EyeIC

Developing a Custom Web-based Application for Medical Use.

EyeIC, Inc. is a cutting-edge healthcare technology firm dedicated to developing innovative software solutions for problems in medical image analysis. The company's first product, MatchedFlicker, arose from research at the University of Pennsylvania's School of Medicine by Dr. Jeffrey Berger and his colleagues. MatchedFlicker is a groundbreaking, 510(k) FDA-cleared application that automatically displays changes between virtually any set of serial (time series) still images. In medicine, MatchedFlicker is being initially developed to aid clinicians with early detection and monitoring of retinal disease and glaucoma.

Published data has shown clear benefits of early detection in these diseases for determining appropriate treatment and improving outcomes. Furthermore, the display of subtle change provided by MatchedFlicker can be shown to patients to help them better understand their disease with the goal of improving compliance and disease management in chronic indications such as diabetes. EyeIC is also investigating the potential for MatchedFlicker in broader applications throughout medicine.

EyeIC initially worked with Mercury Development to develop a Windows desktop-based solution marketed exclusively as licensed software to ophthalmologists and optometrists. The software proved successful in aiding clinicians to better detect and monitor glaucoma and retinal diseases, but the ‘shrink-wrapped’ desktop implementation posed significant challenges to users that limited adoption.

Some difficulties encountered by users included:

• Cost: the licensing model required a considerable upfront investment by clinicians in addition to existing capital expenses for acquiring imaging devices.
• Accessibility: in many cases, MatchedFlicker was not installed on the same machine where images were captured, thus requiring manual transfer of files to enable analysis and limiting the ability for clinicians to collaborate. Also, in many cases users wanted access to images at multiple locations, which was not possible with the desktop implementation.
• Delivery: the software was initially designed with an elaborate update mechanism that was frustrating for users and difficult to manage and required high development overhead to test for all possible combinations of operating system and software.
• Workflow: although the software had excellent tools to facilitate the analysis and reporting required for reimbursement, several manual steps were needed to incorporate the findings from MatchedFlicker into a form suitable for submission to payers.

In mid-2010, Phil Heifetz joined EyeIC as President and CEO and immediately set about evaluating the existing business model and software development strategy. He observed that overall industry trends foretold a rapid conversion to digitized medicine. A key component of this transformation was an increasing demand from patients and medical professionals to have access to medical information and tools via the internet. It was determined that in a few short years, to be competitive, vendors would need to offer solutions that enabled easy access to data across multiple platforms and environments while maintaining high security and compliance with privacy laws.

The team concluded that even in the short term, a pure desktop solution would unlikely be able to gain the kind of adoption needed by EyeIC. After careful analysis and based on prior success and recommendations, EyeIC approached Mercury Development to discuss the best way to move their application to the internet.

In March 2011 Mercury Development agreed to work with EyeIC to create a beta version of a web-based MatchedFlicker. From the beginning, the expert Mercury Development team was involved in all phases of the project, providing valuable functional and UI analysis and helping reduce cost. For example, Mercury Development found ways to reuse key pieces of software code from the original desktop version of MatchedFlicker. At the same time, an improved architecture ensured a higher level of functionality and operability. The new version of MatchedFlicker is a cross-platform, multi-user product that runs on all major browsers, including Chrome, Safari, Internet Explorer, and Opera. Other features included:

• Centralized, secure, scalable storage for patient images
• Host-level rapid and reliable implementation of image matching algorithm
• Backward compatibility with files from the older version of MatchedFlicker.

In addition to meeting these technology goals, the new version of MatchedFlicker provides users with an automated reporting system for reimbursement, a redesigned user interface, and secure, HIPAA-compliant online access and compatibility with virtually any still image capture device.

Finally, to assure success during beta testing, Mercury Development designed and carried out extensive load tests to ensure performance would meet expectations, and to estimate when to schedule cloud capacity expansion with the anticipated increase in users as adoption grows.

Completion of the successful migration of MatchedFlicker from a desktop-based system to a web-based model in just 3 months proved to EyeIC that Mercury Development was a highly reliable partner dedicated to efficiently building exceptional software