Watson Becomes a Guide for Cancer Therapies

May 6, 2015

I read “IBM’s Watson to Guide Cancer Therapies at 14 Centers.” My immediate reaction? Baloney. I am no doc but I recall an experience with docs, hospitals, and assorted health experts.

In the late 1980s, Ziff Communications created the Health Reference Center. The idea seemed like a good one. We packed information about frequent disorders on CD ROMs. Remember those? We put a—for the time—a user friendly interface on the system. We included categories, what today’s marketers call metadata tags, so a person could select a category and see accurate, consumer-centric articles about common maladies. No rocket science for the folks at Ziff’s Information Access unit. But it was a first. We worked out trial placements at some big outfits. We set up the kiosks.

What happened?

The phones began to ring almost immediately. No one wanted the gizmos in their facilities. Docs were among the first and loudest complainers. Docs did not want a kiosk providing information about diabetes or prostrate cancer to their patients. We removed the kiosks and went back to the drawing board. We reengineered the kiosk into what would be called a “cloud service” today. Out of sight, calm returned.

Good lesson for me.

Flash forward to 2015. Have docs morphed from technological knuckle draggers to Silicon Valley surfers? Maybe. The assertion in the headline does not ring true in my ears. IBM’s marketers are doing everything possible to generate revenues from a collection of content processing software. IBM, as you may know, has delivered three years of declining revenue. Since the IBM technology is not delivering substantial, sustainable growth, enter the Watson cancer guide.


James Fennimore Cooper crafted a fictional guide. Is Watson a digital guide like Deerslayer’s sidekick?

According to the write up:

Oncologists will upload the DNA fingerprint of a patient’s tumor, which indicates which genes are mutated and possibly driving the malignancy. Watson, recognized broadly for beating two champions of the game show Jeopardy! in 2011, will sift through thousands of mutations and try to identify which is driving the tumor, and therefore what a drug must target. Distinguishing driver [sic] mutations from others is a huge challenge. IBM spent more than a year developing a scoring system so Watson can do that, since targeting non-driver mutations would not help.

This sounds great. I like the Big Data, the uploading, the DNA touch. The problem is that docs, like lawyers, often cling to methods that look back to the good old days and their time worn methods. Docs use what their employers identify as procedurally appropriate.

Now, in my chats  with docs, I ask about their knowledge of various technologies to which I am exposed and I write; for example, automated processing and report generation. Now keep in mind that I live in Kentucky which is in the lower quartile of education and in the upper quartile of bourbon production and horse racing.

The level of tech knowledge is pretty low. Even more amusing to me is that curiosity about advanced technologies is generally low. Continuing education is more important than digging into bits and bytes in my narrow angle of view.

I have no doubt that the institutions involved are delighted to get access to Watson. I have no doubt that some researchers will explore the system.

I don’t accept the assertion that Watson will guide anything. Don’t agree. If you get cancer, chase down a Watson centric health care provider and let me know how that works out.

I would prefer a specialist with experience and a track record of success. Watson can be a resource but I don’t think Watson makes open source, acquired, and home brew code into Chingachgook. What works in fiction may not transfer to real life except in the mental synapses of marketers.

Stephen E Arnold, May 6, 2015


Comments are closed.

  • Archives

  • Recent Posts

  • Meta