The Future of PACS

Last week was dismal. Beginning of the week, I received a newsletter from Aunt Minnie that proclaimed : PACS is going extinct


While that email was unsettling, what was far worse that before lunch that day I received three “forwards” of that same email from colleagues in GE. I was thinking to myself, how should I interpret these “forwards”  – Satire ? Concern for my well-being ? Interrogative ?

The future of PACS (or mine) did not look that bright that morning.

Let’s travel back to 1997. A summer in New York. The reigning chess champion Gary Kasparov was beaten by a computer – the first ever time it happened in a tournament style match ! The computer was called Deep Blue – a IBM supercomputer which was purpose-built to play chess. While some saw this as a triumph of AI over human intelligence, there were some claims that (particularly from Kasparov) that there were humans playing some of the moves behind the scenes. There was a documentary movie made on this, which claims IBM built Deep Blue to boost their stock value.

IBM Deep Blue was decommissioned and torn down and after years of further research IBM built another computer called Watson which rose to fame last year by defeating two champions in Jeopardy – a very famous quiz show in the US. Watson again, was custom-built to play Jeopardy – mounds of algorithm which do natural language processing to learn from terabytes of text. In fact, according to one source, the entire text of Wikipedia was fed into Watson as an input.


IBM did not tear down Watson. In fact, they are working with several organizations to explore how the learning capabilities of Watson can be used in real-world decision intensive problems and we all know when it comes to data and decisions, there is hardly another domain as complex and riveting as healthcare.

Earlier attempts at artificial intelligence required every possible question and answer to be hard-coded into the system, a time-consuming process with little value in healthcare. Watson uses a probabilistic, evidence-based approach. It generates and scores many hypotheses using an extensible collection of natural language processing, machine learning and reasoning algorithms. It gathers and weighs evidence to refine its hypotheses.

Martin Kohn, M.D., Chief Medical Scientist, Care Delivery Systems, IBM Research.

Dr. Eliot Siegel, MD, Professor of Radiology at University of Maryland is one of those pioneers who is working at bringing the capabilities of Watson, into healthcare, specifically into PACS. At this years’ SIIM, he delivered a talk on Next Generation PACS.

Dr. Siegel on Next Generation PACS [intro]


Dr. Siegel talks about an interesting concept, teaching Watson the same way as a medical student studying for a degree program. Starting with basic anatomy and physiology, learning to interpret charts and understand normal values; finally working up to be an aide next to a human performing diagnosis, just like a “Junior Doctor”.

Dr. Siegel on Watson in Medicine [2011]


That email, it turns out, was actually an advt. from a competitor PACS. For now, we all have a job in PACS. So, let’s get back and focus. There is just way too much work to be done.


Further reading/viewing:

  1. Wishlist for next-gen PACS
  2. A long 52 min video on Watson in Healthcare


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s