Ashish Sharma, Presenter: Nothing to Disclose

Nadir Saghar, Abstract Co-Author: Nothing to Disclose

Darryl Tharpe, Abstract Co-Author: Nothing to Disclose

Tony Pan MS, Abstract Co-Author: Nothing to Disclose

Ross Warren Filice MD, Abstract Co-Author: Nothing to Disclose

Easy reproducible access to imaging and annotation data to support clinical trial results is an important part of clinical trial review and auditing. It is typically challenging to access this data, since it is often stored in proprietary formats and not easily accessible. Measurements are reported on CRFs while images reside on CDs. The latter often do not show the actual measurement or locate the structures that were actually measured. Consequently measurements on images are challenging to verify. Here we describe an informatics-based approach to enable the review of image and annotation data captured during a trial. It utilizes Service-oriented architecture (SOA) to federate databases that manage DICOM images and annotations to facilitate ‘single-click’ access to reproducibly recreate annotated-images.

This project uses two databases that manage images and annotations. DICOM images can be stored in NBIA (an open source research image data management system). Annotations are described using the AIM model and stored as DICOM SR or as XML objects. All databases are exposed via a secure REST API. The REST API for NBIA is similar to WADO-RS. The API for AIM annotations includes the ability to convert the annotations into HTML5 compliant format that can be rendered in web applications. The APIs, also include the ability to federate images and annotations. Finally, a web application was developed that allow a reviewer to review images and annotations.

This pilot replicates the workflow of image review in clinical trials. The use of SOA allows us to easily construct systems that can securely federate images and annotations without overhauling existing data management systems. REST APIs make it easy to create presentation platforms that can advance the adoption of quantitative imaging.

Standardizing image and annotation data facilitates storage of this data in accessible sites along with development of presentation platforms that give 'single-click' access to any image and markup directly from a review document. Such standardization also allows important research applications across clinical trial data sets such as quantitative imaging biomarker development.

http://abstract.rsna.org/uploads/2014/14018789/14018789_7d4v.jpg

Sharma, A, Saghar, N, Tharpe, D, Pan, T, Filice, R, Federated Access to Images, Annotations and Measurements to Optimize the Review and Auditing of Imaging Data Submitted in Clinical Trials.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14018789.html