K062490 · Mercury Computer Systems, Inc. · LLZ · Oct 27, 2006 · Radiology
Device Facts
Record ID
K062490
Device Name
VISAGE PACS/CS, RELEASE VERSION 4.1
Applicant
Mercury Computer Systems, Inc.
Product Code
LLZ · Radiology
Decision Date
Oct 27, 2006
Decision
SESE
Submission Type
Traditional
Regulation
21 CFR 892.2050
Device Class
Class 2
Attributes
Software as a Medical Device
Indications for Use
Visage PACS/CS: is a system for distributing, viewing, and processing medical images and reports within and outside health care environments. It is to be used only by trained and instructed health care professionals. Visage PACS consists of the following components: Visage PACS Storage: Visage PACS Storage offers an archiving option for long-term storage of image data Visage PACS Web: Data that are stored on the Visage PACS Storage server can be accessed simultaneously by multiple web-based viewing stations within or outside a healthcare enterprise through web clients. Visage CS: Visage CS is a client server system that uses thin client technology for distribution of 3D image data generated from image data of state-of-the-art scanning modalities Integration with other hospital information systems (HIS, RIS, CIS) is provided via special interfaces.
Device Story
Visage PACS/CS is a medical image management system for distributing, viewing, and processing images/reports. Components include Visage PACS Storage (DICOM server for archiving), Visage PACS Web (web-based 2D viewing), and Visage CS (thin-client 3D visualization). Input: DICOM images from scanning modalities via hospital network. Operation: Server stores/manages data; Visage CS streams compressed screen content to thin clients, keeping DICOM data on the server. Used in clinical environments/remote locations by physicians/staff. Output: Visualized 2D/3D images and reports on web browsers or thin clients. Supports clinical decision-making by providing diagnostic-quality images; final diagnosis remains the responsibility of the physician. Benefits: Enables remote access to large 3D datasets and integrated 2D/3D workflow across hospital enterprises.
Clinical Evidence
Bench testing only. The device underwent hazard analysis, verification, and validation testing according to the manufacturer's software development process. Integration and system-level testing were performed to confirm hazard mitigation. No clinical data was required or provided.
Technological Characteristics
Software-based PACS/CS running on general-purpose PC hardware (Windows 2003 Server). Connectivity via TCP/IP and DICOM 3.0. Supports RAID, SAN, and NAS storage. 3D visualization via thin-client streaming of compressed screen content. Security via SSL encryption and VPN. Software development follows ISO 9001:2000, ISO 13485:2003, and 21 CFR 820.
Indications for Use
Indicated for use by trained healthcare professionals for the distribution, viewing, and processing of medical images and reports within and outside healthcare environments. Contraindications: Mammographic images compressed by lossy methods and digitized film screen images must not be reviewed for primary image interpretations.
Regulatory Classification
Identification
A medical image management and processing system is a device that provides one or more capabilities relating to the review and digital processing of medical images for the purposes of interpretation by a trained practitioner of disease detection, diagnosis, or patient management. The software components may provide advanced or complex image processing functions for image manipulation, enhancement, or quantification that are intended for use in the interpretation and analysis of medical images. Advanced image manipulation functions may include image segmentation, multimodality image registration, or 3D visualization. Complex quantitative functions may include semi-automated measurements or time-series measurements.
Special Controls
*Classification.* Class II (special controls; voluntary standards—Digital Imaging and Communications in Medicine (DICOM) Std., Joint Photographic Experts Group (JPEG) Std., Society of Motion Picture and Television Engineers (SMPTE) Test Pattern).
{0}------------------------------------------------
K06 2496
Premarket Submission VISAGE PACS/CS Summary
24 Aug 2006
## A.1 510(k) Summary of Safety and Effectiveness
### A.1.1 Company Identification
Mercury Computer Systems, Inc. 199 Riverneck Road Chelmsford, MA 01824-2820 United States of America
#### A.1.2 Official Correspondent
Richard Glasheen 199 Riverneck Road Chelmsford, MA 01824-2820 United States of America Tel.: 978-967-1843 Fax: 978-256-0588 E-Mail: rglashee@mc.com
### A.1.3 Date of Submission
24 Aug 2006
#### A.1.4 Device Name
| Trade name: | VISAGE PACS/CS |
|-------------------------|-----------------------------------------------------|
| Release Version: | 4.1 |
| Common name: | VISAGE PACS/CS |
| Classification Name: | Picture Archiving and Communications System |
| Reference: | per 21 CFR 892.2050 |
| Class: | II |
| Review Panel: | Radiology |
| Product Classification: | 90 LLZ, Picture Archiving and Communications System |
Guidance for the submission of premarket notifications for Guidance document: medical image management systems (issued on july 27, 2000)
#### A.1.5 Substantial Equivalence
The Visage PACS/CS Software is substantially equivalent, in the opinion of Mercury Computer Systems Inc., to TeraRecon AquariusNet Server/Thin client (K012086, Class II).
### A.1.6 Device Description
Visage PACS is a system to distribute, view, and process medical images and reports within and outside of health care environments. It consists of the following components:
- Visage PACS Storage -
- Visage PACS Web -
- Visage CS -
#### Visage PACS Storage
OCT 2 7 2006
{1}------------------------------------------------
A server receives image data in DICOM format via the hospital network. This provides universal connections to archives, modalities and workstations. The modalities that are supported by Visage PACS Storage are listed in the DICOM Conformance Statement.
Visage PACS Storage offers an archiving option for long-term storage of image data. Only the data consistency on archive media is guaranteed, the system provider has to take own appropriate means (e.g. redundancy) for safety against data loss caused by media destruction. Without the archiving option, the Visage PACS system features no components for long-term data archiving. Additional archiving on film or in digital form is therefore necessary.
#### Visage PACS Web
Data that are stored on the Visage PACS Storage server can be accessed simultaneously by multiple web-based viewing stations within a healthcare enterprise or from elsewhere outside through web clients.
The image data transfer is done in DICOM format via the Intranet or the Internet, for example to stations located in a doctor's office, throughout hospitals or a physician's home. Strong data encryption is provided (SSL) to ensure a secure data transfer. Images can be viewed directly within a web browser (Internet Explorer). The system offers simple functions for image manipulation and measurements.
Reports can be viewed together with the images on one page.
#### Visage CS
Visage CS is a client server system that uses thin client technology for distribution of 3D image data generated from image data of state-of-the-art scanning modalities. The thin client viewer allows to view and process 3D image data. No DICOM data is transferred to the client. It remains on the 3D Application Server at any time, ensuring safe and consistent access to large 3D data throughout the hospital enerprise. Instead of image data, a stream of compressed screen content information is transmitted during interaction. If Visage CS is used in un-secure networks (e.g. WAN) third party VPN (Virtual Private Network) solutions have to be used to secure the data transfer between the Visage Server and the client machines.
### A.1.7 Intended Use
Visage PACS is a system for distributing, viewing, and processing medical images and reports within and outside health care environments. It is to be used only by trained and instructed health care professionals. Visage PACS consists of the following components:
- Visage PACS Storage
- । Visage PACS Web -
- Visage CS
-
Integration with other hospital information systems (HIS, RIS, CIS) is provided via special interfaces.
#### Visage PACS Storage
A server receives image data in DICOM format via the hospital network. This provides universal connections to archives, modalities, and workstations. The modalities that are supported by Visage PACS Storage are listed in the DICOM Conformance Statement.
Visage PACS Storage offers an archiving option for long-term storage of image data. The system guarantees the consistency of the data stored on archiving media but does not prevent data loss caused by media destruction.
#### Visage PACS Web
{2}------------------------------------------------
Data that are stored on the Visage PACS Storage server can be accessed simultaneously by multiple web-based viewing stations within or outside a healthcare enterprise through web clients.
The image data are transferred in DICOM format via the Intranet or the Internet, for example to stations located in a doctor's office, throughout hospitals or a physician's home. Strong data encryption is provided (SSL) to ensure secure data transfer. Images can be viewed directly within a web browser (Internet Explorer). The system offers simple functions for image manipulation and measurements. Reports can be viewed together with the images on one page.
#### Visage CS
Visage CS is a client server system that uses thin client technology for distribution of 3D image data generated from image data of state-of-the-art scanning modalities.
The thin client viewer allows viewing and processing of 3D image data. No DICOM data is transferred to the client. It remains on the 3D Application Server at all times, ensuring safe and consistent access to large volumes of 3D data throughout the hospital enterprise. Instead of image data, a stream of compressed screen content information is transmitted during interaction.
If Visage CS is used in insecure networks (e.g. WAN), third party VPN (Virtual Private Network) solutions have to be used to secure data transfer between the Visage Server and the client machines.
#### Important Usage Notes
Visage PACS can support physicians and/or their medical staff in providing their own diagnosis for medical cases. The final decision regarding diagnoses, however, resides with the doctors and/or their medical staff in their own area of responsibility.
Although the web and thin client technologies allow the software to be run on a variety of hardware platforms, for diagnostic purposes the user must make sure that the display hardware used for reading the images complies with state-of-the-art diagnostic requirements and currently valid laws.
Mammographic images compressed by a lossy method and digitized film screen images must not be reviewed for primary image interpretations.
The computers hosting the server-side software of Visage PACS/CS must be operated in a server room ensuring the appropriate operating and environmental conditions and access control.
#### A.1.8 Software Development
Mercury Computer Systems, Inc. certifies that the VISAGE PACS/CS software is designed, developed, tested and validated according to written procedures. These procedures identify individuals within the organization responsible for developing and approving product specifications, coding and testing, validation testing and field maintenance. All employees receive the appropriate quality system training.
The Mercury Computer Systems, Inc. Quality System is in compliance with the following voluntary and mandatory standards and regulations:
| Standard/Regulation | Title |
|---------------------|------------------------------------------------------------------------------------------------|
| 21 CFR 820 | Quality Systems Regulation |
| ISO 9001:2000 | Quality Management Systems - Requirements |
| ISO 13485:2003 | Quality Systems - Medical Devices - Particular requirements<br>for the application of ISO 9001 |
{3}------------------------------------------------
| 93/42/EEC | Medical Device Directive | |
|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|-------------------------------------------|--|
| and and the beginner for the first for the first of the first and the first and the first and the first and the first and the first and the first and the first and the first<br>(IEC) 60601-1-4<br>------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | International Electrotechnical Commission | |
{4}------------------------------------------------
## A.1.9 Substantial Equivalence Comparison Chart
Annotation:
Our Visage product line consists of 2D Web based PACS software and a separate 3D Visage Server and Thin Clients to give the end user a complete 2D/3D PACS experience. The software backend of the Visage product is based on the RADIN 3.0 product that received 510(k) approval (K043483) from the FDA In 2004. RADIN 3.1 was granted 510(k) approval to Sohard AG which was acquired by Mercury Computer Systems, Inc. in 2005.
TeraRecon unites 2D/3D into one System, AquariusNet Server and Thin Clients, but has no PACS functionality. TeraRecon also provides the same Thin Client software suite on a stand- alone workstation, Aquarius Blue.
Image /page/4/Figure/7 description: This image shows a diagram of a medical imaging system. The diagram includes components such as an Aquarius Workstation, AquariusNET, Personal Digital Light Box, and Thin Client. The diagram also shows connections between the different components, indicating how they are integrated within the system. The text in the image includes labels for the different components, as well as the phrases "Powered by VolumePro" and "Powered by TeraRecon F-1".
{5}------------------------------------------------
| | | Substantial Equivalence comparison Chart | |
|--|--|------------------------------------------|--|
|--|--|------------------------------------------|--|
| Product | Visage PACS/CS<br>4.1 | AquariusNet<br>Server/Thin client | RADIN 3.0 |
|------------------------------------------------------------------------------|-----------------------------------------------------------------------------------------------------------------------------------------------------------------|-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| Company Name | Mercury Computer<br>Systems, Inc. | TeraRecon, Inc. | Mercury Computer<br>Systems, Inc.<br>Formally SoHard<br>AG |
| 510 (k) number | TBD | K012086 | K043483 |
| General | | | |
| Networking | TCP/IP | Yes | TCP/IP |
| Image Acquisition/Communication | DICOM Compliant | Yes | DICOM Compliant |
| Image file formats | DICOM 3.0 | Yes | DICOM 3.0 |
| Imaging modalities | Multi Modality | Yes | Multi Modality |
| Platform | PC | PC | PC |
| Operating System | Windows | Windows | Windows |
| Standard Microsoft Technology | Yes | Yes | Yes |
| DICOM 3.0 Compliant | Yes | Yes | Yes |
| Patient Demographics | Yes | Yes | Yes |
| DICOM Storage SCP | Yes | Yes | Yes |
| DICOM Storage SCU | Yes | Yes | Yes |
| DICOM Query/Retrieve SCU/SCP | Yes | Yes | Yes |
| DICOM Storage Commitment | Yes | Yes | No |
| HIPAA compliant | Yes | Yes | Yes |
| Secure data transmission | SSL encryption,<br>VPN encryption | VPN, encryption | Yes |
| User authentication | Yes | Yes | Yes |
| User Account | Yes | Yes | Yes |
| User groups | Yes | Yes | Yes |
| User Levels | Yes | No | Yes |
| Visage PACS/CS Online | | | |
| Scalability | | | |
| Concurrent user model | Yes | Yes | Yes |
| Data Storage | | | |
| Format | | | |
| Original Format | Yes | Yes | Yes |
| JPEG Lossless | Yes | Yes | Yes |
| JPEG Lossy | Yes (5-100%) | Yes | Yes (5-100%) |
| Wavelet | Yes (5-100%) | Yes (0-100%) | Yes (5-100%) |
| Storage Space<br>Management | | | |
| Intelligent storage<br>management | Yes | Yes | Yes |
| Data Storage | | | |
| Device | | | |
| RAID, SAN, NAS | 250GB to several<br>TB | RAID, SAN, NAS | |
| DICOM Network | | | |
| DICOM Confor-<br>mance | DICOM 3.0 | DICOM 3.0 | DICOM 3.0 |
| Hardware and Soft-<br>ware Requirements<br>for server | | | |
| Hardware<br>Software | PC 2x Pentium III<br>1,4 GHz or PC 1x<br>Pentium IV 3GHz<br>(Hyperthreading),<br>>= 1GB RAM<br>2x 60 GB Harddisk<br>SCSI or SATA<br>Windows<br>2003 | Workstation<br>Latest<br>Dell Power Solution<br>VolumePro<br>(Ter-<br>Graphics<br>rarecon)<br>Board,<br>Pentium<br>4<br>Intel<br>processor, 2.0 GHz<br>or faster<br>512MB of SDRAM<br>system memory<br>120GB Hard<br>Disk<br>Drive storage<br>Windows<br>NT<br>4.0. | PC<br>111,<br>Pentium<br>min 500MHz<br>512 MB RAM<br>20 GB Harddisk<br>Windows<br>2003 |
| | Server<br>Internet Explorer 6 | Windows 2000, Win<br>2003 | Server<br>Internet Explorer 6 |
| Workflow Features | | | |
| DICOM report inter-<br>face | Yes | Yes | Yes |
| File based report<br>interface | Yes | Yes | Yes |
| URL interface for<br>OEM integration | Yes | Yes | Yes |
| Creation of patient<br>media (DVD) ac-<br>cording to DICOM<br>standard | Yes | Yes | Yes |
| | Visage PACS/CS<br>3D Server | | |
| Scalability | Yes | Yes | Yes |
| Data Storage For-<br>mat | | | |
| Volume Data | Yes | Yes | |
| Hardware and Soft-<br>ware Requirements | | | |
| Hardware | Minimum 2x dual<br>core Opteron, 8GB<br>RAM, graphic<br>adapter 2x<br>FX4500, 1x 160GB<br>Harddisk | Workstation Latest<br>Dell Power Solution<br>VolumePro (Terare-<br>con) Graphics<br>Board,<br>Pentium 4 Intel<br>processor, 2.0 GHz<br>or faster<br>512MB of SDRAM<br>system memory<br>120GB Hard Disk<br>Drive storage | Pentium 2 GHz<br>512 MB RAM<br>50-100 GB Hard-<br>disk |
| Software | SuSe Linux version<br>10.0 | Windows NT 4.0,<br>Windows 2000, Win<br>2003 | Windows 2003<br>Server<br>Internet Explorer 6 |
| Workflow Features | | | |
| URL interface for<br>OEM integration | Yes | Yes, added capabil-<br>ity | Yes |
| | Visage PACS/CS<br>Archive | | |
| Storage Modules | | | |
| Harddisk RAID | Yes | Yes | |
| SAN systems | Yes | Yes | |
| NAS systems | Yes | Yes | |
| Data Security | | | |
| Data verification on<br>media | Yes | No | Yes |
| Manipulation detec-<br>tion | Yes | No | Yes |
| Database consis-<br>tency check | Yes | No | Yes |
| Archive type | PACS | PACS or UNC<br>mount point storage | PACS |
| Hardware and Soft-<br>ware Requirements | | | |
| Hardware | Pentium 2 GHz<br>512 MB RAM<br>50-100 GB Hard-<br>disk | Standard Dell High<br>Performance Prod-<br>uct | |
| Software | Windows 2003<br>Server<br>Internet Explorer 6 | Windows NT 4.0,<br>Windows 2000 | |
| | Visage PACS/CS<br>WEB Client | | |
| Key Features | | | |
| Full DICOM images<br>on clients | Yes | Yes | No |
| No Software Installa-<br>tion, just Internet<br>Explorer needed | Yes | No, thin client also<br>for 2D part. Also<br>has activeX client<br>that can run within<br>IE and does not<br>require client<br>download. | Yes |
| Supported Modali-<br>ties | CR, CT, DR, DS,<br>DX, ES, GM, IO,<br>MG, MR, NM, PT,<br>OT, RF, RT, US,<br>XA, XC | CR,CT,MR,DR, US,<br>PT, NM, DX, OT,<br>RF, SC, XF, XA | CR, CT, DR, DS,<br>DX, ES, GM, IO,<br>MG, MR, NM, PT,<br>OT, RF, RT, US,<br>XA, XC |
| Supported Image<br>Types | | | |
| Greyscale, Color,<br>Multiframe | Yes | Yes | Yes |
| Image Manipulation<br>Functions | | | |
| Zoom | Yes | Yes | Yes |
| Quick Zoom | Yes | Yes | Yes |
| Magnifying glass | Yes | No | Yes |
| Pan | Yes | Yes | Yes |
| Window leveling | Yes | Yes | Yes |
| Edge enhancement | Yes | Yes | Yes |
| Grayscale inversion | Yes | Yes | Yes |
| Rotating, flipping | Yes | Yes | Yes |
| MPR, MIP | Yes | Yes | No |
| Measurement<br>Functions | | | |
| Distance | Yes | Yes | Yes |
| Angulation | Yes | Yes | Yes |
| Area | Yes | Yes | Yes |
| Greyscale densitiy<br>(probe) | Yes | Yes | Yes |
| Manual distance<br>calibration | Yes | Yes | Yes |
| Cine Mode | Yes | Yes | Yes |
| Workflow features | | | |
| Database Filters | Yes | Yes | Yes |
| DICOM<br>query/retrieve from<br>archives and work-<br>stations | Yes | Yes | Yes |
| Change user and<br>group assignment of<br>patients | Yes | Yes | Yes |
| Multiple series load-<br>ing | Yes | Yes | Yes…
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.