When you have an x-ray, MRI, or PET scan taken, you have reason to believe all of the resulting images would appear the same, no matter which imaging machine was used at what medical institution.
What if the image taken at one medical center did not look or read like the same image at a different facility? Michael Graham, MD, PhD, director of nuclear medicine with UI Health Care, talks about his research evaluating consistent imaging across multiple medical sites:
In the diagnosis of a broken bone or a tumor, is it important that medical images from one medical center be consistent to another center?
With broken bones, the imaging is actually done fairly consistently and this is not a real problem. But with more advanced imaging techniques, such as PET imaging for tumors, there's a lot of possibility for variation in technique If it's not done in a consistent way, the images cannot be interpreted in a reliable fashion.
How did this study come about?
At The University of Iowa, we have a research effort called the Image Response Assessment Team, (IRAT) which is funded through the Holden Comprehensive Cancer Center.
There are eight of these teams around the country and we work together on a national level. We decided to do a survey among the eight IRATs and other high-level clinical institutions to look at the consistency of PET imaging and we found a remarkably large variation.
What is the goal of this research?
The goal is to define the range of variation we found it in things like the dose the patient got and what time they were imaged after the injection. Another goal is to define standardization of the methodology so the images can be done in a reproducible way and interpreted reproducibly. This should lead to better acceptance of our procedures by entities such as the insurance companies and Medicare.
Why is it important, especially in clinical trials, to ensure medical images are standardized from site to site?
Especially in clinical trials, the imaging has to be done consistently or the FDA will not accept the results.
How is this study evaluating the consistency from one medical site to another to determine reliable data being conducted?
We initially sent out a survey with 30 questions. The institutions answered but some weren't filled out completely and so we contacted them to fill in the blanks. It seems to be a high-quality survey of 15 major medical centers throughout the United States.
What are the next steps in determining and ensuring consistency across multiple medical imaging sites?
At the Society of Nuclear Medicine, where I am currently president-elect, we are developing something called a clinical trials network to support and facilitate multicenter clinical trials. This research that will help us define standardization procedures as we set up this clinical trial network.
The goal is to have more than 200 sites capable of doing multicenter clinical trials. When you have that large a number, there's a huge opportunity for variation and that's just not going to be acceptable. |
