Cancer diagnosis meets Angry Birds

Cancer diagnosis is a far cry from Angry Birds, but I found out recently that they have a few things in common. Both the online shooting gallery and a new pathology imaging tool out of UAB are web apps that analyze digital images: one to entertain; the other to see how fast cancer cells are growing.

This snapshot is the result of using an ImageJS module to determine how fast cancer cells are growing. The dark blots are the nuclei of cells dividing as part of the high-speed, abnormal growth seen in tumors.

Bioinformatics experts here this week published the code for a free Web application, called ImageJS. It lets pathologists analyze digitized pathology slides for malignancy based on how cancer cells change color when exposed to standard dyes. More exciting in some ways are future ImageJS modules that promise to deliver genomics analysis and to let doctors compare their patient’s data to similar cases stored in national databases in real time while keeping it private.

Why should we care?  Such future comparisons will make your diagnoses more accurate and rule out treatments that won’t work with your genetic signature, sparing you misery.

Most importantly, the system should realize the promise of Big Data, Web 3.0, cloud computing, the one world computer and the semantic Web. What?  

Many laymen, including me, and some researchers, have little idea about what those things are or what they will mean for healthcare and research in the coming years. Our bioinformatics guy, Dr. Jonas Almeida, tried to help me prepare to write my article on his new system by suggesting I read The Big Switch by Nicolas Carr.

The book adroitly explained the above tech jargon, but my understanding remains basic and my attempts to clearly summarize Dr. Almeida’s work imperfect. Maybe we should have broken the story into a series to make its complexity more digestible, or done separate versions for programmers, cancer docs and patients. Still, we did our best and look forward to your thoughts. 

By the way, cloud computing is much in the news, but coverage is more likely to focus on gaming than medicine. We had hoped that the tech pundits might use our publication as excuse to cover open coding or the cloud in medicine, but the ImageJS system is only a research prototype.

Neither should we argue too forcibly for the immediate importance of this work because bioinformatics as a whole is still in its relative infancy, with plenty of potential for overstatement. Dr. Almeida is the first to argue that ImageJS will only become significant if it becomes popular with pathologists and biostatisticians nationwide.  

As consumers, we can choose from free Web applications by the hundreds, most of them games. Here’s hoping our doctors will soon have access to such cheap and powerful technologies as they care for us.  At least Dr. Almeida has made a start. Just like Angry Birds or Instagram, the ImageJs modules are available at the Google Chrome App store.


PS: Under the broader heading of medical research and social computing, check out this article on 23andMe  in the Nature blog Spoonful of Medicine.