In-Depth

.NET in the ER

• By Jack Vaughan
• August 1, 2004

Keith Brophy, CTO at Robertson Research Institute, Saginaw, Mich., has been working on a Web services application whose objective has special relevance -- saving lives.

Two years ago, Brophy began working on NxOpinion, which the firm envisions as a real-time diagnostic tool for ERs throughout the world.

The project was originated by Institute founder and president Dr. Joel Robertson. When the son of a close friend died due to a lack of relevant diagnostic information in an emergency room, Robertson dedicated some of the Institute’s resources to solving the misdiagnosis of non-chronic illnesses.

Brophy and his team chose Microsoft .NET as the application platform. This opened the project up to the use of another type of protocol: Web services protocols that “bake-in” many common functions that developers heretofore coded uniquely for each project. A primary benefit of the platform, according to Brophy, is the ability to deploy on a variety of systems, ranging from desktop PCs to Tablet and notebook computers.

.NET tools, writing to the CLR and .NET Framework class libraries, saved effort and supported reuse, said Brophy. “We have to make very good use of money ... and have predictable outcome of plans. With .NET, we had a 10-month timetable. We were able to come in ahead of timetable and under budget.”

The goal is not to get wrapped up in technology, but to help patients. “The bar that we are aiming for is to save five lives and to be able to measure it,” said Brophy. But the technology is still a vital enabler. NxOpinion uses a Bayesian decision engine for deep diagnostics, and WinForms as a query medium.

Web service component architecture is a key. As described by Brophy, at the heart of the effort is a drive to use parts that interlock, but are flexible enough to allow for retrofitting or throw away five years “down the road.”

So what about these baked-in protocols, we asked Brophy? Is that the real value of Web services?

“Tools count,” he replied, and they can enable Web services. “In describing what has happened to the development paradigm, I would use a construction analogy. We don’t build walls a board at a time any more. The .NET Framework is a huge efficiency tool.”

In the basic “development portion” of this project, Brophy and crew set aside five months for design “envisioning,” followed by another five months of hands-on development. We asked what advice he had for others.

Said Brophy: “Clarify, clarify, clarify.”