Dose Exposure vs. Diagnostic Accuracy: Radiology's Tightrope Act
How the push for lower radiation doses in imaging challenges effective diagnosis and clinical outcomes
All medical interventions inherently involve some level of risk, and imaging is no exception. In fact, every year an estimated 74 million CT procedures are performed in the U.S., with the related radiation dose representing over 60% of all patient medical radiation exposure.
To mitigate the potential risk of radiation exposure, radiology has taken significant steps towards reducing radiation exposure. For example, the popular Image Wisely® and Image Gently® campaigns focus on reducing unnecessary radiation exposure in adult medical imaging.
Meanwhile organizations like RSNA work to advance dose-reducing technologies, support proper use criteria for exams, educate professionals on low-dose protocols and encourage dose tracking via registries like the American College of Radiology Dose Index Registry.
As a result of initiatives like these, the effective dose per person in the U.S. from diagnostic and interventional X-ray procedures dropped by 20% between 2006 and 2016.
But by reducing one risk, have we created another?
That was the topic of conversation at an RSNA 2025 session about balancing the risks and benefits of medical imaging.
Risks of Prioritizing Dose Reduction Over Image Quality
“Since the amount of radiation utilized for a radiological examination is directly correlated to the quality of the diagnostic image, when we reduce radiation exposure, we could negatively impact diagnostic performance,” said Francesco Ria, DMP, a medical physicist at Duke University Health System in Durham, NC.
To illustrate, Dr. Ria noted how false-negative tests constitute a substantial proportion of all diagnosed cancers within CT screening programs. “There is a reported 8% to 15% occurrence of false negative tests in lung cancer screening studies alone,” he said.
Because there is no consistent way to measure or compare the risks and benefits of radiation, whether these false negatives can, at least in part, be attributed to reductions in radiation dose is anyone’s guess. “We measure radiation risks very thoroughly using a myriad of methods, but we don’t estimate how patients are positively affected by our work,” Dr. Ria noted.
As a result, he said that clinical improvement has focused on reducing radiation dose without studying the implications such reductions on image quality and thus diagnostic outcome.
A Total Risk Index in Diagnostic Imaging
To help fill this gap, Dr. Ria implemented a mathematical model based on a formalism proposed by Ehsan Samei, PhD, chief imaging physicist at Duke University, as part of a recent International Atomic Energy Agency (IAEA) report.
The formalism holistically considers a total risk index of diagnostic imaging, including both the radiation risk and the risk of a negative influence on the clinical benefit of an imaging procedure.
To define the total risk index, the model considers the radiation burden to the patient, disease prevalence, false positive rate, expected life-expectancy loss for an incorrect diagnosis, and the radiologist’s interpretative performance (i.e., AUC). Because risks and benefits are affected by patient sex age, and race, these factors were also included in the mathematical model.
The new model was applied to a population of one million digital twins—virtual representations of individual patients, created using real-world clinical and demographic data, that simulate how those patients might respond to different medical imaging scenarios.
For the simulated population, researchers found that clinical risk outweighed radiation risk by at least 400%.
Contrary to popular belief, the optimization application to typical abdominal CT exams demonstrated that radiation dose should be increased in over 90% of the cases as a strategy for improving diagnostic accuracy and thus reducing the overall risk to the patients.
“By taking into account the disease, conditions and demographics of an individual patient, our risk index model can help radiologists strike the right risk-benefit balance.”
— FRANCESCO RIA, DMP
It’s Not the Amount of Dose but the Right Dose for Each Patient
According to Dr. Samei, the exaggerated dose reductions highlighted by these findings can be detrimental to the quality of the radiological images and can even harm patients.
“We can’t let the risks of imaging outweigh the benefits of detecting a problem,” he said. “We do imaging for the very purpose of benefiting the patient; we cannot afford to compromise the benefit and thus undermine the very purpose of undertaking a medical procedure.”
Dr. Ria stressed that the focus shouldn’t be on the amount of dose but on finding the right dose for each patient.
“By taking into account the disease, conditions and demographics of an individual patient, our risk index model can help radiologists strike the right risk-benefit balance,” he concluded.
The researchers behind the model intend to incorporate demographic, social, epidemiological and etiological factors into the index, as well as extend it to different clinical scenarios, including various cancers, cardiac and pulmonary diseases.
Following extensive tests in virtual settings, combined with observational studies, the model will be ready for clinical use.
For More Information
Pledge to Image Wisely to show your commitment to safer, evidence-based imaging and reduce unnecessary radiation while protecting patients and reinforcing trust in the value of radiology.
Read previous RSNA News articles on radiation exposure: