---
title: "Kaiser's 1,200-scan trial found AI echo drafts matched clinicians and halved read time"
description: "The Permanente Medical Group ran the randomized trial of EchoPrime itself and found the AI drafts matched clinician drafts while cutting cardiologist read time nearly in half. Days later it launched Permanentech, its own vehicle to commercialize clinical AI, so the buy-or-build question just got a benchmark."
publisher: "The Inference"
section: "Health"
published: 2026-11-20T16:30:43.879Z
modified: 2026-11-20T16:30:43.879Z
canonical: https://theinference.org/article/kaiser-s-1-200-scan-trial-found-ai-echo-drafts-matched-clinician-drafts-and-cut-read-time-nearly-in-half
language: en
keywords: "Health, AI, Medical AI, Kaiser Permanente, Cardiology"
---

# Kaiser's 1,200-scan trial found AI echo drafts matched clinicians and halved read time

> The Permanente Medical Group ran the randomized trial of EchoPrime itself and found the AI drafts matched clinician drafts while cutting cardiologist read time nearly in half. Days later it launched Permanentech, its own vehicle to commercialize clinical AI, so the buy-or-build question just got a benchmark.

The Permanente Medical Group, [Kaiser Permanente](https://theinference.org/markets/companies/kaiser-permanente)'s 10,000-doctor group, funded and ran the trial itself: 1,200 echocardiograms from 21 Northern California centers, each read by a sonographer, a cardiologist and EchoPrime, an AI model licensed to [InVision Medical Technology](https://theinference.org/markets/companies/invision-medical-technology) [1]. Blinded cardiologist overreaders made substantial changes to half the AI drafts and half the clinician drafts, [meeting the trial's noninferiority bar](https://www.tctmd.com/news/ai-echo-holds-well-against-clinician-interpreters) [1]. David Ouyang, the Kaiser Permanente cardiologist behind the model, is now preparing [the FDA clearance needed to deploy it in practice](https://theinference.org/article/epic-won-fda-clearance-for-its-deterioration-score-with-rival-agilemd-s-algorithm-as-the-benchmark) [5].

## Parity on accuracy, decisive on time

Accuracy is not the money here; seconds are. A cardiologist finalizing a report from an AI draft spent nearly half the time of reading with no draft at all [1]. And the AI draft does what no human draft can do for free: it erases the nearly ten minutes a sonographer spends writing the preliminary interpretation before a cardiologist ever looks [1]. Echo demand keeps climbing as the population ages while the imaging workforce stays the same size, and Ouyang says that backlog, not accuracy, is what motivated the trial [5].

*Chart: **Any preliminary draft roughly halves the cardiologist's read time** Bar chart of mean cardiologist finalization time per echocardiogram report: 507 seconds with no preliminary draft, 282 seconds with a sonographer's draft, 276 seconds with a fellow cardiologist's draft and 274 seconds with the AI draft.*

*Mean seconds a cardiologist spent finalizing one echocardiogram report, by source of the preliminary draft, in the AI ECHO INSIGHT randomized trial of 1,200 echoes at 21 Kaiser Permanente Northern California centers. The AI draft also removes the roughly ten minutes a sonographer spends writing the preliminary interpretation. Source: TCTMD report of AI ECHO INSIGHT, AHA 2026. [1]*

## The caveats are the price list

The trial was not powered to show any effect on patient outcomes, the AI was inferior to cardiologists reading alone, and overreaders guessed which interpreter wrote a draft barely better than chance [1]. None of that kills the model. It prices it: EchoPrime is validated as a drafting tool whose final read a cardiologist still owns, not a replacement [1][5].

## Models do not travel, validated deployments do

The same week, Kaiser's research division published 49 models that predict pregnancy complications from routine medical records for roughly 443,000 pregnancies, and showed [the models lose accuracy when moved to a nationwide dataset](https://www.nature.com/commshealth/research-articles) until they are locally adapted [2][6]. That is the pattern for clinical AI: models do not travel, validated deployments do.

## The value lands with whoever owns the read layer

No listed company sells EchoPrime; it is Cedars-Sinai intellectual property licensed to the private InVision Medical Technology [5]. The value lands with whoever holds the read layer: the integrated operator banks a labor line per scan and can cap what it pays vendors, because it has now proven it can trial a vendor's model in its own halls and publish the result [1]. Permanentech, [the wholly owned innovation company TPMG launched this week](https://finance.yahoo.com/healthcare/articles/permanente-medical-group-launches-permanentech-130500797.html) to partner with outside AI developers and build physician-created tools, is the hedge on both sides of that trade [3][4].

The model gets the headline. The validation pipeline gets the margin. Kaiser has just shown it can test the best outside model against its own clinicians, and that it intends to own the pipeline that makes such tests possible.

## Takeaway

Any preliminary draft nearly halves a cardiologist's read time; the AI draft also erases the sonographer's ten minutes, and Kaiser has now proven it can trial an outside model in its own halls and publish the result.

## Sources

1. [TCTMD, AI-Echo Holds Up Well Against Clinician Interpreters, 19 November 2026](https://www.tctmd.com/news/ai-echo-holds-well-against-clinician-interpreters)
2. [Kaiser Permanente Division of Research, Electronic health records contain information that may help predict pregnancy complications, November 2026](https://divisionofresearch.kaiserpermanente.org/ehr-predict-preg-complications/)
3. [Fierce Healthcare, The Permanente Medical Group launches tech incubation company, 16 November 2026](https://www.fiercehealthcare.com/health-tech/permanente-medical-group-launches-tech-incubation-company)
4. [Business Wire via Yahoo Finance, The Permanente Medical Group Launches Permanentech to Accelerate Physician-Driven Transformation of Care Delivery, 19 November 2026](https://finance.yahoo.com/healthcare/articles/permanente-medical-group-launches-permanentech-130500797.html)
5. [Medscape, Cardiologists in the Driver's Seat for AI-Echo Success (interview with David Ouyang), 18 November 2026](https://www.medscape.com/viewarticle/cardiologists-drivers-seat-ai-echo-success-2026a10018zf)
6. [Communications Health (Nature Portfolio), Evaluation of machine learning-based prediction of pregnancy complications and adverse birth outcomes using routine data, Parent, Moraitaki, Ghassemi et al., 11 November 2026](https://www.nature.com/commshealth/research-articles)


---

Kaiser's 1,200-scan trial found AI echo drafts matched clinicians and halved read time — The Inference. Canonical: https://theinference.org/article/kaiser-s-1-200-scan-trial-found-ai-echo-drafts-matched-clinician-drafts-and-cut-read-time-nearly-in-half
