Healthcare AI Market and FDA Clearance Dataset
Get 704 verified entities and 487 ready-to-join relationships connecting FDA-cleared AI devices with manufacturers, device classes, companies, specialties, training programs, career roles, and published research citations.
This release turns separate healthcare AI records into a market map you can actually join. Start with an FDA clearance, follow its applicant and device class, connect it to a company profile when the source supports the match, then move into specialties and research citations.
A packaged healthcare AI research set
The release contains 295 normalized clearance records that were verified for this product on October 8, 2026. It is a curated package, not a claim to reproduce every entry on the FDA website. The FDA says its public AI-enabled medical device list is updated periodically and is not comprehensive.
What version 1.1 connects
Stable IDs connect clearances to 126 normalized applicant identities and 25 canonical device classes. The relationship layer adds 295 device joins and 192 citation-to-topic joins without guessing at company matches.
The package is useful for market research, data products, retrieval systems, and editorial work. It does not contain patient data, medical advice, clinical outcome claims, or live market prices.
Use the healthcare datasets for machine learning guide to decide whether your project needs this market map, the Healthcare Search Dataset, or a credentialed clinical source.
Designed For
- healthcare AI market maps
- FDA clearance research
- company and product intelligence
- nonclinical AI research systems
Coverage
Schema Preview
| Field | Type | Description | Example |
|---|---|---|---|
k_number | string | FDA 510(k) clearance number | K243400 |
device_name | string | Cleared device name | AI-assisted imaging device |
product_code | string | FDA product code | QIH |
decision_date | date | FDA decision date | 2025-01-31 |
company_structure | string | Curated company ownership structure | public company |
applicant_id | string | Stable normalized FDA applicant identifier | manufacturer:aidoc-medical-ltd |
relationship_id | string | Stable identifier for one explicit cross-table relationship | relationship:k243400:computed-tomography-x-ray-system-qih |
company_link_status | string | States whether the record has a source-supplied curated company link | linked_curated_company |
quality_flags | string | Defined data-quality notes or none | none |
source_url | url | Originating FDA, Europe PMC, or official source | https://www.accessdata.fda.gov/ |
last_verified | date | Most recent record verification date | 2026-10-08 |
Collection And Preparation
- Normalize FDA 510(k) records by clearance number, device class, product code, applicant, specialty, decision date, and source URL, then assign stable namespaced IDs.
- Publish an explicit relationship table that connects each clearance to its manufacturer, canonical device class, specialty, and source-linked company for immediate cross-table analysis.
- Keep 126 normalized applicant identities separate from the 35 curated company profiles. A company link is added only when the source provides a matching vendor slug.
- Connect company profiles to factual market access fields while excluding share prices, market capitalization, and other time-sensitive quote data.
- Retain research citation facts and source links while excluding article abstracts from the commercial release.
- Fold one older SFH source class into the newer verified class with its original slug and title retained as aliases.
- Preserve source URLs and verification dates so each record can be checked against its originating record.
Known Limits
- The release contains no patient records, protected health information, claims, diagnoses, treatment outcomes, or clinical predictions.
- FDA clearance does not establish clinical performance for a specific patient or use case.
- Research citations support discovery. Buyers should read the linked papers before drawing conclusions.
- Company and security relationships can change after the stated verification date.
Useful Applications
- connect FDA clearances to device classes and applicants
- map healthcare AI companies and public market relationships
- compare specialties by clearance and device class coverage
- seed retrieval systems with sourced nonclinical records
- link research citations to healthcare AI subject groups
Change Log
- Clarified that the 295 clearance records are a verified packaged subset and linked the official FDA list and dataset selection guide.
- Published version 1.1.0 with 704 unique entities, 487 explicit relationships, normalized manufacturer identities, deduplicated research citations, stable IDs, provenance status, quality flags, and a corrected device-class alias.
- Published version 1.0.0 with 583 verified records, public samples, a complete codebook, and source links.