]. The most comprehensive PACS-CDW integration of the related work so far has been shown for i2b2, which primarily requires selecting a patient population in the CDW using clinical data that then becomes the basis for a PACS query in another downstream module [15].…”