m3taversal
a756745c18
- What: Converted 132 broken wiki links to plain text across 41 health domain files. Added Vida to the Active Agents table in CLAUDE.md. - Why: Leo's PR #15 review required these two changes before merge. - Details: Broken links were references to claims that don't yet exist (demand signals). Brackets removed so they read as plain text rather than broken links. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
26 lines
2.5 KiB
Markdown
26 lines
2.5 KiB
Markdown
---
|
|
description: The FDA has authorized 1356 AI medical devices with 1039 in radiology and Aidocs foundation model covers 14 CT conditions at 97 percent sensitivity while Viz.ai saves 31 minutes in stroke treatment where each minute costs 4 disability-adjusted life years
|
|
type: claim
|
|
domain: health
|
|
created: 2026-02-17
|
|
source: "FDA AI device database December 2025; Aidoc foundation model clearance January 2026; Viz.ai ISC 2025 multicenter study; Paige and PathAI FDA milestones 2025"
|
|
confidence: likely
|
|
---
|
|
|
|
# AI diagnostic triage achieves 97 percent sensitivity across 14 conditions making AI-first screening viable for all imaging and pathology
|
|
|
|
The FDA has authorized 1,356 AI-enabled medical devices as of December 2025, up 8.5% from the prior report. Radiology dominates: 1,039 devices (77% of all authorizations), growing from 6 clearances in 2015 to 221 in 2023. In January 2026, Aidoc received clearance for healthcare's first comprehensive foundation model -- a single triage solution covering 14 conditions on CT scans at 97% mean sensitivity (up to 98.5%) and 98% mean specificity (up to 99.7%).
|
|
|
|
Viz.ai operates across 1,700+ hospitals with 13 FDA-cleared algorithms. Clinical data showed implementation reduced stroke treatment time by an average of 31 minutes. Given that every 1-minute delay to endovascular therapy costs 4 disability-adjusted life years, this is clinically transformative. Pathology AI hit critical milestones: Paige received Breakthrough Device designation for PanCancer Detect (first AI detecting both common and rare cancer variants), and PathAI's AIM-MASH became the first AI pathology tool FDA-qualified for clinical trials. DermaSensor became the first AI device for non-dermatologist skin cancer detection, cutting missed cancers in half.
|
|
|
|
By 2035, every imaging study, lab result, and vital sign stream passes through an AI filter before human review. AI does not replace diagnosticians -- it ensures nothing gets missed. The safety net model (catching what humans miss at speed) creates the highest clinical value of any AI application in healthcare.
|
|
|
|
---
|
|
|
|
Relevant Notes:
|
|
- [[the physician role shifts from information processor to relationship manager as AI automates documentation triage and evidence synthesis]] -- diagnostic triage is one of the five AI value creation layers
|
|
- [[AI middleware bridges consumer wearable data to clinical utility because continuous data is too voluminous for direct clinician review]] -- the same AI middleware pattern applies to clinical imaging data
|
|
|
|
Topics:
|
|
- livingip overview
|
|
- health and wellness
|