K252481 · Roche Molecular Systems, Inc. · PAB · Nov 3, 2025 · Microbiology
Device Facts
Record ID
K252481
Device Name
cobas CMV
Applicant
Roche Molecular Systems, Inc.
Product Code
PAB · Microbiology
Decision Date
Nov 3, 2025
Decision
SESE
Submission Type
Traditional
Regulation
21 CFR 866.3180
Device Class
Class 2
Indications for Use
cobas CMV is an in vitro nucleic acid amplification test for the quantitation of cytomegalovirus (CMV) DNA in human EDTA plasma. cobas CMV is intended for use as an aid in the management of CMV in solid organ transplant patients and in hematopoietic stem cell transplant patients. In patients receiving anti-CMV therapy, serial DNA measurements can be used to assess viral response to treatment. The results from cobas CMV must be interpreted within the context of all relevant clinical and laboratory findings. cobas CMV is not intended for use as a screening test for blood or blood products.
Device Story
cobas CMV is an automated in vitro diagnostic test for quantifying CMV DNA in human EDTA plasma. It utilizes the cobas 5800/6800/8800 systems for fully automated sample preparation, PCR amplification, and real-time detection. The process involves nucleic acid extraction using magnetic glass particles, followed by selective PCR amplification of the CMV DNA polymerase (UL54) gene and an internal DNA Quantitation Standard (DNA-QS). Detection is achieved via TaqMan probe-based fluorescence. The system software automatically calculates viral load results, which are reviewed by clinicians to manage transplant patients and monitor treatment response. The device is intended for clinical laboratory use.
Clinical Evidence
Clinical performance was established in the original PMA (P160041). This 510(k) provides bench-level method comparison data comparing system versions 2.0 and 1.4 using 30 positive and 31 negative clinical specimens. Results showed <0.2 log10 difference in viral load and 100% negative percent agreement.
Technological Characteristics
Quantitative real-time PCR assay. Targets conserved region of CMV DNA polymerase gene (UL54). Uses magnetic glass particles for nucleic acid extraction. Detection via TaqMan-based fluorescent reporter/quencher probes. Instrumentation: cobas 5800/6800/8800 systems. Fully automated platform. Standardized against 1st WHO International Standard for Human CMV DNA.
Indications for Use
Indicated for solid organ and hematopoietic stem cell transplant patients, including those receiving anti-CMV therapy, to quantitate CMV DNA in human EDTA plasma for viral load management.
Regulatory Classification
Identification
A quantitative cytomegalovirus (CMV) nucleic acid test for transplant patient management is identified as a device intended for prescription use in the detection of CMV and as an aid in the management of transplant patients to measure CMV deoxyribonucleic acid (DNA) levels in human plasma and/or whole blood using specified specimen processing, amplification, and detection instrumentation. The test is intended for use as an aid in the management of transplant patients with active CMV infection or at risk for developing CMV infection. The test results are intended to be interpreted by qualified healthcare professionals in conjunction with other relevant clinical and laboratory findings.
Special Controls
*Classification.* Class II (special controls). The special controls for this device are:(1) The labeling required under § 809.10(b) of this chapter must include:
(i) A prominent statement that the device is not intended for use as a donor screening test for the presence of CMV DNA in blood or blood products.
(ii) Limitations, which must be updated to reflect current clinical practice. The limitations must include, but are not limited to, statements that indicate:
(A) Test results are to be interpreted by qualified licensed healthcare professionals in conjunction with clinical signs and symptoms and other relevant laboratory results;
(B) Negative test results do not preclude CMV infection or tissue invasive CMV disease, and that CMV test results must not be the sole basis for patient management decisions.
(iii) A detailed explanation of the interpretation of results and acceptance criteria must be provided and include specific warnings regarding the potential for variability in CMV viral load measurement when samples are measured by different devices. Warnings must include the following statement, where applicable: “Due to the potential for variability in CMV viral load measurements across different CMV assays, it is recommended that the same device be used for the quantitation of CMV viral load when managing CMV infection in individual patients.”
(iv) A detailed explanation of the principles of operation and procedures for assay performance.
(2) Design verification and validation must include the following:
(i) Detailed documentation of the device description, including all parts that make up the device, reagents required for use with the CMV assay but not provided, an explanation of the methodology, design of the primer/probe sequences, rationale for the selected gene target, and specifications for amplicon size, guanine-cytosine content, and degree of nucleic acid sequence conservation. The design and nature of all primary, secondary, and tertiary quantitation standards used for calibration must also be described.
(ii) A detailed description of the impact of any software, including software applications and hardware-based devices that incorporate software, on the device's function.
(iii) Documentation and characterization of all critical reagents (
*e.g.,* determination of the identity, supplier, purity, and stability) and protocols for maintaining product integrity throughout its labeled shelf life.(iv) Stability data for reagents provided with the device and indicated specimen types, in addition to the basis for the stability acceptance criteria at all time points chosen across the spectrum of the device's indicated life cycle, which must include a time point at the end of shelf life.
(v) All stability protocols, including acceptance criteria.
(vi) Final lot release criteria, along with documentation of an appropriate justification that lots released at the extremes of the specifications will meet the claimed analytical and clinical performance characteristics as well as the stability claims.
(vii) Risk analysis and documentation demonstrating how risk control measures are implemented to address device system hazards, such as Failure Modes Effects Analysis and/or Hazard Analysis. This documentation must include a detailed description of a protocol (including all procedures and methods) for the continuous monitoring, identification, and handling of genetic mutations and/or novel CMV stains (
*e.g.,* regular review of published literature and annual in silico analysis of target sequences to detect possible primer or probe mismatches). All results of this protocol, including any findings, must be documented.(viii) Analytical performance testing that includes:
(A) Detailed documentation of the following analytical performance studies: Limit of detection, upper and lower limits of quantitation, inclusivity, precision, reproducibility, interference, cross reactivity, carryover, quality control, specimen stability studies, and additional studies as applicable to specimen type and intended use for the device.
(B) Identification of the CMV strains selected for use in analytical studies, which must be representative of clinically relevant circulating strains.
(C) Inclusivity study results obtained with a variety of CMV genotypes as applicable to the specific assay target and supplemented by in silico analysis.
(D) Reproducibility studies that include the testing of three independent production lots.
(E) Documentation of calibration to a standardized reference material that FDA has determined is appropriate for the quantification of CMV DNA (
*e.g.,* a recognized consensus standard).(F) Documentation of traceability performed each time a new lot of the standardized reference material to which the device is traceable is released, or when the field transitions to a new standardized reference material.
(ix) Clinical performance testing that includes:
(A) Detailed documentation of device performance data from either a method comparison study with a comparator that FDA has determined is appropriate, or results from a prospective clinical study demonstrating clinical validity of the device.
(B) Data from patient samples, with an acceptable number of the CMV positive samples containing an analyte concentration near the lower limit of quantitation and any clinically relevant decision points.
(C) The method comparison study must include predefined maximum acceptable differences between the test and comparator method across all primary outcome measures in the clinical study protocol.
(D) The final release test results for each lot used in the clinical study.
Predicate Devices
cobas CMV Quantitative nucleic acid test for use on the cobas 5800/6800/8800 systems (P160041)
Submission Summary (Full Text)
{0}
FDA
U.S. FOOD & DRUG
ADMINISTRATION
# 510(k) SUBSTANTIAL EQUIVALENCE DETERMINATION DECISION SUMMARY
ASSAY AND INSTRUMENT
## I Background Information:
A 510(k) Number
K252481
B Applicant
Roche Molecular Systems, Inc.
C Proprietary and Established Names
cobas CMV
D Regulatory Information
| Product Code(s) | Classification | Regulation Section | Panel |
| --- | --- | --- | --- |
| PAB | Class II | 21 CFR 866.3180 - Quantitative Cytomegalovirus Nucleic Acid Tests For Transplant Patient Management | MI - Microbiology |
## II Submission/Device Overview:
A Purpose for Submission:
To obtain clearance of cobas CMV assay on the upgraded cobas 6800/8800 systems.
B Measurand:
Cytomegalovirus (CMV) DNA
C Type of Test:
Quantitative Molecular assay - nucleic acid amplification test
Food and Drug Administration
10903 New Hampshire Avenue
Silver Spring, MD 20993-0002
www.fda.gov
{1}
K252481 - Page 2 of 9
## III Intended Use/Indications for Use:
### A Intended Use(s):
cobas CMV is an in vitro nucleic acid amplification test for the quantitation of cytomegalovirus (CMV) DNA in human EDTA plasma.
cobas CMV is intended for use as an aid in the management of CMV in solid organ transplant patients and in hematopoietic stem cell transplant patients. In patients receiving anti-CMV therapy, serial DNA measurements can be used to assess viral response to treatment.
The results from cobas CMV must be interpreted within the context of all relevant clinical and laboratory findings.
cobas CMV is not intended for use as a screening test for blood or blood products.
### B Indication(s) for Use:
See Intented Use.
### C Special Conditions for Use Statement(s):
Rx - For Prescription Use Only
### D Special Instrument Requirements:
For use on the cobas 5800/6800/8800 systems.
## IV Device/System Characteristics:
### A Device Description:
The cobas CMV is a quantitative test for use on the cobas 5800 system, cobas 6800 system and cobas 8800 system. The cobas CMV enables the detection and quantitation of CMV DNA in EDTA plasma of infected transplant patients. The viral load is quantified against a non-CMV DNA quantitation standard (DNA-QS), which is introduced into each specimen during sample processing. The DNA-QS also functions to monitor for the entire sample preparation and PCR amplification process. In addition, the test utilizes three external controls: a high titer positive, a low titer positive, and a negative control.
**cobas CMV Reagents and Controls**
cobas CMV kit
cobas CMV Control kit (Low Positive Control (CMV L(+)C), High Positive Control (CMV H(+)C))
cobas NHP Negative Control kit (Normal Human Plasma Negative Control (NHP-NC))
Reagents for sample preparation (cobas omni reagents)
**CMV primers and probes**
cobas CMV targets a conserved region of the DNA polymerase gene (UL54). One forward primer and two reverse primers are used for amplification, and detection is effected by a single probe.
{2}
Instrument and Software
The cobas 5800 system and cobas 6800/8800 systems provide a fully automated and integrated platform to perform PCR testing (specimen preparation, extraction, amplification and detection).
B Principle of Operation:
Nucleic acid from patient samples and added lambda DNA-QS molecules is simultaneously extracted. In summary, viral nucleic acid is released by addition of proteinase and lysis reagent to the sample. The released nucleic acid binds to the silica surface of the added magnetic glass particles. Selective amplification of target nucleic acid from the sample is achieved by the use of target virus-specific forward and reverse primers which are selected from highly-conserved regions of the CMV DNA polymerase (UL54) gene. Selective amplification of DNA-QS is achieved by the use of sequence-specific forward and reverse primers which are selected to have no homology with the CMV genome. A thermostable DNA polymerase enzyme is used for amplification. The target and DNA-QS sequences are amplified simultaneously utilizing a universal PCR amplification profile with predefined temperature steps and number of cycles.
The cobas CMV master mix contains one detection probe specific for CMV target sequences and one for the DNA-QS. The probes are labeled with target-specific fluorescent reporter dyes allowing simultaneous detection of CMV target and DNA-QS in two different target channels. During the PCR amplification step, hybridization of the probe to the specific single-stranded DNA templates results in cleavage by the 5'-to-3' nuclease activity of the DNA polymerase resulting in separation of the reporter and quencher dyes and the generation of a fluorescent signal. With each PCR cycle, increasing amounts of cleaved probes are generated and the cumulative signal of the reporter dye is concomitantly increased. Real-time detection and discrimination of PCR products is accomplished by measuring the fluorescence of the released reporter dyes for the viral targets and DNA-QS.
Results Interpretation
| Results | Interpretation |
| --- | --- |
| Target Not Detected | CMV DNA not detected.
Report results as “CMV not detected.” |
| < Titer Min | Calculated titer is below the Lower Limit of Quantitation (LLoQ) of the assay. Report results as “CMV detected, less than (Titer Min).”
Titer min = 34.5 IU/mL |
| Titer | Calculated titer is within the Linear Range of the assay – greater than or equal to Titer Min and less than or equal to Titer Max.
Report results as “(Titer) of CMV detected”. |
| > Titer Max^{a} | Calculated titer is above the Upper Limit of Quantitation (ULoQ) of the assay.
Report results as “CMV detected, greater than (Titer Max).”
Titer max = 1.0E+07 IU/mL |
a Sample result > Titer Max refers to CMV positive samples detected with concentrations above the upper limit of quantitation (ULoQ). If a quantitative result is desired, the original sample
K252481 - Page 3 of 9
{3}
should be diluted with CMV-negative human EDTA plasma and the test should be repeated. Multiply the reported result by the dilution factor.
## C Instrument Description Information:
1. Instrument Name:
cobas 5800/6800/8800 systems
2. Specimen Identification:
The samples are identified by the barcode on the sample tube automatically via the instrument barcode scanner or manually by using the handheld barcode reader or virtual keyboard.
3. Specimen Sampling and Handling:
Specimen processing is fully automated on the cobas 5800/6800/8800 Systems.
4. Calibration:
The user does not perform calibration of the cobas 5800/6800/8800 Systems.
5. Quality Control:
Cobas CMV control kits consist of one NHP Negative Control [(-) C] and two CMV Positive Controls, a low positive control [CMV L (+) C] and a high positive control [CMV H (+) C].
## V Substantial Equivalence Information:
A Predicate Device Name(s):
cobas CMV
B Predicate 510(k) Number(s):
P160041
## C Comparison with Predicate(s):
| Device & Predicate Device(s): | K252481 | P160041 |
| --- | --- | --- |
| Device Trade Name | cobas CMV | Same |
| General Device Characteristic Similarities | | |
K252481 - Page 4 of 9
{4}
K252481 - Page 5 of 9
| Intended Use/Indications for Use | cobas CMV is an in vitro nucleic acid amplification test for the quantitation of cytomegalovirus (CMV) DNA in human EDTA plasma.
cobas CMV is intended for use as an aid in the management of CMV in solid organ transplant patients and in hematopoietic stem cell transplant patients. In patients receiving anti-CMV therapy, serial DNA measurements can be used to assess viral response to treatment. The results from cobas CMV must be interpreted within the context of all relevant clinical and laboratory findings.
cobas CMV is not intended for use as a screening test for blood or blood products. | Same |
| --- | --- | --- |
| Conditions for use | For prescription use | Same |
| Sample Types | Human EDTA plasma | Same |
| Subject Status | Transplant patient, hematopoietic stem cell transplant patient, patients receiving anti-CMV therapy | Same |
| Sample Collection Devices | BD Vacutainer PPT Plasma Preparation Tubes for Molecular Diagnostic Test Methods or in sterile tubes using EDTA as the anticoagulant. | Same |
| Analyte Targets | Cytomegalovirus DNA | Same |
| Amplification Technology | Real-time PCR | Same |
| Detection Chemistry | Paired reporter and quencher fluorescence labeled probes (TaqMan Technology) | Same |
| Result Analysis | PCR Cycle threshold analysis | Same |
| Controls | DNA Quantitation Standard (DNA-QS) (internal control)
cobas CMV Control Kit (external positive control)
cobas NHP Negative Control Kit (external negative control) | Same |
{5}
| Traceability/Standardization | 1st WHO International Standard for Human Cytomegalovirus DNA for Nucleic Acid Amplification Technology Assay | Same |
| --- | --- | --- |
| General Device Characteristic Differences | | |
| Control Scheduling | Default setting will remain the same as the predicate device.
Additional setting possible for alternate control frequency based on lab requirements and local regulations.
Note:
- Controls will be required at least for each reagent lot change and every 72 hours.
- The new control concept is identical to the one with cobas 5800 system. | Positive control and negative control included on every amplification/detection plate |
| Instrumentation | cobas 5800 system and cobas 6800/8800 systems 2.0 | cobas 5800 system and cobas 6800/8800 systems 1.4 |
VI Standards/Guidance Documents Referenced:
None.
VII Performance Characteristics (if/when applicable):
A Analytical Performance:
1. Precision/Reproducibility:
Refer to P160041.
2. Linearity:
Refer to P160041.
3. Analytical Specificity/Interference:
Refer to P160041.
K252481 - Page 6 of 9
{6}
4. Assay Reportable Range:
Refer to P160041.
5. Traceability, Stability, Expected Values (Controls, Calibrators, or Methods):
Refer to P160041.
6. Detection Limit:
Refer to P160041.
7. Assay Cut-Off:
Refer to P160041.
8. Accuracy (Instrument):
Refer to P160041.
9. Carry-Over:
Refer to P160041.
## B Comparison Studies:
1. Method Comparison with Predicate Device:
CMV-positive and CMV-negative EDTA plasma specimens collected from the transplant patient population were used to assess the performance equivalency across three cobas CMV kit lots on one cobas 6800/8800 systems 2.0 and one cobas 6800/8800 systems 1.4. Positive samples covered three targeted viral load ranges: 3.45E+01 – 2E+03 IU/mL, 2E+03 – 2E+05 IU/mL and 2E+05 – 1E+07 IU/mL. Figure 1 shows the results of Deming Regression analysis showing the relationship between viral load concentrations measured by cobas 6800/8800 systems 2.0 and cobas 6800/8800 systems 1.4. The Deming regression analysis results have correlation coefficient (r) of 0.999, a slope of 0.984 (95% CI of 0.965 – 1.004%), and intercept of 0.031.
Figure 1. Method Comparison: Deming Regression Analysis
K252481 - Page 7 of 9
{7}
| Variable | Mean | Std Dev | Variance Ratio | Correlation |
| --- | --- | --- | --- | --- |
| Log10 titer cobas® 6800/8800 1.4 | 4.101 | 1.182 | 1 | 0.999 |
| Log10 titer cobas® 6800/8800 2.0 | 4.069 | 1.163 | | |
2. Matrix Comparison:
N/A
C Clinical Studies:
1. Clinical Sensitivity:
Refer to P160041.
2. Clinical Specificity:
Refer to P160041.
3. Other Clinical Supportive Data (When 1. and 2. Are Not Applicable):
N/A
D Clinical Cut-Off:
N/A
E Expected Values/Reference Range:
N/A
F Other Supportive Instrument Performance Characteristics Data:
K252481 - Page 8 of 9
{8}
N/A
## VIII Proposed Labeling:
The labeling supports the finding of substantial equivalence for this device.
## IX Conclusion:
The submitted information in this premarket notification is complete and supports a substantial equivalence decision.
K252481 - Page 9 of 9
Predicate graph will load when search results are available.
Embedding visualization will load when search results are available.
PDF viewer will load when search results are available.
Loading panels...
Select an item from Submissions
Click any panel, subpart, regulation, product code, or device to see details here.
Section Matches
Results will appear here.
Product Code Matches
Results will appear here.
Special Control Matches
Results will appear here.
Loading collections...
Loading
My Alerts
You will receive email notifications based on the filters and frequency you set for each alert.
Sort by:
Create Alert
Search Filters
Agent Token
Create a read-only bearer token for Claude, ChatGPT, or other agents that can call HTTP APIs.
Copy this now. It will not be shown again.
Connected apps
Apps you authorized through browser sign-in. Disconnecting revokes their access immediately.
Learn the FDA Browser
Two short videos show you everything — or skip straight to the written tutorial if you'd rather read. You can reopen this any time from the Tutorial button in the top bar.
Part 1 — Search, results, and everyday workflows 16 min
Part 2 — Embeddings: the galaxy map 3 min
1. Search: exact and fuzzy
Type a phrase like "coronary artery calcification" into the search box. You get two kinds of results. Exact results match the literal phrase — prefix searches work ("coronary artery calcificati") but suffix searches do not. Fuzzy results match on the meaning and intent of your phrase rather than the exact words, and are sorted by relevance score. Hover over the Exact or Fuzzy badge on any row to see exactly why it matched.
Use the checkboxes above the results to narrow: SaMD keeps only software-only devices, AI / ML keeps only devices with AI.
Exact vs. fuzzy search: what's the difference?
Exact matches on the literal phrase (prefix search works, suffix does not). Fuzzy matches on the meaning and intent of the phrase rather than the exact words. Hover over the badge on any row to see why it matched.
You search "coronary artery calcification" and want only software devices with AI. What two filters do you apply?
Narrow by SaMD (software-only devices), then narrow by AI/ML (devices with AI).
2. The results table
Scroll right in the results table. The intended use is extracted for you — no need to open the PDF. The device story gives a high-level snapshot of what the device does and how it's used. The AI Performance sub-table shows each output name, acceptance criteria, observed values, and development/test dataset descriptions — the same format Innolitics uses for regulatory strategy outputs, and the fastest high-level fingerprint of an AI device. It is AI-generated but has been very reliable in practice.
Where do you find a device's intended use without opening the PDF?
Scroll right in the search results table. The intended use column is extracted for you; no need to dig into the 510(k) summary PDF.
What does the AI Performance sub-table show, and why is it useful?
Output name, acceptance criteria, observed values, development dataset description, and test dataset description. It's the same format we use for regulatory strategy output and Fast 510(k) input, and the fastest high-level fingerprint of an AI device. AI-generated but reliable in practice.
3. Judging fuzzy relevance
Fuzzy results trail off in relevance as you scroll. Use three signals to decide how far down to go: the fuzzy badge explanations, the intended use column, and whether your target output (e.g., Cobb angle) still appears in the AI Performance sub-table. Once it stops appearing, you're past the relevant zone. A top hit with a low score (~0.4) and a stretched explanation is a hint the closest predicates are far away — the project may be headed for De Novo. Note the fuzzy search is a pattern match: it doesn't handle negation ("not") well, and hardware devices can appear — filter by SaMD/AI ML to cut them.
How do you judge how far down fuzzy search results to go?
Use the relevancy signals: the fuzzy badge explanations, the intended use column, and whether the target output (e.g., Cobb angle) still appears in the AI Performance sub-table. Once it stops appearing, results are trailing off in relevancy.
4. Device detail page: chat and citations
Click a device name to open its detail page: device facts on the left, a chat window on the right. Ask something like "Describe the training data". The answer carries little citation bubbles — click one to jump to the highlighted passage in the source PDF, so you can verify every AI answer against the document. There's also a Download PDF button for sharing.
How do you verify an AI chat answer on the device detail page?
Click the citation bubbles to jump to the relevant highlight in the source document.
Reading rule for every project: how many summaries do you read in full?
At least the three most relevant 510(k) or De Novo summaries, in full. After that, use targeted chat questions to confirm your memory quickly. The tool supports this professional habit — it doesn't replace it.
5. Side-by-side comparison
Select multiple rows in the results table (aim for under ~10), then open the PDF Viewer tab. Ask one question — it goes to all selected devices in parallel, each with citations. This is the fastest way to compare and contrast devices: training data, PCCP scope, how they handled adding new scanners, and so on.
What does the side-by-side PDF viewer mode do?
Select multiple devices, open the PDF viewer tab, and ask one question (e.g., "Describe the training data"). It queries all selected devices simultaneously with citations, so you can compare and contrast quickly.
6. Collections
With rows selected, go to the Collections tab and create a labeled collection (e.g., "Cobb Angle Project"). Reload that selection any time — before a client call, pull up the collection and ask questions across all of its devices at once.
How do you save a set of selected devices for later use?
Select the rows, go to the Collections tab, and create a labeled collection (e.g., "Cobb Angle Project"). You can reload the selection anytime and carry it into the PDF viewer and other tabs that support selections.
7. Product codes and the regulations tree
Click a product code in the results to jump to it in the regulations tree — identification text, sibling product codes, and devices you can open in a PDF viewer on the right. Click a regulation number to see its identification, special controls, and related product codes. You can also search by product code or regulation number at the top of the tree. Always read the special controls if any exist for your device — it broadens your search and sharpens pre-kickoff research.
What can you do from the regulations tree view?
Browse product codes and regulation numbers, read the identification text and special controls, browse sibling product codes, open device PDFs on the right, and search by product code or regulation number at the top of the tree.
8. Chart view
Click Show Chart and segment by regulation number (or product code) to see which regulations dominate your result set. Clicking a regulation takes you into the regulations tree. Great for spotting that most matches are, say, hardware laparoscopic devices — a cue to go back and filter.
How do you see which regulations dominate a search result set?
Click "Show Chart" and segment by Regulation Number. Clicking a regulation takes you to the regulations tree.
9. The predicate graph
Open the Predicates tab for a family-tree view of predicate relationships. Click a node to trace its parents and children; selections from search carry over pre-selected. Commonly predicated devices are worth reading — a lot of people predicated them for a reason. The visual lineage is also handy on client calls, e.g. to show how a predicate family evolved and justify why your predicate still holds.
In the predicate graph, why are commonly predicated devices worth reading?
A lot of people predicated them for a reason. Clicking a node traces parents and children, and selections from search carry over pre-selected.
10. Embeddings: the galaxy map
The Embeddings tab plots every matching document in a 2-D "galaxy map" where semantically similar devices cluster together. Hover or click clusters to explore, and let AI label the clusters for you. Embeddings beat product codes for grouping: two devices can carry different product codes (LLZ vs. QIH) yet do the same thing — the embedding captures the meaning of the intended use and device story. This is also exactly how retrieval-augmented generation (RAG) works under the hood, and it makes a great visual on client calls.
Try it yourself
Head to the search page and work through a few of these AI/ML fuzzy searches to build intuition: perivascular fat on CT · aortic valve calcification opportunistic screening on noncontrast CT · breast cancer prediction on digital pathology slides · autism detection · gestational age prediction · a hearing aid that can also detect a pulse · foundation model based analysis of ECG · large language models · penetration test. Watch how the relevance scores, intended use, and AI Performance tables tell you when results stop being meaningful.
