Arik Hesseldahl

Recent Posts by Arik Hesseldahl

A Cloud Application That Saves Lives

Say what you will about about the constant appearance of cloud computing as the buzz phrase of the moment in tech circles. For all its vaunted simplification of corporate computing environments that lead to billions in cost savings, here’s a metric that trumps them all: Saving five lives.

I learned today about a complicated five-way kidney transplant operation that took place on Friday at San Francisco’s California Pacific Medical Center involving five surgeons, four anesthesiologists, 10 nurses in the operating room, and no fewer than 40 support staff. So large a kidney swap operation is so rare that this constitutes only the second time its been done in the United States, and the first was nearly five years ago.

When you need a new kidney, finding a donor can be a complicated process. Once upon a time you needed only a willing donor of the same blood type, or to wait for one to become available from a compatible deceased person. Now the science of matching donors to recipients focuses also on matching six specific antigens. This makes the mathematical complexity of engineering swaps substantially more complicated. When a spouse or sibling needs a kidney and a compatible match isn’t readily available, spouses or family members who would otherwise donate to their loved ones in need offer instead to swap with another recipient/donor pair in a similar situation. (The image above illustrates the scenario.)

When you consider that there are more than 80,000 people in the U.S. in need of a new kidney, the complexity of engineering these swaps given all the factors involved becomes incredibly difficult. I’m no mathematician, and I don’t even know what it means, but I’m told that when a sufficiently large number of people are involved, the problem reaches a level of complexity known as “NP-hard.”

It’s the sort of problem that the software engineering minds of Silicon Valley would in theory be adept at solving. And David Jacobs would qualify as one of them. After 25 years working for companies like Microsoft, Marimba (now a unit of BMC) and Macromedia (now part of Adobe), he came down with a kidney disease that required a transplant. He spent more than three years waiting and endured more than a year of dialysis before receiving a donated kidney from a friend in 2004. After that he started exploring ways to put his software brains to work on this kidney swap problem.

The result is Silverstone Matchgrid, a cloud-based application that takes all the characteristics involved in matching willing kidney donors with potential recipients in need–all the factors including blood type as well as all those antigens. Estimates are that some 6,000 people would qualify for a Kidney Paired Donation if they knew about the option, and an efficient national exchange system could result in as many as 3,000 transplants from living donors annually.

Hospitals and transplant organizations can use the application to speed up the search to match patients in need of kidney transplants who have a qualified, incompatible donor with an alternate compatible donor. The software not only generates all the potential paired donations possible, but determines the maximum number of transplants that can happen from a pool of candidates. Not bad for an application that runs on Amazon Web Services, eh?

It’s a fascinating story that’s better demonstrated than described. Jacobs demonstrated Matchgrid at the DEMO 2009 conference. A video of that presentation is below.