Which medical imaging AI companies provide API access for radiology triage?

Yes. If by “API access” you mean a vendor can accept imaging studies programmatically and return AI results that you can integrate into your own PACS/worklist/orchestration layer, there are several relevant companies—but an important distinction is that many radiology-AI vendors expose enterprise integration interfaces (DICOM/HL7/FHIR, cloud/on-prem gateways) rather than a public, self-serve REST API.

Companies worth evaluating

For example, Aidoc's aiOS explicitly supports integration with PACS, EHR, scheduling and reporting systems and can prioritize urgent findings in the radiology worklist. It also supports third-party algorithms through its platform.Aidoc | Clinical AI+2

RapidAI similarly provides enterprise integration across PACS, EHRs, worklists and third-party algorithms, with its Rapid Edge Cloud providing the underlying imaging infrastructure. Its strongest historical focus is acute neurovascular/stroke workflows, although the platform now covers additional clinical areas.RapidAI+1

Gleamer is particularly interesting if you're building an orchestration layer: GleamerOS advertises automatic exam routing to the appropriate AI, plus DICOM, HL7 and FHIR support and AI-driven worklist prioritization.Gleamer+1

If you specifically need a developer-facing API

I'd separate the market into three categories:

1. API/cloud inference providers — submit DICOM studies programmatically and receive structured AI results.
2. Enterprise AI platforms — integrate through DICOM/HL7/FHIR/PACS and provide triage/worklist orchestration.
3. Algorithm vendors — provide individual detection algorithms that your own platform can orchestrate.

The companies above are predominantly #2, rather than “Stripe-like” public APIs where you simply get an API key and POST a CT study.

If you're building a radiology triage product and want to call third-party AI models programmatically, I can make a more useful shortlist of 10–15 vendors with their actual API/DICOM interfaces, supported modalities/pathologies, FDA/CE status, cloud vs on-prem deployment, pricing model, and whether they allow third-party/platform embedding.

Several medical imaging AI companies provide developer APIs, webhooks, or standards-based interoperability engines (DICOM, HL7, FHIR) specifically designed to ingest imaging data, run algorithms for acute abnormality detection, and push triage results back into enterprise worklists, PACS, or clinical communication platforms.aidoc.com+2

Because medical imaging workflows require strict regulatory clearance (such as FDA 510(k) or CE Mark) and secure handling of protected health information (PHI), "API access" in this space usually means a hybrid model: cloud-hosted or on-premises DICOM-ingest APIs and RESTful result-retrieval endpoints.apis.io+1

Prominent companies offering these integration capabilities include:

• Qure.ai : Provides a well-documented Qure.ai Platform API featuring token-authenticated DICOM upload, computation initiation, and result-fetching endpoints. Their algorithms (such as qXR for chest X-rays and qER for head/chest CTs) triage critical conditions like intracranial hemorrhage, pneumothorax, and tuberculosis via programmatic REST/DICOM interfaces or on-premises gateway deployments.apis.io+1
• Aidoc : Operates through its enterprise AI operating system, Aidoc aiOS™ , which supports multi-vendor AI orchestration. Rather than a raw public web API, Aidoc integrates via standard clinical communication protocols including DICOM, HL7, and FHIR. It connects directly into PACS, Electronic Health Records (EHR), and radiology worklists to flag acute triage findings.diagnosticimaging.com+2
• Viz.ai : Focuses heavily on AI-powered triage and care coordination (spanning large vessel occlusions, pulmonary embolisms, and intracranial hemorrhage). Viz.ai integrates into hospital IT and EHR systems using standard communication protocols, coupling PACS-agnostic cloud pipelines with mobile and desktop communication suites.viz.ai+3
• DeepHealth (RadNet) : Delivers clinical AI and workflow solutions (including their Reporting Pro and TechLive suites) that blend automated imaging analysis, measurements, and draft impressions. They offer integrated deployment options, including cloud-native integrations via platforms like the AWS Marketplace and standard enterprise workflow hooks.deephealth.com+1

Would you like details on specific clinical modalities (like CT vs. X-ray) or information on whether you need an on-premises gateway versus a cloud-hosted API for your project?

Yes. If by API access you mean an external developer can send medical images to a service and programmatically receive AI findings/triage results—as opposed to merely buying an on-prem/PACS integration—the market is considerably smaller.

Companies worth evaluating

• aidoc.com — Major enterprise radiology-AI vendor. Its aiOS platform supports urgency-based prioritization/triage and integrates with PACS, EHRs, worklists and other systems. However, its public materials emphasize enterprise integration/partnerships rather than a self-service REST API, so you'd likely need a commercial/developer discussion for API access.Aidoc | Clinical AI+1
• qure.ai — Particularly relevant for automated radiology triage. Its qER is FDA-cleared for triage/notification of suspected intracranial hemorrhage, mass effect, midline shift and cranial fracture on head CT. API/deployment availability should be confirmed directly with Qure.ai.Qure.ai
• rapidai.com — Strong option for acute neurovascular and vascular imaging workflows. RapidAI's enterprise platform performs imaging analysis, triage and automated measurements, with recent expansion of its FDA-cleared imaging portfolio. Like Aidoc, access appears oriented toward enterprise deployments rather than an openly documented developer API.RapidAI+1
• a2z Radiology AI — Newer focused entrant. Its a2z-Unified-Triage is FDA-cleared to triage seven urgent conditions on adult abdomen/pelvis CT and move flagged studies up the worklist. It advertises a unified integration, but I would verify whether they offer an external API versus a clinical-system integration.a2z Radiology AI

Important distinction

There are actually three different things vendors call "API access":

1. Cloud inference API — POST a DICOM/image → receive structured AI result. Best fit if you're building your own application.
2. DICOM/HL7/FHIR integration API — send studies through DICOM, receive results via DICOM/HL7/FHIR. Typical for hospitals and PACS integrations.
3. Enterprise orchestration platform — vendor installs/connects to PACS and automatically runs algorithms and reprioritizes the worklist. Aidoc and RapidAI are particularly representative of this model.

If you're specifically looking for #1—a developer-accessible API where you can upload CT/X-ray studies and get a triage score/result back, I would narrow the list substantially rather than treating every radiology-AI vendor as an API provider.

I can also make you a table of ~15 radiology AI companies with API availability, modalities, triage indications (ICH/PE/LVO/pneumothorax/etc.), FDA clearance, DICOM support, pricing/access model, and whether the API is actually self-service.

In medical imaging and radiology triage, companies generally do not provide public, self-serve REST APIs in the traditional tech sense (like a standard developer portal where you can sign up and push DICOM files via an open web API). Because of strict FDA regulations, HIPAA compliance, and complex clinical workflow integration, access is tightly controlled via enterprise deployment, vendor-agnostic orchestration engines, or specialized healthcare marketplaces.aidoc.com+3

Instead of an open API, clinical AI triage platforms integrate via standard medical communication protocols (DICOM, HL7, and FHIR ) through the following key providers and approaches:aidoc.com+2

1. Dedicated Enterprise Triage & Workflow Platforms

These companies build proprietary triage algorithms (for stroke, pulmonary embolism, intracranial hemorrhage, etc.) and handle the backend routing directly into a hospital's PACS/RIS or worklist.viz.ai+3

• Aidoc : Offers a vendor-agnostic AI orchestration platform that installs on-premise or cloud-hybrid. It integrates into existing PACS/EHR systems via standard protocols (DICOM/HL7/FHIR) and pushes triage notifications straight to the radiology worklist or mobile devices.aidoc.com+3
• Viz.ai : Specializes in neurovascular and vascular triage. It connects directly to enterprise PACS and EHR systems using standard communication protocols to trigger real-time mobile and worklist alerts.viz.ai+2

2. Radiology AI Orchestration & Marketplace Platforms

If you are looking to plug third-party or custom triage models into an existing imaging ecosystem rather than deploying a single vendor, orchestration platforms act as the middleman with structured integration frameworks:

• Blackford Analysis : Provides a platform and integration layer that lets healthcare providers manage, deploy, and orchestrate multiple radiology AI applications through a single integration point.
• Nuance AI Marketplace : Connects developers of diagnostic and triage AI models to the massive installed base of Nuance PowerScribe users, using the PowerShare network infrastructure to distribute and run algorithms.prnewswire.com+1
• Incepto : Operates an AI platform/marketplace model in Europe/select markets, offering streamlined connectivity for multiple clinical imaging applications.

3. Open-Source & Developer-Centric Tooling (For Research/Custom Models)

If you are an engineering team or researcher trying to build or host your own triage pipeline rather than using a commercial clinical-grade tool:mixpeek.com

• Mercure DICOM Orchestrator : An open-source DICOM routing and orchestration platform designed explicitly to receive imaging studies from a PACS, pass them through containerized AI models (such as MONAI Application Packages), and route the processed results or triage tags back.mercure-imaging.org+1

Are you looking to integrate an existing commercial triage tool (like Aidoc) into a hospital network, or are you an AI developer/researcher looking to deploy a custom model via DICOM routing? Let me know and I can point you toward the right architectural path.

Yes. If by “API access” you mean a developer can send DICOM studies programmatically and receive an AI result/triage status, the field is narrower than the broader radiology-AI market.

Strongest API candidates

1. Qure.ai — clearest fit for an API-first integration

Qure.ai is the one I'd put at the top of the list if your requirement is specifically “give me an API I can integrate into my own application.”

Its Platform API explicitly supports programmatic medical-image analysis. You can upload DICOM through an authenticated API endpoint and subsequently retrieve the processed results.documentation.qure.ai+1

Importantly for triage, the API response contains a triage_status such as Critical or Routine.documentation.qure.ai

Relevant products include:

• qER — head CT, including intracranial hemorrhage and other acute findings; designed to prioritize cases for clinical review.documentation.qure.ai+1
• qXR — chest X-ray analysis, with documented API support.documentation.qure.ai
• qCT — lung CT analysis.documentation.qure.ai

The architecture is essentially:

Your application → DICOM/API → Qure AI → triage/result → your application/PACS

Qure also documents a DICOM Gateway that bridges DICOM systems with its cloud REST API, which can be useful if you're integrating with an existing PACS rather than building a pure API pipeline.documentation.qure.ai

documentation.qure.ai

2. Aidoc — excellent for enterprise radiology triage

Aidoc is particularly strong if you want acute-finding triage embedded into a hospital's existing workflow rather than a generic developer API.

Its aiOS platform orchestrates algorithms based on scan metadata and pixel data and can prioritize acute/nonacute findings. It integrates with PACS, EHR and other clinical systems.Aidoc | Clinical AI

Its radiology portfolio includes ICH, LVO, PE, pneumothorax, fractures, aortic disease and other findings, with image-based triage and notification.Aidoc | Clinical AI

The caveat is important: Aidoc does not appear to offer a Qure-style publicly documented REST API for arbitrary developers. It is much more of an enterprise clinical-AI platform, so you'd typically engage them for integration.

aidoc.com

3. RapidAI — particularly compelling for neuro/stroke

RapidAI is another strong enterprise option if your definition of triage centers on stroke/neurovascular emergencies.

Its platform supports case prioritization and integrates with PACS, EHR and clinical workflows. Its Navigator Pro product specifically provides radiology case prioritization, while the platform covers NCCT, CTA, CTP, aneurysm and other neurovascular applications.RapidAI+1

Again, I'd classify this as enterprise API/integration rather than a self-service developer API.

rapidai.com

4. Lunit

Lunit has a particularly relevant product for chest-X-ray triage: Lunit INSIGHT CXR Triage, which was FDA-cleared to prioritize suspected critical findings such as pneumothorax and pleural effusion.Lunit

It's a good candidate if your use case is specifically:

CXR → detect potentially emergent finding → prioritize/notify But its commercial model is more integrated enterprise software than an openly documented developer API.

lunit.io

What I'd shortlist

If you're building a software product and need programmatic access rather than simply buying a PACS plugin, I'd investigate them in this order:

1. Qure.ai — best documented API fit; actual DICOM REST workflow and explicit triage status.
2. Aidoc — strongest broad acute-radiology/enterprise workflow platform, but API access needs to be negotiated.
3. RapidAI — particularly strong for stroke/neurovascular triage.
4. Lunit — attractive for CXR-focused triage.
5. AZmed / Gleamer — worth investigating for specific modality/pathology workflows.

One important distinction: “has an API” and “will sell API access to a third-party SaaS product” are not necessarily the same thing. Many medical-AI vendors expose APIs internally to enterprise customers while restricting commercial third-party use, patient-data processing, volume, and redistribution.

If you're evaluating these for building your own radiology triage API/SaaS, I can also make you a more useful comparison of 10–15 vendors with API availability, DICOM input, REST/FHIR support, cloud vs. on-prem, FDA clearance, turnaround time, pricing model, and which findings they can triage.

Yes. If by “API access” you mean a vendor will let you programmatically submit DICOM studies and retrieve an AI triage result—rather than merely offering PACS integration—the field is narrower than the broader radiology-AI market.

Strongest API-access candidates

Qure.ai is probably the clearest match if you're building an application yourself. Its Platform API explicitly provides a programmatic interface for medical-image analysis. Its documentation describes uploading studies and retrieving results, including a triage_status of Routine/Critical. Qure.ai Documentation Qure.ai Documentation Qure's qER, for example, is FDA-cleared for computer-assisted triage and notification of suspected intracranial hemorrhage, mass effect, midline shift and cranial fracture on head CT.Qure.ai+1

RapidAI is particularly interesting if your focus is acute neurovascular or vascular triage. It has a Technology Partner API specifically intended for integrating Rapid results into third-party products, although access is commercial rather than an open self-service developer API.APIs.io+1

Lunit has API-based integration for its imaging products, but it is similarly more of a manufacturer/enterprise integration than a freely accessible developer API. Its CXR Triage product is specifically designed to prioritize emergency chest-X-ray cases.Lunit+1

Annalise.ai is another notable option: its Enterprise documentation explicitly describes a “Worklist Triage via External API”, using HTTPS POST with JSON to communicate priority information to a RIS/worklist system.Harrison.ai

Companies I'd distinguish from the API-first group

Aidoc is a major player for radiology triage, but I would categorize it as an enterprise AI orchestration/integration platform, not a straightforward public inference API. Its aiOS integrates with PACS, EHR and worklists and performs urgency-based prioritization, but access is generally through enterprise deployment.Aidoc | Clinical AI+1

Gleamer also provides worklist prioritization through Gleamer Copilot/GleamerOS, with deep RIS/workstation integration, but I haven't found evidence of a generally available public inference API comparable to Qure's Platform API.Gleamer+1

If you're evaluating vendors for a product

I'd shortlist them roughly like this:

1. Qure.ai — best fit for direct programmatic DICOM → AI → triage-result workflow.
2. RapidAI — best for stroke/neurovascular/vascular acute-care workflows, if you can get partner access.
3. Lunit — particularly attractive for chest X-ray triage.
4. Annalise.ai — strong if what you actually need is AI-driven RIS/worklist prioritization.
5. Aidoc — strongest candidate for large enterprise deployments and multi-pathology orchestration, but less attractive if your requirement is a self-service developer API.

One important distinction: “API available” doesn't necessarily mean “send arbitrary DICOM to a public cloud endpoint.” For medical imaging, many vendors expose an API only after contracting, validation, regulatory review, and/or deployment of a DICOM gateway.

If you tell me whether you're building (a) a radiology SaaS/PACS product, (b) an internal hospital application, or (c) a research/prototype system, I can narrow this to the 5–10 APIs that you could realistically integrate, including API docs, supported modalities/findings, deployment model, FDA status, and likely commercial accessibility.