Annalise.ai

Multi-finding chest X-ray and head CT triage AI with FDA clearance on named findings, not on the whole model

As of 1 September 2026, Annalise.ai's pricing is not published; the vendor quotes on request. Annalise. Softwr lists it under Healthcare. Annalise.ai is made by Harrison.ai, available on Web, Self-hosted.

Visit siteVisit Annalise.ai

Overview

What Annalise.ai does

Annalise.ai is the clinical radiology arm of Harrison.ai, an Australian company founded by Aengus Tran and Dimitry Tran. Its two products are Annalise Enterprise CXR, which reports on more than one hundred and twenty findings on chest radiographs, and Annalise Enterprise CTB, which covers around one hundred and thirty findings on non-contrast head CT. The models run against studies as they arrive and push results into the worklist and the reading workflow, with active and passive notification for time critical findings. The distinguishing commercial fact is the regulatory split. In Australia, Europe and the United Kingdom, Annalise carries broad regulatory clearance covering the full multi-finding output as a decision support device. In the United States the position is different: roughly ten findings across chest X-ray and head CT have 510(k) clearance for triage and notification, including pneumothorax, pleural effusion and intracranial haemorrhage, and the device notifies rather than diagnoses. A United States buyer who evaluates the Australian product and then contracts in the United States will find a materially narrower cleared indication. Harrison.ai has also begun marketing the critical care offering under the Harrison.ai name, so the Annalise brand is being folded upward, which is worth clarifying in a multi-year contract. Buyers are hospital emergency and radiology departments, national screening programmes and teleradiology providers. The trade off is that triage AI changes worklist order rather than clinical decisions, and the benefit is only realised if the department actually restructures how urgent studies are picked up. Validation is real but specific: the published work reports improvement in radiologist detection of individual findings under reader study conditions, which is not the same as demonstrating shorter time to treatment in your own emergency department.

What people use it for

An emergency department wanting pneumothorax and intracranial haemorrhage pushed to the front of the reading queue overnight when cover is thinnest
A teleradiology provider covering many small hospitals that needs consistent triage before a human reads
A screening programme in Europe where the broader CE marked decision support indication, rather than the narrow United States triage clearance, is what is being deployed
A radiology department building an AI governance programme that wants one vendor covering both chest radiograph and head CT rather than contracting two separate algorithm suppliers

Where it falls short

Concrete and checkable, so you can decide whether any of them matter to you. This is the half of a review a vendor will not write about Annalise.ai.

The United States cleared indication covers roughly ten findings for triage and notification, far fewer than the model reports, so a buyer who evaluates the full multi-finding output and then contracts in the United States gets a much narrower cleared use.
Triage AI reprioritises a worklist but does not change clinical decisions on its own, so departments that do not redesign how urgent studies are picked up see no measurable improvement in time to treatment.
Published validation is largely reader study evidence on detection of individual findings; prospective evidence of shorter time to treatment in routine emergency care is thin.
The Annalise brand is being migrated under the Harrison.ai name for the critical care offering, which creates ambiguity in multi-year contracts about what product name is actually being supported.
A high finding count raises the alert burden, and departments report that low prevalence findings generate false positives that erode radiologist trust unless the alerting thresholds are tuned locally.

Cross-shopped

What people choose instead of Annalise.ai

Each pairing was judged by two reviewers asking whether a buyer would genuinely weigh the two against each other. The ones that failed were deleted rather than published.

Annalise.aivs
Rad AI
Rad AI: If reporting time rather than detection is the bottleneck, and you want to avoid a device validation programme
Annalise.aivs
Gleamer
Gleamer: If the priority is musculoskeletal trauma on plain film rather than chest and head
Annalise.aivs
Sectra
Sectra: If you would rather buy algorithms through an imaging platform marketplace than contract directly with each AI vendor

Pricing

What Annalise.ai costs

Taken from the vendor's own pricing page. Prices move, so check before you buy.

Annalise Enterprise

On request

Priced per study analysed or as an annual volume subscription
CXR and CTB licensed separately
Cleared finding set differs by market, which changes what you are buying
Deployment, validation and integration scoped per site

Capabilities

Features

Annalise Enterprise CXR
Multi-finding chest radiograph analysis covering over one hundred and twenty findings
Annalise Enterprise CTB
Non-contrast head CT analysis covering around one hundred and thirty findings
Triage and notification
Active and passive alerting for time critical findings such as intracranial haemorrhage
Worklist reprioritisation
Reorders the reading queue rather than requiring a separate application
Localisation overlays
Highlights the region driving each finding on the image
Confidence output
Per finding confidence values rather than a single binary flag
PACS-native deployment
Results surface in the existing viewer and worklist
On-premise or cloud inference
Deployment choice for organisations with data residency constraints

Answered, with sources

Questions people ask

Each answer names the page it came from, so you can check it rather than take our word for it.

How many findings are FDA cleared?

Around ten across chest X-ray and head CT, for triage and notification only. The model reports far more findings than that; the rest are not cleared in the United States.

Is the European position different?

Yes. CE marking under the medical device regulation covers a broader decision support indication than the United States triage and notification clearances, so the same software is a different regulatory product depending on where you deploy it.

Does it diagnose?

No. In the United States it is a triage and notification device: it flags studies for earlier review, and the radiologist makes the diagnosis.

Is Annalise.ai becoming Harrison.ai?

The critical care offering is now marketed under the Harrison.ai name. Ask the vendor to name the contracted product explicitly before signing a multi-year deal.

Keep looking

Where to go from Annalise.ai

Best Healthcare software for

Compare Annalise.ai with

Other Healthcare software

Gleamer
CE marked and FDA cleared imaging AI suite for bone trauma and chest, now owned by RadNet
Pricing on request12 researched notes
Rad AI
Generative reporting and follow-up tracking for radiology, positioned outside the diagnostic device boundary
Pricing on request12 researched notes
Sectra
Enterprise imaging and PACS from a Swedish vendor that also sells the pathology and cardiology modules on one platform
Pricing on request12 researched notes
Aidoc
FDA cleared radiology AI for triage and notification of time critical findings
Pricing on request11 researched notes
Ibex Medical Analytics
Cancer detection AI for breast, prostate and gastric pathology deployed in routine clinical service, not just pilots
Pricing on request12 researched notes
Paige
Digital pathology AI holding the first FDA De Novo authorisation for a pathology algorithm, now part of Tempus
Pricing on request12 researched notes
Philips IntelliSpace PACS
Enterprise radiology PACS and diagnostic image viewing from Philips
Pricing on request12 researched notes
Fujifilm Synapse
Enterprise imaging platform covering PACS, VNA, cardiology and 3D from Fujifilm Healthcare
Pricing on request12 researched notes
Visage Imaging
Server-side streaming PACS built for radiologists reading very large studies over thin connections
Pricing on request12 researched notes
Lunit
FDA cleared and CE marked AI for mammography, tomosynthesis and chest radiography
Pricing on request11 researched notes
Techcyte
AI digital pathology across clinical, veterinary and environmental laboratory disciplines
Pricing on request11 researched notes
PathAI
Digital pathology image management with an FDA cleared platform for primary diagnosis
Pricing on request11 researched notes
Epocrates
Mobile clinical reference and decision support app
Free plan6 researched notes
Figure 1
Medical photography and knowledge sharing platform
Free plan4 researched notes
Fullscript
Supplement dispensing for practitioners, free to use because it earns on the products
Free plan9 researched notes
GNU Health
Free hospital and health information system from a small NGO, built on the Tryton framework
Open source12 researched notes
G
GoodRx
Compare prescription drug prices and find coupons
Free, then $14.99/mo12 researched notes
Halaxy
Practice management for allied health that charges nothing for the software and earns on payments
Free plan12 researched notes

Softwr does not host reviews and shows no star rating for Annalise.ai, because a rating we did not collect is not ours to publish. What is here is the pricing and platform detail from the vendor’s own pages, limitations we could state concretely, and alternatives a reviewer confirmed people weigh against it. Tell us if any of it is wrong.

Best Healthcare software alternatives

View all Healthcare software