MDDX
Revealing Stories from Research Data

Dan Gebow, PhD., while working on a large global CT imaging registry in collaboration with two cardiologists at Johns Hopkins, came across a revelation that despite their requirement for the anonymized images, they found that de-identification was often only superficial. Patient names, dates of birth, and numerous other identifiers were littered throughout the images. To complicate the situation, the images were from famous personalities including former U.S. presidents, members of Congress, Silicon Valley leaders, and more. In some cases, CDs with the image data were being opened by customs officials and examined. “Besides the obvious potential HIPAA-violation penalties, I was always concerned about the potential political or economic fallout that could result from disclosing heart conditions of the famous personalities,” remarks Gebow. He founded MDDX and led the team to design the gold-standard platform for collecting, screening, storing, and sending pristine and HIPAA-compliant image data. The result was a platform, powered by big data and machine-learning technologies that integrated over 1500 screening and de-identification rules to detect image issues and automatically de-identify any image type. “If you are going to sit on terabytes of image data for decades, you want to make sure it is both de-identified and error free, our platform is the answer.”

The cornerstone of MDDX platform is Moving Pictures. This application uses proprietary algorithms and technology to set a new standard for image data collection. MDDX offers a 100 percent de-identification guarantee, a zero percent error rate for images that passes their image Quality Control, and 20-year online storage to exceed even the most stringent record-retention requirements. The system is cloud-based with data centers covering every corner of the world where clinical trials or post-market surveillance are performed—so that upload and download speeds is never a problem for users.

With our 100 percent de-identification guarantee, MDDX effectively de-risks the trial for the data providers, sponsors, CROs, and labs

HIPAA-HITECH makes the sponsor responsible for auditing their subcontractors for regulatory compliance—means the sponsor has exposure if a lab has data breach and the sponsor’s PHI-laden data is leaked. “With our 100 percent de-identification guarantee, MDDX effectively de-risks the trial for the data providers, sponsors, CROs, and labs,” explains Gebow.

Additionally, MDDX has launched an entire suite of seven applications that make up their “Trial Application Platform”, ranging from the firm’s EDC application to their endpoint adjudication modules, an imaging core lab workflow management system, and eConsent. “If you have a tricky data collection schema and can draw it out on paper, we can auto-generate an online application in under one day—that is automatically Part 11 and HIPAA compliant,” says Gebow.

In an instance, MDDX assisted a client who had special request to have large CTs parsed into three different versions. They needed a full-sized case for their online repository, a medium one with specific views to be sent to the core lab, and a small one to go to their clinical specialists. “We added those custom rules to Moving Pictures, and now it can send out all three versions in under three hours of the image upload. To date, over 3000 CTs have gone through this special process, greatly accelerating the pace of the trial,” explains Gebow.

For the future, “we are rapidly moving beyond cardiology and have just started to gain recognition in oncology, musculoskeletal, and neurological imaging,” marks Gebow. The company also wants to demonstrate how their platform enabled the vision that they had while forming MDDX: to give client’s data life beyond the timeline of the trial. “Your research data has a story to tell—it should be an epic,” concludes Gebow.