A review of more than 1,300 oncologic PET/CT scans found that a University of Toledo radiologist who received additional training in nuclear medicine was just as competent as three body imaging radiologists in reading PET/CT studies. All four readers reached sensitivity, specificity, and accuracy levels of at least 90%.
"Although the sample size is small for such an assessment, it suggests that current training is adequate in producing competent readers," said lead author and third-year radiology resident Dr. Omar Ashour from the University of Toledo.
Whose job is it?
With the rise of modalities such as PET/CT and PET/MRI, radiologists are being called upon to interpret an increasing number of hybrid images, with pressure to advance their education and skills, Ashour said during a presentation at this year's Society of Nuclear Medicine and Molecular Imaging (SNMMI) annual meeting.
"Some turf wars have been going on, mainly between nuclear medicine physicians and radiologists," he told SNMMI attendees.
There have also been issues with variations in nuclear medicine training. He cited a study by Zijlstra et al that found substantial disparities among nuclear medicine physicians in interpreting FDG-PET/CT scans of lymphoma patients, particularly an increase in false negatives.
In addition, Suzuki et al concluded that the interpretation of FDG-PET was adequately reproducible, but the group also found somewhat contradictory results in healthy subjects whose results were less reproducible among six general radiologists in the study.
"From my literature review, it seemed that nuclear medicine physicians were the specialists who had to go through more training to read PET/CT, because nuclear medicine studies in general are less anatomically inclined," Ashour told AuntMinnie.com. "So there was a bigger learning curve to learn the CT portion of the study."
How is it going?
To help smooth the rough training terrain, the American College of Radiology and the American Board of Nuclear Medicine in 2010 collaborated to accelerate the pathway to dual certification by the American Board of Radiology in diagnostic radiology and nuclear radiology.
The University of Toledo has implemented a program to better prepare radiologists by giving them training in nuclear medicine interpretation. The question is, how much progress has been made?
In the study presented at SNMMI 2016, Ashour and colleagues reviewed PET/CT interpretations made by a University of Toledo radiologist who had been trained in nuclear medicine interpretation. They compared his results with those of three radiologists trained in body imaging.
The analysis included a total of 1,307 PET/CT exams from 2011 through 2013. The nuclear medicine-trained radiologist interpreted 595 cases, while the three body imaging radiologists -- who had 25 years, 10 years, and one year of experience, respectively -- read a total of 712 cases.
Each study was categorized based on the final call of the presence or absence of malignancy or metastasis and whether it was positive or negative. The interpretations were compared with the gold standard of pathology results or clinical follow-up, with each radiologist's performance evaluated by the calculation of sensitivity, specificity, and accuracy.
The researchers found no statistically significant difference collectively between the four participants in sensitivity, specificity, and accuracy (p > 0.05). In addition, there was no statically significant difference individually between the nuclear medicine-trained radiologist and the three body imaging-trained radiologists (p > 0.05).
Confidence interval: 0.91-0.99 for all results
|Performance by specialty
|Nuclear medicine-trained radiologist
|Body imaging-trained radiologist 1
|Body imaging-trained radiologist 2
|Body imaging-trained radiologist 3
"As it turned out, at least for our criteria, both [specialties] are similarly well-trained," Ashour said. "For a nuclear medicine-trained radiologist, he or she is combining the best of both worlds."
In the future, the job of interpreting hybrid imaging could become more challenging, as the adoption of PET/MRI increases in imaging centers. The American College of Radiology and SNMMI have already established competency criteria for the training and experience needed to adequately interpret images from PET/MR brain scans.
Ashour and colleagues plan to continue their research to see how well nuclear medicine-trained and general radiologists perform as new PET tracers are introduced for clinical applications such as prostate evaluation and whole-body scans.
Copyright © 2016 AuntMinnie.com