By Ricky N. Arbuckle, CRA, MBA, RT(R)(CV)
In the opening session of the 2018 Spring Conference advanced track, Dr. Woojin Kim answered the question: What is value? He defined it as Value = Quality / Cost. To increase value, either the cost must be reduced or the quality increased. If quality is moving higher, but the cost is also moving higher, the value will be same or, at worst, it will decrease.
Dr. Kim presented a visual tool that was a triangle, with value on top and lines extending down to quality and analytics. The main focus is the line between analytics and quality. If quality is strong, analytics must be tracking and reviewing the data to ensure the information has an on-going statistic.
Business Intelligence (BI) tells us how to:
- Cut Costs
- Improve workflow
- Identify new business
The relationship between value and BI is cost. Keep cost low and quality high as new business or revenue streams are investigated.
The main product from radiology is a final report. Without a final report, the whole chain of services is worthless. Structured reports are currently at 20% for healthcare, but to increase data mining the structured reporting will move to the forefront of quality analytics. Dr. Kim stated, “AI and BI are at the center of all data points in imaging.”
One portion of the presentation focused on following up with patients. Sixty-eight to seventy-one percent of patients do not follow up with additional imaging 3 to 6 months down the road. The follow-up may become linked to payments, and this would change the dynamic of not just generating a report but the full care of the patient. In one example, a facility found a new revenue stream by employing someone to handle all follow-ups in lung screening studies. The ROI was 4x the cost of the FTE, and this was just one sub-set of imaging.
Dr. Kim then covered three main topics surrounding Artificial Intelligence (AI):
Fear: Will the machines take over jobs and change the way we do business? In imaging specifically, will it remove the radiologist? There have been a few articles suggesting that sub-specialists could be the first to feel the AI experience. AI and radiologists working together would be the first steps. Just like CAD in mammography, AI is an aid that helps the radiologist and doesn’t generate the report without any help.
Hope: A combo between radiologist and AI will be the best fit, as AI can make mistakes. A computer can be fooled to recognize item in the wrong way, and in imaging this is critical.
Hype: Currently the hype of AI has reached its highest point, and there will be a period of disillusion of the product. In 2015 the machines could identify structure and items at a rate better than a human.
This was a very thought provoking presentation with the future coming faster than some would like.
Ricky N. Arbuckle, CRA, MBA, RT(R)(CV) is a technologist at Diagnostic Imaging Services in New Orleans, LA. He can be reached at firstname.lastname@example.org.