Radiology · Chest CT decision support — image retrieval, nodule and diffuse-lung quantification
contextflow ADVANCE Chest CT
contextflow GmbH
contextflow is a Medical University of Vienna and TU Wien spin-off whose chest CT suite grew out of the European KHRESMOI image-retrieval project, and its distinguishing idea is retrieval rather than a single verdict — show the radiologist visually similar, labelled cases and let the differential follow. The regulatory footprint is real but singular: an EU MDR Class IIa certificate obtained in June 2021, no FDA clearance, and no evidence of authorisation outside the European Economic Area. The published evidence is genuine and prospective but thin and manufacturer-affiliated: the flagship reader study cut interpretation time by roughly a third across eight radiologists, yet the accompanying gain in diagnostic accuracy did not reach statistical significance, and no randomised or outcome trial has been published. Real European deployment — a Dutch teaching hospital using it in routine care since 2022, a Belgian university pilot, and a hundred-plus German practices through a distribution partner — and a clean public safety record earn the device one mark; a single jurisdiction, the absence of FDA clearance and of any independent product validation, and an unproven effect on diagnostic accuracy hold it to one.
Performance Metrics
Clinical Evidence
The load-bearing clinical evidence for contextflow ADVANCE Chest CT is a single prospective, manufacturer-affiliated reader study, and its results should be read for exactly what they show and no more. In that study (Röhrich et al., European Radiology 2023; PMID 35779087), eight radiologists — four residents and four attendings — interpreted 108 chest CTs spanning twenty-two clinically and histopathologically verified diagnoses of diffuse parenchymal lung disease. Each reader read half the cases with the content-based image retrieval system and half without, in a crossover design. Using the retrieval system cut reading time by 31.3% (p<0.001), a clear and statistically significant efficiency gain. Diagnostic accuracy trended upward with the system — 42.2% versus 34.7% — but that difference did not reach significance (p=0.083). The honest reading is that the tool demonstrably speeds interpretation and points toward, but does not prove, better accuracy; the study was not powered to settle the accuracy question and several authors disclosed relationships with contextflow. Beyond that reader study the product-specific literature is sparse. A 2026 methods paper in Biology Methods and Protocols (PMID 42238221) used ADVANCE Chest CT to quantify interstitial lung disease in systemic sclerosis, but its purpose was to test whether non-contiguous CT slice-sets could be rebuilt for AI analysis rather than to validate the product's diagnostic performance. A 2021 Nature Communications paper (PMID 34584080) from contextflow-linked authors addresses continual-learning methods in medical imaging rather than the commercial device. No randomised controlled trial and no clinical-outcome study of the product has been published as of the review date. The device's demonstrated value is standardised retrieval and quantification that makes the radiologist faster and more consistent, not autonomous diagnosis.
| Study | Design | n | Sensitivity | Specificity | AUC | Published |
|---|---|---|---|---|---|---|
| Röhrich S, Sima DM, Langs G, et al. (manufacturer-affiliated; crossover reader study) | ProspectiveProspective | 108 | — | — | Diagnostic accuracy 42.2% vs 34.7% (p=0.083); reading time −31.3% (p<0.001) | Eur Radiol, 2023; PMID 35779087 (108 chest CTs, 22 verified diagnoses, 8 readers) |
Clinical Pulse
Verified clinician sentiment
Verified clinician reviews launching soon · Apply to contribute →
Reviews are signed by clinicians verified against their regulator's public register.
Inside the algorithm
Editorial featureHow contextflow turns a chest CT into a reference and a report.
Five stages — from raw input to verdict — drawn from manufacturer documentation and the public regulatory record.
- INGEST
- NORMALISE
- DETECT
- LOCALISE
- VERDICT
Stage 01 · INGEST
A routine chest CT arrives from the worklist — no extra scan.
contextflow ADVANCE Chest CT interfaces with the and receives standard chest CT studies as they are acquired. No additional acquisition, contrast or radiation is required beyond the CT the clinician already ordered, and the source images are never altered — the outputs are separate quantitative reports and retrieved reference cases.
The product is a concurrent aid: it runs in the background while the radiologist prepares to read, rather than gating the study or removing it from the worklist.
Input
Chest CT DICOM
Inference
Background, per-study
Inside the Auris+ Listing
Five more sections complete this device’s Auris+ Listing.
Decision Ledger
ProPro unlocks a private, cross-vendor log of every case you read on this device — what the AI called, what you concluded, and a one-line reason if you overrode it. Ready for the EU AI Act's deployer logging obligations when they land in 2028.
Clinical Evidence Deep Dive
ProPro unlocks the structured clinical-evidence summary — study count, target patient population, and a tabular accuracy-metrics view drawn from peer-reviewed sources.
Peer-Reviewed Publications
ProPro unlocks the curated peer-reviewed publication list with PubMed cross-links — the citation backbone of every editorial verdict.
Post-Market & Regulatory Conditions
ProPro unlocks the post-market surveillance summary, recall record, and the conditions of approval that bound real-world use.
AI Algorithm Version History
ProPro unlocks the chronological record of algorithm version changes — what changed when, drawn from manufacturer changelogs and regulatory filings.
Regulatory Approvals
Safety Record
No contextflow recall, field safety corrective action or field safety notice was identified in publicly available sources as of July 2026. Because the product carries no FDA registration there is no US MAUDE or enforcement record to search, and the EU EUDAMED vigilance module could not be queried directly from this environment, so this should be read as "none found in public reporting" rather than an exhaustive audit. The suite is concurrent decision-support software — it generates findings, quantitative reports and retrieved reference cases alongside the radiologist's read, and does not remove studies from the worklist or modify source images — which bounds the harm surface. The principal documented limitation is that the one prospective reader study showed a significant reduction in reading time but only a non-significant trend toward higher diagnostic accuracy, so the retrieval and quantification outputs should inform, not replace, the radiologist's independent interpretation.
Intended Use & Indications
contextflow ADVANCE Chest CT is a software-only radiology aid that runs on routine chest CT studies and returns quantitative and reference outputs alongside the standard read. Its defining capability is content-based image retrieval: the radiologist marks a region on the CT and the system searches a reference database of expert-labelled thin-section scans for visually similar patterns, presenting the associated diagnoses as a differential aid. The wider suite adds detection of lung nodules and incidental pulmonary embolism, quantitative and qualitative characterisation of nineteen lung-tissue patterns associated with interstitial lung disease and COPD, emphysema quantification, and tracking of nodules across prior studies. A separately partnered module contributes a Malignancy Similarity Index that assigns a probability-style score to detected nodules. The product generates findings and quantitative reports for the radiologist to weigh against the images; it does not remove studies from the worklist, alter the source images, or issue a diagnosis autonomously. It is marketed and available in the European Economic Area and carries no United States clearance.