FDA AI/ML Device Predicate Networks

Explore the FDA predicate relationships of AI medical devices

1,614 devices · 988 internal predicate links · FDA-listed decisions through 29 June 2026
About
Filters All devices
19952026
Your browser does not support canvas. Use the search field to look up a device by submission number.

Predicate Provenance

Same-specialty AI-listed
Other-specialty AI-listed
Outside-list predicate
Predicate identity unavailable
No predicate required

FDA Pathway

510(k)
De Novo
Premarket
Scroll page · + / − or pinch to zoom · Drag to pan
Years are ordered but not evenly spaced: each is sized by how many devices it holds
Press Tab to focus the network, then use the arrow keys to move between devices in date order, Home and End for the oldest and newest, and Escape to clear the selection. The Devices tab lists the same data as a table.

Device details

Click a device node to view details

About this dashboard

Who we are

This dashboard was developed by researchers at the Institute of Global Health Innovation, Imperial College London, as part of an ongoing programme examining how AI medical devices are regulated by the FDA.

It brings together 1,614 devices from the dated FDA AI/ML list, with decisions through 29 June 2026, and the primary predicate relationships identified from public FDA documents. Alongside the full list, the Surgery, Cardiology and Neurology datasets reflect clinical scoping reviews conducted by our group. Their coverage differs, as explained below. Devices are counted by FDA submission number; separate entries can represent versions or supplements of the same commercial product.

How the FDA authorises AI devices

The devices in this dashboard entered through three regulatory pathways:

  1. 510(k) clearance: the manufacturer demonstrates substantial equivalence to a legally marketed predicate. Supporting evidence may include laboratory, software and clinical performance data, depending on the device.
  2. De Novo classification: a route for novel devices without a suitable predicate when general controls, or general and special controls, can address their risks. An authorised device can then serve as a predicate for later 510(k) submissions.
  3. Premarket Approval (PMA): a route for class III devices, through which the FDA evaluates scientific evidence of safety and effectiveness.

FDA guidance: 510(k) · De Novo · PMA.

Why does this matter?

The FDA publishes a list of AI medical devices, but their predicate relationships are not directly visible in that list. This dashboard makes those connections explorable. Predicate comparisons can support proportionate evaluation of incremental innovation, allowing new devices to build on established uses. For clinicians, the useful question is how a device and its supporting evidence relate to the patients and clinical task for which it will be used.

How to use the clinical comparison

Find a device, compare its description with its direct predicate, and open both FDA sources. Consider the intended patients, clinical task, inputs and performance evidence. Network shows the wider connections; Lineage follows a family or the selected device’s ancestry. A chain’s length alone does not establish safety, benefit or the adequacy of evidence.

FDA lead specialty and mapped specialty

FDA lead specialty is the classification assigned in the FDA AI/ML list. Radiology accounts for 76% of devices under this classification.

Mapped specialty reflects our research group’s clinical reviews. The original reviews cover 661 devices, of which 659 occur in the current list. Some devices are relevant to more than one specialty. For example, a radiological tool for detecting intracranial haemorrhage may also be relevant to neurosurgical and stroke care.

Both classifications appear in the filters and device details. The specialty datasets are historical subsets with different coverage, and the retained same/other specialty labels do not establish similarity of clinical application. Not mapped identifies devices without a clinical mapping here.

Key observations

Across all listed devices, 555 (34.4%) cite a predicate outside the FDA AI/ML list. This describes list membership; it does not establish whether the predicate uses AI. Predicate identity could not be established for 10 devices (0.6%). A further 61 devices (3.8%) are listed under De Novo or PMA, including PMA supplements, for which a 510(k) predicate is not required.

The graph records one selected primary predicate per submission. Shared ancestry does not establish equivalence between every pair of devices, and relationships beyond the dataset boundary are not fully reconstructed. The Devices and Trends views provide the underlying data and filtered results for download.

Data, methods and citation

Read the dataset coverage, extraction methods and suggested citation, or download the data with its field dictionary. We intend to update this resource regularly as new FDA records become available. For corrections, contact Ahmad Guni.

Key terms

Hover over a term, select it with the keyboard, or tap it for a definition.

Clinical comparison

Data and methods

Explore the earlier devices used as predicates for FDA clearance of AI medical devices, and compare their clinical descriptions and FDA documents.

Using the dashboard

Search for a device and select it to compare it with its direct predicate. Network shows the wider connections; Lineage shows one family or the selected device’s ancestor path. Devices provides a searchable table, and Trends contains annual results and network statistics.

Scroll to move down the page. Use the + and − buttons or pinch to zoom the network; Ctrl/⌘-scroll also zooms. Drag the network to pan. Figures and data can be downloaded from the relevant view.

Reading the connections

Each link represents one selected primary predicate. A family includes earlier devices, descendants and other branches connected to them. Known predicates outside the selected dataset appear at its boundary; their earlier ancestry and additional predicates are not fully reconstructed.

“Outside-list” means that a predicate is absent from this dated FDA AI list; it does not establish whether it uses AI. “Predicate identity unavailable” means that the identity could not be established from the available public documents. De Novo and PMA submissions do not require a predicate.

Data and coverage

Clinical interpretation and regulatory context

Predicate comparisons can support proportionate evaluation of incremental innovation. When comparing a device with its predicate, consider the intended patients, clinical task, inputs and supporting performance evidence. Network position or chain length alone does not establish safety, clinical benefit or the adequacy of evidence.

For 510(k) clearance, the new device must have the same intended use and meet the requirements for technological characteristics and supporting information. De Novo provides a classification route for eligible devices without a suitable predicate; PMA assesses safety and effectiveness through a separate approval process. The evidence required depends on the device and pathway.

FDA 510(k) requirements · FDA De Novo classification

Research team and citation

Developed by researchers at the Institute of Global Health Innovation, Imperial College London. For corrections, contact Ahmad Guni. Each device links to its FDA source.