Aidoc: Radiology AI Company Profile and FDA-Cleared Device Overview

Company Overview

Aidoc is an Israeli-American medical AI company headquartered in Tel Aviv, with significant US commercial operations. Founded in 2016, the company focuses almost exclusively on radiology AI — specifically on tools designed to flag time-sensitive findings in CT imaging so that radiologists can prioritize worklists before completing full reads.

The company's commercial model is built around a platform called aiOS, which aggregates multiple AI algorithms under a single integration layer connecting to hospital PACS and RIS systems. Rather than selling individual point solutions, Aidoc positions aiOS as an infrastructure layer that can run its own cleared algorithms alongside third-party tools from other vendors. This platform approach distinguishes it from competitors that sell standalone single-indication products.

FDA-Cleared Device Portfolio

Aidoc has accumulated one of the larger portfolios of FDA-cleared radiology AI indications among US-market vendors. Nearly all clearances have come through the 510(k) pathway, with the company identifying predicate devices from earlier computer-aided detection (CAD) software categories. The cleared indications span neurological emergencies, pulmonary pathology, vascular events, and musculoskeletal findings.

Each cleared algorithm is intended as a triage and notification tool — not a diagnostic replacement. The standard intended use language across Aidoc's clearances positions the software as an aid for radiologists to identify cases warranting expedited review, not as a standalone diagnostic.

Cleared Indications by Clinical Area

The aiOS Platform: What It Actually Does

The aiOS platform is worth understanding separately from the individual cleared algorithms, because it changes how Aidoc competes in the market. Most radiology AI vendors sell point solutions — a PE detection tool, an ICH detection tool — that integrate directly with a hospital's PACS. Aidoc's pitch is different: aiOS acts as a middleware layer that sits between the imaging infrastructure and the AI algorithms, handling routing, notification, and workflow integration once.

The practical implication is that hospitals adopting aiOS can, in principle, add new Aidoc algorithms or approved third-party algorithms without re-integrating at the PACS level each time. Aidoc has positioned this as a marketplace model, allowing partner AI vendors to distribute through aiOS to Aidoc's installed base.

• Worklist prioritization: Cases flagged by any active algorithm surface at the top of the radiologist's reading queue, with visual overlays indicating the finding type and confidence.
• Multi-modality routing: aiOS handles CT, CTA, and MRI series routing to the appropriate algorithm based on study type and body part.
• Notification infrastructure: Alerts can be routed to referring clinicians, ED physicians, or stroke teams — not only to radiologists — depending on the finding and configuration.
• Third-party integration: Partner algorithms from other cleared vendors can run within aiOS, with Aidoc handling the integration layer. This is a commercially significant feature that affects procurement decisions.
• Analytics dashboard: Aggregate utilization data, turnaround time metrics, and alert volumes are accessible to radiology administrators.

Clinical Evidence: What Exists and What Doesn't

Aidoc's evidence base is uneven across indications, which is typical for a company that has cleared a large number of algorithms in a short period. The ICH and LVO detection tools have the most peer-reviewed support; the newer indications have thinner published evidence.

Intracranial Hemorrhage — Best-Supported Indication

Multiple retrospective studies have evaluated Aidoc's ICH detection algorithm, and a prospective implementation study published in Radiology reported that AI-assisted triage reduced time-to-diagnosis for ICH cases. That study, conducted across multiple academic centers, found statistically significant reductions in door-to-diagnosis time when the algorithm was active — though it was not a randomized controlled trial and the effect size varied by site. A 2023 retrospective analysis across a large US health system found the algorithm's sensitivity for ICH in the 0.92–0.95 range depending on hemorrhage subtype, with specificity around 0.87–0.90.

The evidence for LVO detection is similarly retrospective-dominant. Published AUC values for LVO identification on CTA have generally been in the 0.90–0.96 range across independent validation sets, though performance varies with CT scanner type, contrast protocol, and patient population.

Pulmonary Embolism and Other Indications

The PE detection algorithm has peer-reviewed validation data, with sensitivity figures in the 0.88–0.93 range reported in published studies. However, false positive rates on CTPA — particularly for subsegmental PE — remain a documented concern. At least two published analyses noted that alert fatigue from PE false positives was a real-world deployment challenge, not just a theoretical one.

For newer indications — bowel obstruction, vertebral fracture, pericardial effusion — peer-reviewed independent validation is limited as of mid-2026. Most available data comes from Aidoc's own validation studies submitted as part of the 510(k) process, which are not publicly accessible in full. Clinicians and procurement staff should treat these indications as having regulatory clearance but limited independent evidence.

Funding and Company Stage

Aidoc has raised approximately $250 million in disclosed venture funding across multiple rounds, with General Atlantic leading a Series D round in 2023. The company remains private as of mid-2026 and has not disclosed plans for a public offering. Earlier investors include UST Global, TLV Partners, and Bessemer Venture Partners.

The company has publicly stated it serves over 1,000 hospital sites globally, though the breakdown between US and international deployments is not disclosed at the site level. Revenue figures are not public.

Deployment Footprint and Clinical Partnerships

Aidoc has disclosed partnerships or deployments with a number of US academic medical centers and large health systems, including publicly announced relationships with health systems in the Northeast and Southeast US. The company has also been referenced in deployment studies at sites in Europe and Israel, though the majority of its FDA-cleared product activity is US-focused.

One notable deployment context is the company's involvement with CMS-adjacent reimbursement discussions. Aidoc has been among the radiology AI vendors advocating for dedicated CPT codes covering AI-assisted triage services, a policy area that remains unresolved as of mid-2026. No CMS reimbursement pathway specific to Aidoc's tools has been established.

Known Limitations and Considerations

• Performance variability by scanner and protocol: Published studies consistently note that algorithm performance varies with CT scanner manufacturer, slice thickness, and contrast protocols. Hospitals using non-standard acquisition parameters should request site-specific validation data before deployment.
• Alert fatigue: Multiple implementation studies have identified alert fatigue as a real operational problem, particularly for PE and pneumothorax indications where false positive rates are higher. Threshold calibration at the site level is possible but requires IT and radiology collaboration.
• Population representation: Aidoc's published validation studies have been conducted predominantly in US and European hospital populations. Performance in underrepresented demographic groups — including patients with atypical presentations or comorbidities that alter imaging appearance — is not consistently reported.
• Thin evidence for newer indications: FDA clearance through 510(k) does not require randomized trial evidence. Several of Aidoc's more recently cleared indications have limited peer-reviewed independent validation. Clinicians should not assume that clearance implies equivalent evidence quality across all indications.
• Third-party algorithm quality on aiOS: When aiOS is used as a platform for third-party algorithms, each partner tool carries its own clearance status and evidence base. Procurement teams need to evaluate those tools independently — aiOS integration does not transfer Aidoc's regulatory standing to partner products.

Regulatory History

Aidoc has no publicly disclosed FDA warning letters, enforcement actions, or mandatory recalls as of mid-2026. The company's 510(k) submissions are on public record through the FDA's CDRH database, and the submission numbers for cleared indications can be verified there. The company has not pursued any De Novo or PMA submissions, which is consistent with its strategy of identifying predicate devices in the CAD and computer-aided triage software categories.

Aidoc's aiOS platform itself — as an integration and routing layer — is not independently cleared as a medical device. The clearances attach to the individual algorithms running on the platform. This distinction matters for hospital compliance teams assessing how to classify the software in their technology inventories.