Weekly AI digest
Radiology & medical imaging AI · week of 30 August to 5 September 2026
239 peer-reviewed papers · 6 industry and regulatory items · conference and KOL highlights. Adapted from my weekly intelligence report. Full report: FR (PDF) · EN (PDF) · RU (PDF).
Governance takeaway
What to do about it this week
This week the monitoring layer itself came under the microscope. The ACR's Assess-AI registry, the first national imaging AI registry, monitors AI outputs by having an LLM extract findings from signed reports, and its authors admit the extractor's own accuracy is not yet independently validated. Read that next to CIDER, where an extraction LLM was held to per-field accuracy and repeated-run reproducibility standards, and next to a TI-RADS study where a prompt change alone moved scoring accuracy by 28 points.
The action for a radiology leader: inventory every place an LLM silently turns text into data in your department, including inside your monitoring tools, and give each one the same acceptance test you would give a diagnostic model, with the prompt version-locked and test-retest agreement measured before its output counts as evidence.
Regulation & AI Act watch
Vara: first CE mark for autonomous breast screening triage
Certified as a Class IIb MDR device to close normal screening mammograms without radiologist review, anchored on the prospective PRAIM study of 461,818 women and the ATMON real-time supervision system. The one question to ask the vendor: what is the sensitivity in the autonomously closed normal stream, not just overall, and who receives ATMON's alarms in your program.Heartvue.Proton: FDA 510(k) for cardiac MRI quantification, with a PCCP
Cleared for viewing, post-processing and quantitative evaluation of cardiovascular MR, automating dimensions, volumes, ejection fraction and flow, on vendor-submitted validation with no published peer-reviewed data. The authorized Predetermined Change Control Plan lets models update without a new 510(k), so ask which changes it covers and how you will know a new version shipped.Aidoc First Read: Breakthrough Device Designation, not a clearance
The AI drafting preliminary chest radiograph report text received Breakthrough designation for four life-threatening findings and remains investigational. The question to hold for market entry: hallucination and omission rates per draft against final signed reports.Capital keeps flowing to imaging AI infrastructure
Elucid raised an oversubscribed 55 million dollar Series D for coronary CT analysis with FFR-CT under FDA review, and Scan.com closed 220 million to run US imaging access as an API platform. Ask any coronary AI vendor for head-to-head evidence against invasive FFR in a population like yours; treat network platforms as vendor-management, not clinical-validation, problems.CARPL and GenServe wire AI findings to autonomous patient outreach
The partnership connects radiology AI outputs to voice-AI outreach, scheduling and follow-up workflows. Before piloting, establish who clinically reviews a finding before a patient is called, and what audit trail links each call to the model version that triggered it.
Evidence you can use
Selected from 239 papers indexed in PubMed this week.
AI Monitoring AI with LLMs: The American College of Radiology's Imaging AI Registry
Journal of the American College of Radiology, 3 September 2026Impact of AI-Triaged Worklists and AI-Assisted Report Generation on Radiology Turnaround Times: Prospective Real-World Study
Journal of Medical Internet Research, 31 August 2026A Secure, Scalable Large Language Model-Based System (CIDER) for High-Throughput Clinical Data Extraction From Medical Reports: Retrospective Validation Study
Journal of Medical Internet Research, 2 September 2026From Redaction to Restoration: Deep Learning for Medical Image Deidentification and Reconstruction
Journal of Imaging Informatics in Medicine, 31 August 2026Streamlined TI-RADS Reporting From Bilingual Speech Recognition to LLM-Assisted Conclusion Generation Using a Two-Stage Workflow
Academic Radiology, 1 September 2026SlideChat is a multimodal generative artificial intelligence assistant for whole-slide computational pathology across cancer types
Nature Cancer, 1 September 2026Automatic Kidney Image Segmentation During Robot-Assisted Partial Nephrectomy Using a Deep Learning Model Based on a Multiannotator Dataset: Model Development and Validation Study
JMIR Medical Informatics, 2 September 2026Multi-task Deep Learning via UNETR for PSMA PET/CT Image for Prostate Cancer: Simultaneous Assessment of Prostate Cancer Disease Burden, Treatment Selection and Survival Outcomes
Journal of Imaging Informatics in Medicine, 31 August 2026
Post-market signal
The stronger signal this week is that post-market monitoring is becoming a certified product component rather than a paper obligation. Vara's autonomous screening CE mark rested on ATMON, a real-time supervision system grounded in seven years of continuous real-world monitoring, and the company will license ATMON independently so providers could wrap it around other autonomous tools. The regulator accepted autonomy only with a live safety net attached; if a vendor proposes any autonomous operation, ask what their equivalent of ATMON is and who receives its alarms.
The second signal is a model of post-deployment measurement from Erasmus MC, published in Radiology on 1 September: 19,190 chest CT exams the year before deploying a commercial nodule tool versus 20,133 the year after. Median reporting time fell 14.6 percent, with the largest gains for ECG-gated exams and thoracic subspecialists, while emergency department exams got 7.1 percent slower. Aggregate benefit coexisted with a measurable subgroup harm signal, visible only because the department segmented its own routine data.
Playbook snippet
One step for your AI committee
Anchor: EU AI Act Article 15 (accuracy, robustness and cybersecurity).
Step: Add reproducibility as a separate acceptance test for any LLM-based tool: run the same 30 of your own documents or cases through it twice, days apart, at the vendor's production settings. Report test-retest agreement alongside accuracy and set a threshold below which the tool does not go live, with the prompt and model version recorded in the acceptance file.
Why: A model that answers differently on Tuesday is not deployable however accurate it was on Monday, and accuracy testing alone will never surface this.
Worked example: Posta M et al. validated the CIDER extraction pipeline on 2073 Hungarian histopathology reports with exactly this design: robustness across temperatures 0 to 2.0 and technical reproducibility across 3 independent runs at temperature 0.1, alongside per-field exact-match accuracy from 99.5 percent (sex) down to 78.1 percent (tumor size). J Med Internet Res, PMID 42686197, DOI 10.2196/95780.
From the field
The ACR spent the week in the payment arena: comments filed on the CY2027 HOPPS proposed rule, cautioning against site-neutral imaging payment and pressing CMS on how Software as a Medical Service gets categorized and paid, plus three new draft QCDR measures announced for 2027 MIPS. How algorithmic services are reimbursed in the outpatient setting is being negotiated now, ahead of the final rule this fall.
RSNA's inaugural Across Africa: Advancing Cancer Imaging course is running 4 to 5 September in Cape Town, combining cancer imaging clinical sessions with capacity-building case studies from Egypt, Tanzania, Uganda and Kenya, and closing on AI innovation in global cancer care.
On LinkedIn, Curt Langlotz greeted the first autonomous breast screening approval with two words, "It's happening...", a compact marker that the debate has shifted from whether to under what supervision. Woojin Kim previewed his KCR 2026 lectures on foundation models and agentic AI and amplified the new ACR buyer's guide on calculating the AI errors that matter to your own population.
Pranav Rajpurkar reported the ReXGroundingCT Challenge at MICCAI 2026 heating up: 36 teams, top Dice up from 0.332 to 0.367, ten days left. Grounding, making a model show which pixels support each generated sentence, is the capability report-drafting AI needs before its output can be audited rather than merely read. Amine Korchi, meanwhile, announced a move into long-form video for the ideas that need more room than a feed post.
Source articles are indexed in PubMed with verified DOIs. Manuscript-stage work is excluded. The full weekly report is produced in French.
Get the weekly digest
One email a week on what matters in radiology and imaging AI. Researched by a radiologist and imaging informatics professional. Free, no vendor sponsorship.