Advances in brain imaging captured the largest share of AuntMinnie’s Top 10 most-read radiology news stories of 2025.
Whether you consider our most popular article from 2025 to be encouraging for the predictive value of quantitative PET imaging, a sign of Alzheimer’s disease as the collective worry, or something else entirely, the findings are worth reading in case you missed them.
This collection of our most popular pieces also calls attention to radiology billing arrangements, CT’s possible connection to new cancers, and big bets on a cutting-edge approach to treating prostate cancer, among the hottest topics in radiology in 2025. Snapshots we’ve curated below make it easy for you to review what AuntMinnie’s members read the most.
1. PET scans identify people at early risk for cognitive decline
Our most-read article of the year focused on quantitative PET imaging and the work of German researchers. They used F-18 florbetaben brain PET scans to measure beta-amyloid protein accumulation over five years in people experiencing subjective cognitive decline.
For their study, Guilherme Kolinger, MD, of Life Molecular Imaging in Berlin, and colleagues zeroed in on the relationship between conversion to mild cognitive decline and baseline amyloid levels. Estimating the annualized rate of change in Centiloid levels for four groups, the group found significant differences in longitudinal amyloid accumulation depending on baseline Centiloid level.
Subjective cognitive decline was defined as the self-perception of cognitive decline that is not yet captured on standard cognitive tests, and the subjective cognitive decline population provides an early window to study and possibly understand Alzheimer’s disease before the brain is irreversibly damaged, the researchers explained. However, they were uncertain if the cognitive decline observed in all participants who converted to mild cognitive impairment was related to Alzheimer’s disease.
Currently, there is no definitive consensus in the PET community of which Centiloid level values should be used to define certain at-risk groups, including beta-amyloid positivity, according to the authors, but this study presents one model and a window into quantitative PET imaging research. It was published April 23 in the European Journal of Nuclear Medicine and Molecular Imaging.
2. Aetna sues Radiology Partners in alleged NSA billing scheme
AuntMinnie’s second most popular story of 2025 involved a lawsuit in which insurance giant Aetna accused Radiology Partners (RP) of improper radiology billing strategies.
Among Aetna’s complaints is that RP misrepresented its Florida practice Mori, Bean, and Brooks (MBB) when it used the practice’s name and tax identification number to bill for radiology services performed by other RP-affiliated radiology practices across Florida.
The lawsuit spotlights the difference between in-network and out-of-network payment rates, shining a light on a high-profile payer-provider reimbursement battle since the enactment of the No Surprises Act (NSA) and its independent dispute resolution (IDR) process.
According to Aetna, RP also wrongfully initiated tens of thousands of disputes under the IDR process. The insurer terminated its in-network contract with MBB in July 2022, after an influx of billings it said were aimed at obtaining high in-network rates, according to an updated October 2025 filing.
Then, billing on an out-of-network basis triggered allegations of violating the No Surprises Act. Many of RP’s Florida practices are in network, according to Aetna.
RP portrayed the lawsuit as an example of a profit-driven insurer dodging its existing payment obligations and using litigation to influence public perception; Aetna portrayed RP as weaponizing the IDR process to coerce a new and lucrative in-network contract for MBB. Aetna wants money back for its troubles. RP is seeking arbitration for alleged contract disputes. The lawsuit is ongoing with a jury trial set for February 1, 2027, unless the case is resolved sooner.
3. CT estimated to cause 5% of new cancer cases
CT scan radiation exposure data left a major impression on the radiology community when it was revealed that CTs could eventually account for 5% of all new cancer cases annually if current practices continue.
Rebecca Smith-Bindman, MD.
University of California, San Francisco (UCSF) researchers investigated, raising the curtain on the National Cancer Institute’s (NCI) Radiation Risk Assessment Tool (RadRAT) and other data sources used to investigate how many future cancers could result from radiation exposure from annual CT exams in the U.S.
Published April 14 in JAMA Internal Medicine, the study analyzed patient-level metadata from UCSF’s International CT Dose Registry of CT examinations from 143 U.S. hospitals and outpatient facilities associated with 22 healthcare organizations in 20 states.
Between 2007 and 2023, the number of CT exams performed in the U.S. rose 35% from 68.7 million to 93 million, an increase not completely explained by population growth, according to the study’s first author, Rebecca Smith-Bindman, MD, from UCSF. The largest number of cancers was projected to result from abdominal and pelvic CT in adults, an estimated 37,500 (37%), followed by 21,500 (21%) from chest CT, the researchers found.
Smith-Bindman and colleagues recommended reducing CT use among low-risk patients, offering imaging alternatives such as ultrasound and MRI, and reducing variation in radiation techniques across imaging centers.
In response, the American College of Radiology (ACR) issued a statement highlighting the benefits of CT.
4. SPECT/CT predicts overall survival in Pluvicto patients
For a glimpse into our continuing coverage of the rise of theranostics, this story highlighted a head-to-head comparison of post-therapeutic PET/CT and SPECT/CT imaging for monitoring patients receiving lutetium-177 prostate-specific membrane antigen (PSMA) radioligand therapy (Pluvicto, Novartis) for metastatic castration-resistant prostate cancer (mCRPC).
Researchers from University Hospital LMU Munich in Germany, University of California, Los Angeles (UCLA), Johns Hopkins, and University Grenoble Alpes in France presented new evidence that SPECT/CT scans combined with serum PSA levels are better than PET/CT scans for evaluating early patient responses to Pluvicto.
Led by Lena Unterrainer, MD, the study aimed to develop an evidence-based clinical protocol to monitor the efficacy of Lu-177 PSMA. Unterrainer and colleagues found that progressive disease determined by baseline SPECT/CT was associated with shorter overall survival compared with stable disease and with partial response, the authors noted.
Their study results were published April 24 in the Journal of Nuclear Medicine.
5. China limits exports of rare earth elements, targets CT x-ray tubes
The medical imaging community feared supply interruptions for MRI contrast agents and radiopharmaceuticals when China announced April 4 that it would limit the export of rare earth elements, including several used in medical imaging and treatments.
The export control list included gadolinium, lutetium, and yttrium. Manufacturers, however, tried to assuage fears. Bayer said it had a robust global manufacturing and supply network and did not anticipate any immediate impact. Guerbet said it didn’t source raw materials (gadolinium) from China.
At the same time, China announced two investigations related to CT x-ray tubes. Both involved Chinese x-ray tube supplier Kunshan Yiyuan Medical Technology.
First, the company requested an investigation over its concerns that certain ball-bearing CT x-ray tubes originating from the U.S. and India the company was were being “dumped” on the market in China. The situation is presumed to be still under investigation. The second investigation was said to be broader for the purpose of assessing the impact of imported CT x-ray tubes on Chinese companies.
6. New method proposed for diagnosing Alzheimer’s
In March, AuntMinnie covered interesting Alzheimer’s disease biomarker research using brain PET.
Researchers at the University of Cologne in Germany proposed that PET-derived “fill states” of amyloid, tau, and neurodegeneration — hallmark pathologic abnormalities associated with Alzheimer’s — may provide additional information for monitoring disease progression and screening at-risk individuals.
Diagnosis and monitoring of Alzheimer’s disease typically uses average standardized uptake value (SUV) ratios of PET radiotracers in regions known to be affected by the disease, explained first author Elena Doering, MD, and colleagues. However, that approach largely ignores the spatial distribution of pathologies, they said.
Therefore, Doering and colleagues used a convolutional neural network to compute fill states for amyloid, tau, and neurodegeneration using percentages of significantly abnormal voxels, the smallest unit in a PET image. They then tested the efficacy of fill states versus standard SUV ratios for reflecting stages of the disease, from subjective cognitive decline to dementia.
Higher fill states were associated with higher stages of cognitive impairment, according to the results. Suggesting fill states as a new Alzheimer’s disease biomarker, the researchers noted that one potential application of fill-state data could be to serve as a readout for clinical trials, pending proof of longitudinal efficacy.
This paper was published March 25 in Radiology.
7. USPSTF boosts DEXA in osteoporosis screening
The U.S. Preventive Services Task Force (USPSTF) kicked off the new year recommending dual-energy x-ray absorptiometry (DEXA) bone mineral density (BMD) testing, with or without fracture risk assessment, for osteoporosis screening.
The task force reported then that the prevalence of osteoporosis among U.S. residents 50 years or older is 12.6%, but higher among people 65 years or older (27.1% in women and 5.7% in men).
Aside from DEXA BMD, the USPSTF’s recommendations remained consistent with its 2018 position, as follows:
- Universal screening is recommended for women 65 years or older.
- Screening is also recommended for postmenopausal women younger than 65 years with one or more risk factors for osteoporosis.
- For men, the USPSTF found there is not enough evidence to make a recommendation about screening for osteoporosis.
Evidence reported in 28 publications prompted the new advice, according to the task force. Published January 14 in JAMA, USPSTF’s recommendations also highlighted additional tools that can be used to assess osteoporosis risk.
8. BRAID study shows promise for imaging dense breasts; experts sound off
The Breast screening Risk Adapted Imaging for Density (BRAID) randomized controlled trial caught the attention of leading breast imaging researchers, educators, and advocates — and AuntMinnie readers.
BRAID took place at 10 U.K.-based breast screening sites and involved more than 6,305 women, all undergoing supplemental imaging using abbreviated MRI and contrast-enhanced mammography (CEM).
The major finding was that abbreviated MRI and CEM detected more invasive cancers than automated breast ultrasound (ABUS), and at smaller sizes. Both modalities detected three times as many invasive cancers, according to lead author Fiona Gilbert, MD, and colleagues, who also acknowledged that ABUS resulted in lower recall and biopsy rates.
|
Comparison between supplemental breast imaging modalities in BRAID study |
|||
|
Measure |
ABUS |
Abbreviated MRI |
CEM |
|
Cancer detection rate (per 1,000 exams) |
4.2 |
17.4 |
19.2 |
|
Invasive cancer detection rate (per 1,000 exams) |
4.2 |
15 |
15.7 |
|
Recall rate |
4% |
9.7% |
9.7% |
|
Biopsy rate |
1.5% |
4.9% |
4.4% |
The breast imaging community raised several points in response:
- A standard mammogram is not enough for women with dense breasts.
- CEM is not approved by the U.S. Food and Drug Administration for screening.
- Women who want to have supplemental screening have options.
The paper was published May 21 in The Lancet.
9. Teleradiology misread lawsuit results in over $15M jury award
Lawyers leveraged a 2024 Georgia Court of Appeals ruling to establish gross negligence against Quality Nighthawk Teleradiology Group and its remote employee teleradiologist. A jury awarded $15.5 million in the case that stemmed from an overnight remote radiology read of a CT scan.
An attorney representing the plaintiffs said that unlike prior case law that required the provider to be physically located in the emergency department for gross negligence to apply, the July 2024 ruling used in this case found that it did not matter where a person is located when they provide emergency medical care to a patient located in the emergency department.
Among the key points raised, the remote radiologist’s preliminary report of completely normal, only degenerative changes differed from the emergency department radiologist’s final report that noted possible fractures. Counsel also raised the point that the read time for two critical cervical CTs was abbreviated (about 5 minutes) compared with times reported for similar studies (15 minutes), and that the teleradiology firm deviated from the standard of care.
The outcome was devastating for the patient who had fallen down some stairs and went into a Gwinnett, GA, hospital wearing a cervical collar. Based on the remote radiologist’s preliminary report, the cervical collar was removed, and within hours, the patient became a quadriplegic. He died nearly three years later. Law.com published its analysis April 4.
10. If AI finds an abnormality that a radiologist misses, who’s at fault?
Rounding out AuntMinnie’s top stories at number 10, researchers are already thinking about the value of AI in the courtroom.
A study published in May considered under which conditions a jury would be more likely to side with a plaintiff in a hypothetical court case of a radiologist being sued for missing either a brain bleed or cancer.
With the knowledge that a diagnostic AI application was used, jurors perceived the radiologist as more culpable in situations where radiologists missed an abnormality and the AI didn’t. When AI flagged the brain bleed and the radiologist did not, the participants sided with the plaintiff 72.9% of the time.
However, presenting error rates (false omission and false discovery rates) for the AI mitigated that tendency to judge radiologists more severely in the presence of dissenting AI results, wrote a team led by Grayson Baird, PhD, from Brown University in Providence, RI.
For the cancer vignette, the agreement of AI in missing the pathology neither helped nor harmed the defendant, researchers reported.
With the increased use of AI-based software assisting with diagnostics comes the concern that its use will become a factor in determining liability in court cases. Radiology researchers continue to study how the public — and potential jurors in particular — will perceive the role of these AI applications and their accuracy compared to that of human readers.
This study was published May 22 in NEJM AI.