Radiology & Imaging AI
M

MEDIVIS

Medivis makes surgical imaging and augmented reality navigation software. Medivis Studio loads CT and MRI from PACS, segments anatomy automatically for routine structures such as the spine, brain tumors and full body CT and interactively for tumors and vessels, fuses MRI onto CT, and supports planning of trajectories and screws, viewed on standard screens, a 3D monitor, Microsoft HoloLens 2 or Apple Vision Pro. Medivis Navigation guides instruments during cranial and thoracolumbar spine surgery from an NVIDIA powered workstation with infrared tracking, shown on a headset, smart glasses or AR loupes, with body, cervical and fluoroscopic navigation awaiting clearance. FDA clearances cover Studio, first cleared in 2019 as SurgicalAR for image display and 3D rendering, Spine Navigation (July 2024) and Cranial Navigation, cleared as NeuroAlign in October 2025, and Medivis markets Studio as cleared for AI segmentation, which the 2019 clearance does not describe. Maia, a harness that lets vision language models call Medivis tools, and surgical robotics are research programs, not cleared products. Studio runs on Microsoft Azure or on site.

Medivis holds 46 patents granted and pending, and its research page lists case series in cranial and spine surgery, intensive care drain placement and ovarian cancer planning at Memorial Sloan Kettering. Founded in 2016 by physicians and engineers including a neurosurgeon and a radiologist, it is based in New York and raised a $5 million seed round led by Initialized Capital, then a $20 million Series A in June 2023 led by Thrive Capital, with Mayo Clinic participating.

AI Health Index verifiedNovember 20, 2026
Compare MEDIVIS with other vendors
Founded
2016
Headquarters
New York, New York
Categories
radiology-and-imaging-ai
Indexed Products
Medivis Studio, Medivis Spine Navigation, Medivis Cranial Navigation, Medivis Body Navigation (clearance pending), Workstation Pro, Maia (research)
Buyer Segments
Hospital neurosurgery departments, Spine surgery programs and ambulatory surgery centers, Academic medical centers, Medical education programs
Assessment

Capability Axes

An AI Health Index grade measures what a buyer can verify from public sources on the date shown. It is not a rating of how good the product is. A vendor can build an excellent system and grade low on an axis because it publishes nothing an outsider can check. How grades read

AI Capability
CC on AI CentralityArtificial intelligence is a feature layer on a product whose value stands without it. A learned component is a model whose parameters are estimated from data: machine learning, or a statistical model fitted to population data, including one then updated on the patient measurements. A fixed formula, a rule set, or a feedback controller that adjusts a dosing factor from the patient own readings is not one, however patient specific its output and whatever the vendor calls the adjustment.
Vendor Published

Automatic segmentation and fusion speed up planning on a platform whose core is visualization and navigation. Studio segments the spine, brain tumors and full body CT from a few seed points and registers MRI onto CT automatically, and Maia, a research program, lets vision language models call Medivis tools.

What surgeons buy, though, is 3D rendering, augmented reality display and instrument tracking, which work without any model, and none of the three FDA clearances describes a machine learning function. Studio and the navigation systems do what they were cleared to do without the models, with segmentation done by hand.

BB on Autonomy and Oversight ModelThe oversight structure is described and one part is missing, commonly the threshold at which the system stops or what happens after it is wrong.
Regulatory Filing

The surgeon operates the system, and the labeling sets its limits. Spine Navigation (K233725) is cleared as an aid for locating anatomy, and its labeling warns that the headset's virtual display should not be relied on alone for absolute position. Cranial Navigation pairs the headset with a touchscreen workstation as the primary input. In Studio, routine structures segment automatically while surgeons carve out tumors and vessels interactively.

In the research programs, Maia's model proposes and verified tools act, and robotics runs autonomy skill by skill with a surgeon supervising. How an automatic segmentation is reviewed before it guides a plan is not described.

CC on Model and Technology TransparencyThe architecture is described in general terms with nothing identified. Proprietary is asserted rather than explained.
Vendor Published

Studio segments routine structures from a few seed points and fuses MRI onto CT automatically, and Maia passes imaging, algorithms and live case state to whatever vision language model is in use, local or cloud, then calls Medivis's own verified tools. No model type, training data, version or update practice is given for the segmentation, and no model is named for Maia. The cleared devices' summaries describe registration, rendering and tracking, so the clearance record does not say what produces a segmentation either.

CC on Model Supply Chain DisclosureThe architecture is described and no model provider is named. Naming a hosting provider alone does not lift a record out of this band. Record the host in the note, because it matters for residency and breach scope, and grade on the model layer, which is the question this axis is named for.
Vendor Published

Studio runs on Microsoft Azure when it is not on site, and the segmentation tools are presented as Medivis's own, though the company does not say so outright. Maia is built to work with any vision language model, local or cloud, and names none, so which outside providers receive imaging and case state, and on what retention and training terms, is not stated. The website privacy policy names Google Analytics and RingCentral, and no subprocessor list for the products appears.

BB on Clinical and Operational EvidenceNamed deployments with dated outcome figures and enough method to test them, or published research short of independent validation.
Vendor Published

The research page lists 12 publications from teams using the technology, including single center case series in cranial and cerebrovascular surgery, augmented reality guided ventricular drain placement in intensive care, rhizotomy, ovarian cancer planning at Memorial Sloan Kettering, a positional accuracy study against an FDA standard and a Veterans Affairs survey.

Cranial Navigation's clearance reports bench error within 2 mm and 2 degrees, and the cranial page claims five times better catheter accuracy than free hand placement, citing the drain placement literature rather than a trial of its own. No study measures how accurate the automatic segmentation is or whether it changes outcomes, and the current site names no hospital as a customer.

CC on AI Safety and PHI StewardshipGeneral assurances of privacy and security that do not answer the questions artificial intelligence raises: what is retained, what reaches a model, and what happens to it there.
Vendor Published

Deployment choices shape where imaging goes, and little else is said. Studio can run on Microsoft Azure or on site, and the Maia research page notes that local models suit data that must not leave the building, which gives a hospital a way to keep scans inside its walls. Nothing states whether patient imaging trains or tunes the segmentation models, how long images or plans are kept in an Azure deployment, or which model providers Maia sends imaging to when it uses cloud models. Maia chat logs on the website are kept for 12 months, and guardrails and how a safety event is handled are not described.

Regulatory and Compliance
CC on HIPAA and BAA PostureA status is stated but not supported: compliance is claimed, or business associate status stated, without the underlying document, including where the only privacy notice published covers the website rather than the service that handles patients. The mirror case grades here as well: a substantive privacy document that reaches the service, with no statement of business associate status anywhere and no scope position taken. One of the two elements the band above asks for is present, which is a grade below it and not a grade at the floor.
Vendor Published

A business associate agreement is listed, and the only privacy document covers the website. The Studio page includes HIPAA and BAA among its enterprise features beside single sign on, without the agreement or any statement of business associate status.

The privacy policy, last updated February 2023, defines the service as the website alone, mentions HIPAA only to exclude covered health information from California privacy law, and asks visitors not to share patient information with the Maia chat guide. How Studio, the navigation systems or an Azure deployment handle the patient imaging they load is not addressed.

DD on Security Certifications and Trust CenterControls are asserted with nothing independent behind them, or nothing is published. Read the note before concluding anything: this is the grade most often corrected on a second pass, because assurance material frequently sits on a parent domain or inside an old announcement rather than on the product pages.
Vendor Published

Single sign on is listed, and no certification or outside check appears. The Studio page lists SAML and OIDC sign on and HIPAA and BAA among enterprise features, and nothing else on security: no SOC 2, ISO 27001 or HITRUST, no penetration testing, no trust center and no description of encryption or access control. The FDA clearances test registration accuracy, latency and display, not how Medivis runs security. For software that loads patient imaging into a cloud service and streams it to headsets in the operating room, a buyer has no independent assurance to check.

DD on FDA and Regulatory StatusThe positioning is unclear or unsupported: clinical claims with no regulatory position stated, or language implying a status the databases do not carry. An open enforcement action that names the product also grades here, whatever the positioning: an FDA warning letter with no close out letter posted, or a consent decree in force. The note says which of the two applies, and the grade returns to the positioning bands when FDA closes the action.
Regulatory Filing

The Studio page calls Medivis Studio an FDA cleared, CE marked platform for imaging, AI segmentation, 3D reconstruction and surgical planning, and the company timeline says the imaging, segmentation and planning platform was cleared in 2019. That clearance, K190764 (May 2019, product code LLZ, as SurgicalAR), covers image display, MPR, 3D volume rendering and measurements, and its summary describes no segmentation or machine learning.

K233725 (July 2024, OLO) cleared Spine Navigation, which the spine page calls AI powered, and K231897 (October 2025, HAW) cleared Cranial Navigation as NeuroAlign; neither summary describes a learned function. Body, fluoroscopic and cervical navigation are marked pending and not for sale, and Maia is labeled research.

DD on AI Governance and Bias DisclosureNothing published on how model behavior is governed or tested. Multilingual operation with no subgroup performance sits here when the vendor markets recognition quality as a strength, because a caller the system failed to understand leaves no complaint and no record.
Vendor Published

How the segmentation models are tested or governed goes unsaid. Automatic segmentation of the spine, brain tumors and full body CT feeds surgical plans, but no accuracy results, testing across scanners, contrast protocols, patient ages or unusual anatomy such as prior fusion hardware or deformity, review of model changes or outside audit appears. The FDA summaries for all three cleared devices report bench testing only and describe no machine learning. Maia is framed as a joint development program in which verified tools do the work the model proposes, which describes a design choice rather than an evaluation.

BB on AI Liability and RecourseA published falsifiable commitment, or a real correction route the affected person can exercise against the vendor. A published error rate with its method and denominator grades here. A statutory right that runs to the covered entity rather than to the vendor does not reach this band on its own: every other route here asks something of the vendor, and being located in a particular jurisdiction is not conduct.
Regulatory Filing

Cranial Navigation's clearance (K231897) reports system error within 2 mm and 2 degrees in bench testing to ASTM F2554, and Spine Navigation's labeling tells surgeons not to rely on the headset alone for absolute position. Those figures cover tracking, not the accuracy of an automatic segmentation used in a plan, which carries no measured error.

The public terms, updated February 2023, cover only the website, provide it as is and cap liability at the amount paid or $100, and no product warranty, remedy or correction path for a wrong segmentation is offered.

Integration and Deployment
CC on EHR and Interoperability DepthIntegration is claimed through standards or a middleware layer with no system named and nothing to verify.
Vendor Published

Studio lists full PACS integration, DICOMweb, a modality worklist and single sign on through SAML or OIDC identity providers, and Cranial Navigation imports images from PACS or local files. No PACS vendor, EHR, surgical scheduling system or marketplace listing is named. Plans, trajectories and screw plans are saved to the case inside Studio and opened again in the operating room, and nothing says whether plans or segmentations are written back to the imaging archive or the record. In the operating room, Studio connects to the HoloLens 2, Apple Vision Pro, a 3D monitor and third party navigation tools.

BB on Deployment Model and Data ResidencyOptions and residency are stated with isolation or the processing path left open.
Vendor Published

Studio runs on Microsoft Azure or on site, as Enterprise with server GPUs feeding thin clients or as a Workstation on a local GPU, and navigation runs on an imaging workstation in the operating room. Maia can use local models for data that must not leave the building. Which Azure region holds imaging, how one hospital's data is separated from another's and how long it stays are not stated, and the privacy policy says data may be processed wherever the company and its partners operate.

Commercial
CC on Commercial TransparencyNo price is published and the posture is discoverable: a buyer can establish how the product is sold and what drives the cost before contacting the vendor. Most of the index sits here.
Vendor Published

Studio deploys as Enterprise, on server GPUs with thin clients on Microsoft Azure or on site, or as a Workstation on a local GPU, and navigation adds an NVIDIA powered workstation, infrared tracking and a HoloLens 2 headset, with smart glasses and AR loupes now offered for spine. Buyers request a free trial of Studio or contact enterprise sales, and the spine product is sold to hospitals and ambulatory surgery centers. No price is given, and how software, hardware, headsets and support are charged, whether per seat, per site or per case, is not stated.

BB on Setting and Specialty CoverageCoverage is named with validation behind part of it.
Regulatory Filing

Spine Navigation (K233725) covers open and minimally invasive thoracolumbar spine procedures, including pedicle screws, with cervical spine and 2D to 3D fluoroscopic registration pending; Cranial Navigation (K231897) covers catheters, shunts and craniotomies in operating rooms, intensive care units and interventional suites. Body navigation, led by prostate biopsy, is under investigation.

Studio serves surgical planning across specialties and anatomy education through AnatomyX, and published case series cover neurosurgery, intensive care drain placement and gynecologic oncology planning. Results by setting, surgeon experience or patient anatomy are not reported, and the automatic segmentation has no validation by anatomy.

Alternatives

Alternatives to MEDIVIS

Other products that do the same kind of work as MEDIVIS, and what each one does.

MediView XR builds XR90, an augmented reality visualization and navigation system for image guided interventional procedures, and markets OmnifyXR for interventional suites.

Treatment Pad is an iPad and iPhone app that uses computer vision to produce before and after photographs for cosmetic physicians and aesthetic clinics.

MEDICAL IP builds AI segmentation and quantification software for medical imaging.

Seoul based MRI AI company with US operations. SwiftMR is a deep learning enhancement tool that reduces noise and sharpens MRI images after acquisition, so sites can scan faster or get better images from the same scan.

Deep learning software that enhances and accelerates image acquisition on existing scanners rather than detecting disease.

Belgian company, founded in 2011 in Leuven and now part of GE HealthCare, whose icobrain software measures the brain on MRI and CT.

Commercial

Pricing

Vendor-published figures are labeled as such. Figures labeled “Estimated” are derived from third-party sources and have not been confirmed by the vendor.

Entry Price Pricing Basis BAA Tier Implementation Source
Not published.
Medivis sells through enterprise sales and offers a free trial request for Studio. Studio deploys as Enterprise, on server GPUs with thin clients on Microsoft Azure or on site, or as a Workstation on a local GPU, and navigation adds Workstation Pro hardware, a HoloLens 2 headset and optical tracking. How software, hardware and support are charged is not stated. Not published. The Studio page lists HIPAA and BAA among its enterprise features. None stated. Vendor Published

No figures are given for software, hardware or support.