Encrypted login | home

Program Information

OpenVCT: A Framework for Open-Source Virtual Clinical Trial Resources


A Maidment

A Maidment1*, P Bakic1 , B Barufaldi1 , D Higginbotham1 , D Pokrajac2 , A Avanaki3 , K Espig3 , A Xthona3 , T Kimpe4 , (1) University of Pennsylvania, Philadelphia, PA, (2) Delaware State University, Dover, DE, (3) Barco, Beaverton, OR, (3) Barco, Kortrijk, Belgium

Presentations

WE-G-601-6 (Wednesday, August 2, 2017) 4:30 PM - 6:00 PM Room: 601


Purpose: Virtual clinical trials (VCTs) are computer simulations of clinical trials; in this presentation, we focus on trials of imaging devices. VCTs have multiple purposes. They aid in the design and development of novel imaging systems, prototype clinical trials, augment clinical trials with simulated data, and aid in regulatory approval. OpenVCT is a precompetitive effort to standardize VCTs.

Methods: We have developed a computational framework for the formulation and conduct of VCTs. VCTs consist of two phases – virtual subject image accrual, and virtual reader studies. The OpenVCT framework (OpenVCT.org) provides standards for: (1) encoding and communicating real and simulated anatomic models; (2) specifying and executing the sequence of steps in the VCT; (3) encoding the output of the VCT.

Results: The standard data format uses a multipart compressed directory and file structure containing prescribed XML and binary documents. XML files are used to specify the VCT steps, and record the VCT output. The binary files encode the simulated anatomy, lesions, lesion masks, etc. While multiple image file formats are supported, we recommend that files be written and stored in DICOM format. Trial results can be stored in XML format or in statistical software formats for data analysis. All software is open-source. Libraries for common actions (file read/write, XML operations, etc.) are provided. Sample code, sample output, lexical and other necessary tools are also provided. At the current time, we can simulate a tomosynthesis exam of a simulated breast in under 60s using off the shelf GPU-enabled computers, offering the potential to simulate clinical trials of hundreds of thousands of women.

Conclusion: It this presentation, we will describe the key aspects of the OpenVCT standard, discuss the open-source software resources for conducting VCT, and demonstrate the use of VCT in the realm of digital mammography and digital breast tomosynthesis.

Funding Support, Disclosures, and Conflict of Interest: This work is supported in part by the Burroughs Wellcome Fund, the Komen Foundation, NIH/NCI, Barco NV, and Real-Time Tomography (RTT) LLC. ADM is a scientific advisor to RTT. ADM is the spouse of an employee and shareholder of RTT.


Contact Email: