Medical AI Evaluation

We aggregate 130 FDA approved AI devices and report how each device was evaluated.

View the Code Read the Paper

Device Database

approval_number
device_name
classification_product_code
date_received
decision_date
indications_for_use
body_area
modality
predicate_numbers
sample_size
num_sites
inc_exc_criteria
has_demographics
subtype_analysis
significance
severity
risk_level
keywords
summary_link

Visualizations

Device Summaries

Device Name:
Approval Number:
Decision Date:
Indications for Use:
Body Area:
Modality:
Sample Size:
Number of Sites:
Risk Level:
Has Inclusion/Exclusion Criteria?:
Has Demographic Data?:
Has Subtype Analysis?:

Methodology Notes

Evaluation feature details:
  • Note: The data collected for the database is based only on publicly available information from FDA.gov.
  • Sample_size: Some reports contain multiple sub-studies with different sample sizes. In these cases, we report only the maximum sample size across sub-studies.
  • Risk_level: Refers to the overall importance of the device toward treatment of a disease. The four tiers are based on IMDRF guidelines detailed here.
  • Is_prospective: Prospective studies are defined as clinical studies in which the data is collected concurrently with device deployment.
  • We report 'NA' values when the summary report does not contain information regarding the feature.