By Mark Watts
In my effort to explain artificial intelligence (AI), I am struck by how hard it is to explain human intelligence.
For us to use AI effectively we must have it perform a task that we value in human intelligence. How is that task selected?
I would suggest using designed thinking to understand the “right” questions or tasks for AI to accomplish.
Design thinking is a human-centered approach to innovation that starts with what is desirable from a human point of view, along with what is technologically feasible and economically viable. There are six phases in the design thinking process:
- Frame a question
- Gather inspiration
- Generate ideas
- Make ideas tangible
- Test to learn
- Share the story
This seems like a rational approach for AI task selection. In health care we may want to track a patient’s journey from start to finish.
- Choose a process or journey that you want to map.
- Write down the steps. Make sure to include even small steps that may seem trivial. The goal is to get you to consider the nuances of the experience that you may normally overlook.
- Organize the steps into a map. Display the steps sequentially in a timeline. The map may include branches to show alternative paths in the patient journey. You could also use a series of pictures or whatever method fits your data.
- Look for insights. What patterns emerge? Anything surprising or strange? Question why certain steps occur, the order they occur in, and so forth. Ask yourself how you might innovate with AI each step.
- If possible, show the map to people familiar with the journey and ask them what you’ve overlooked or gotten out of sequence.
Think about a trip to the hospital’s emergency room. Of course, the most important moment is at the point of care when the doctor is diagnosing the problem or delivering treatment. But when people complain about (or, less commonly, rave about) their emergency room experience, it’s not usually the skill of the doctor they are talking about. A simple version of the design thinking patient journey might include moments like these:
- Experience pain or discover the symptom.
- Consider home treatment versus going to the hospital: the go/no go decision.
- Choose transportation to the hospital.
- Arrive and park (or pay the taxi, etc.).
- Enter the hospital and find the emergency room.
- See the triage nurse.
- Fill out the insurance forms.
- Wait. And wait some more.
- Get ushered into a treatment room.
- Put on an uncomfortable hospital gown and wait some more.
- See multiple preliminary nurses and technicians.
- See the doctor for an assessment and sometimes preliminary diagnosis.
- Undergo additional blood tests, X-rays, and so forth.
- Receive a firmer diagnosis, which can lead to getting instructions for home care, an outpatient procedure, a prescription, a follow-up appointment with a general practitioner or specialist, or admission to the hospital.
As you lay out every step, ask yourself how you might use AI cost-effectively innovate and turn the ordinary experience into something extraordinary.
Because emergency health care is often high anxiety, I have discovered that patients are calmer if you spell out only the journey ahead. It sometimes called “journifying” the journey – taking an amorphous or scary process and breaking it down into tangible, predictable steps. I have found “journifying” helps people not only in the emergency room, but in any number of health care situations: taking your newborn home from the hospital, going in for surgery, or beginning a new treatment regimen.
The pathways or journeys in health care can be improved by the thoughtful application of AI.
Through designed thinking your organization will make better choices for this new tool.
Mark Watts is an experienced imaging professional who founded an AI company called Zenlike.ai.
