A recent Radiology digest,
as sent to subscribers

Study Type: Retrospective analysis (multicentre EHR database study)

Key Question: How has utilisation of emergency department neuroimaging changed across the United States between 2016 and 2025?

Key Findings:

• Abstract content insufficient to extract specific numerical findings, effect sizes, or statistical results — the full data are not present in the submission provided.
• Study draws on over 3 million ED encounters from the Cosmos EHR dataset, suggesting substantial statistical power.
• Temporal trends in CT and/or MRI use for neurological presentations across a 9-year period are the primary focus.

Clinical Relevance: Understanding neuroimaging utilisation trends is directly relevant to UK radiology departments managing ED demand, workforce planning, and referral appropriateness — particularly given NHS pressures around imaging capacity and GIRFT (Getting It Right First Time) priorities.

Limitations: As only the title and author information were provided rather than a full abstract, a complete methodological appraisal is not possible.

Study Type: Commentary / educational review (part of a special series)

Key Question: What are the clinically relevant differences between the three FDA-approved PSMA-targeted PET radiopharmaceuticals, and how should they guide agent selection in prostate cancer imaging?

Key Findings:

• The three agents ([⁶⁸Ga]Ga-PSMA-11, [¹⁸F]DCFPyL, [¹⁸F]rhPSMA-7.3) share PSMA-binding specificity but differ in radioisotope properties, ligand structure, and clearance pathways (renal vs hepatic), affecting lesion detectability near the urinary tract or liver.
• Hepatic clearance agents may complicate pre-therapy dosimetry, where liver uptake serves as a standard reference.
• Current evidence does not demonstrate overall diagnostic superiority for any single agent; choice should be guided by clinical context and operational factors.

Clinical Relevance: As PSMA PET becomes embedded in UK prostate cancer pathways, understanding agent-specific properties is essential for accurate reporting and appropriate agent selection, particularly where access may vary by centre.

Limitations: No original data presented; intrapatient comparative evidence between agents remains sparse.

Study Type: Commentary / podcast editorial

Key Question: What career pivot experiences and insights can early-career radiologists share to inform professional development?

Clinical Relevance: Career trajectory decisions are relevant to radiology trainees and early-career clinicians in the UK, where workforce planning and subspecialisation pathways are increasingly significant within NHS radiology departments.

Key Findings:

• Abstract contains no extractable data; this is a podcast episode summary or editorial piece
• Specific content, themes, or findings cannot be determined from the available information
• No quantitative or qualitative research data are presented

Study Type: Retrospective post-market evaluation (cohort study)

Key Question: Do commercially available AI nodule detection tools perform equivalently in real-world CT practice, and do measurement differences affect guideline-based management recommendations?

Key Findings:

• Detection of actionable nodules (BTS definition) was statistically comparable across all three tools (S1: 375, S2: 341, S3: 373; p=0.73), despite significant differences in total detections and false findings
• Small absolute measurement variations caused significantly different management recommendations under BTS guidelines (S2 triggered more 1-year follow-ups than S3; p<0.001), but not under Fleischner Society guidelines
• False/spurious detections were substantial across all tools (S1: 965, S2: 720, S3: 1169 wrong findings), raising overdiagnosis concerns if used without radiologist oversight

Clinical Relevance: UK radiology departments adopting AI nodule tools should be aware that BTS guideline thresholds are particularly sensitive to inter-software measurement variability, potentially generating inconsistent follow-up pathways.

Limitations: Hypothetical management recommendations were modelled retrospectively without reference to a ground-truth standard or radiologist adjudication.

Study Type: Retrospective multicentre cohort study

Key Question: Are the probability-based HCC diagnostic categories in LI-RADS v2018 and KLCA-NCC v2022 interchangeable in clinical practice?

Key Findings:

• On MRI, LR-5 had higher PPV than KLCA-NCC "definite HCC" (93.9% vs. 91.7%); intermediate categories (LR-4 vs. "probable HCC") showed comparable PPV (68.2% vs. 64.1%)
• On CT, "probable HCC" outperformed LR-4 (83.8% vs. 73.9%, p=0.012), whilst LR-5 and KLCA-NCC "definite HCC" performed equivalently (92.9% each)
• The "indeterminate" categories diverged significantly on MRI (LR-3 HCC proportion 13.2% vs. 40.8% for KLCA-NCC, p<0.001), indicating non-equivalence

Clinical Relevance: UK centres using LI-RADS should not assume equivalent risk thresholds when interpreting literature or guidelines derived from KLCA-NCC, particularly for intermediate-risk lesions guiding surveillance or biopsy decisions.

Limitations: Retrospective design with potential selection bias; findings may not generalise beyond gadoxetic acid-enhanced MRI protocols used in an East Asian HCC-risk population.

Study Type: Retrospective cohort study

Key Question: Does temporal bone remodelling after venous sinus stenting for pulsatile tinnitus differ between isolated sigmoid sinus wall dehiscence and diverticulum-associated dehiscence?

Key Findings:

• Short-term clinical success (≤3 months) achieved in 92.9% (65/70) of patients; long-term success in 97.4% of those with confirmed remodelling
• Remodelling occurred significantly more often with diverticulum than isolated dehiscence (85.7% vs 47.1%; adjusted OR 8.08, 95% CI 1.31–49.73)
• Older age independently predicted remodelling (adjusted OR 1.14 per year, 95% CI 1.03–1.26)

Clinical Relevance: For UK neuroradiologists managing pulsatile tinnitus, pre-procedural CT characterisation of the anatomical subtype may refine patient counselling on expected structural and symptomatic outcomes following venous stenting.

Limitations: Small sample with wide confidence intervals, and follow-up CT was available in only 52/70 patients, introducing potential selection bias.

Study Type: Retrospective case-controlled study

Key Question: Does functional biliary and hepatic data from gadoxetate-enhanced MRI improve prognostic prediction in primary sclerosing cholangitis compared to established clinical scoring systems?

Key Findings:

• On multivariate analysis, only Mayo Risk Score (p=0.001) and relative enhancement of proximal extrahepatic bile ducts (REPD) (p=0.035) independently predicted adverse events; REPD <4.64 gave 70% sensitivity and 63% specificity
• ANALI non-gadolinium score (p=0.0003) was the sole significant predictor of liver transplantation need; score >3 achieved 87.5% sensitivity and 74.2% specificity
• REPD measurement is considered impractical for routine reporting but potentially useful in clinical trials

Clinical Relevance: UK radiologists reporting gadoxetate-enhanced MRI in PSC patients should consider including the ANALI score, as it predicts transplantation need and is feasible in routine practice.

Limitations: Small single-centre retrospective cohort (n=73) limits generalisability and statistical power.

Study Type: Prospective cohort study (cross-sectional MRI analysis within longitudinal cohort)

Key Question: What is the association between midlife cardiovascular risk factors and carotid plaque composition (calcification, LRNC, IPH) on 3T MRI across different ethnic groups?

Key Findings:

• Hypertension and smoking were independently associated with plaque calcification (OR 4.02, 95% CI 1.59–10.12 and OR 4.94, 95% CI 1.79–13.62 respectively); smoking also associated with LRNC (OR 2.88, 95% CI 1.30–6.40)
• IPH prevalence was low (3.9%), precluding meaningful risk factor associations
• Moroccan participants had lower calcification prevalence than Dutch; no difference was seen between South-Asian Surinamese and Dutch participants

Clinical Relevance: In a UK context with a similarly diverse population, these findings support ethnicity-aware cardiovascular risk stratification and suggest carotid plaque MRI could help target preventive intervention in high-risk midlife groups.

Limitations: Small, single-centre sample (n=356) limits statistical power for rarer plaque components and ethnic subgroup comparisons.