Session
A case loaded from PACS by name or MRN. O-Arm and C-Arm scans, USB, CD and the library land in the same session — saved once, recovered whole.
Medivis’ intelligence is embedded in every step of the surgeon’s workflow. Medivis Studio turns imaging into a plan. Medivis Navigation carries that plan into the operating room as real-time guidance. Frontier closes the loop, with systems that learn from every case.
The FDA-cleared, CE-marked imaging platform for everything pre-operative — every case below begins as data, becomes a plan, and integrates with all the devices in the room.
Everything a case is made of — CT, MRI, Ultrasound, Fluoroscopy, PET — the session it lives in, the volumes inside it, and the fusion of one modality with another.
Data
A case loaded from PACS by name or MRN. O-Arm and C-Arm scans, USB, CD and the library land in the same session — saved once, recovered whole.
Everything a case is made of — CT, MRI, Ultrasound, Fluoroscopy, PET — the session it lives in, the volumes inside it, and the fusion of one modality with another.
A case loaded from PACS by name or MRN. O-Arm and C-Arm scans, USB, CD and the library land in the same session — saved once, recovered whole.
The 3D dataset cropped, cut and lit in real time — bone, soft-tissue, brain and lung presets, scout planes, and photorealistic rendering when the case calls for it.
MRI fused onto CT with automatic registration. Swap main and fused, tune opacity and tint, unfuse or recompute at any point.
Imaging becomes anatomy, and anatomy becomes a plan — segmented, annotated and templated on the patient’s own scan.
Plan
Automatic segmentation across spine, brain-tumour and full-body CT targets; a few seed points generate the 3D mesh, with STL import and export.
Imaging becomes anatomy, and anatomy becomes a plan — segmented, annotated and templated on the patient’s own scan.
Automatic segmentation across spine, brain-tumour and full-body CT targets; a few seed points generate the 3D mesh, with STL import and export.
Target and entry trajectories dropped on the patient’s own scan — sync mode drives the localizer in one step, with millimetre measurements on any slice.
Screw planning, level by level: pick levels on the spine schematic, drag diameter, length, body, head and tip — and keep editing live in the OR.
The hardware around the case — sensors, headsets, screens, feeds and probes — connected to the same session so the room can see the plan.
Devices
Select from a range of head-mounted displays including AR loupes (surgeon-specific IPD/prescription) and immersive headsets (e.g. Apple Vision Pro).
The hardware around the case — sensors, headsets, screens, feeds and probes — connected to the same session so the room can see the plan.
Select from a range of head-mounted displays including AR loupes (surgeon-specific IPD/prescription) and immersive headsets (e.g. Apple Vision Pro).
Glasses-free, eye-tracked 3D on the spatial monitor — engage the 3D view and point in depth. No headset required.
Cameras, endoscopes and microscopes routed into any viewport — or in stereo, straight into the headset.
Wireless, tracked ultrasound in the AR environment — B-mode, colour and power Doppler, situated at the probe tip or on a virtual monitor, with depth, gain and TGC in reach.
The optical tracker that sees the tools and the patient — live connection status, a tracker view with positioning lasers, and hybrid tracking with the headset.
Tracked guidance extended to robotic positioning. Talk to us about integration.
Two ways to deploy
The instruments the surgeon already uses, made trackable: add them to a toolkit, fix a reference to the patient, calibrate the tip — then navigate.
Tools
Adapters make the instruments the surgeon already uses trackable, from probes and taps to a catheter clip for EVD, and the reference array anchors the scan to the patient with a warning the moment tracking is lost. Reusable and consumable alike, all from one toolkit library.
The instruments the surgeon already uses, made trackable: add them to a toolkit, fix a reference to the patient, calibrate the tip — then navigate.
Adapters make the instruments the surgeon already uses trackable, from probes and taps to a catheter clip for EVD, and the reference array anchors the scan to the patient with a warning the moment tracking is lost. Reusable and consumable alike, all from one toolkit library.
The moment the scan is locked to the patient on the table — six ways to do it, one rule: verify before you guide.
Register
Painting the surface: forehead and the bridge of the nose traced, checked against the isosurface, then registered.
The moment the scan is locked to the patient on the table — six ways to do it, one rule: verify before you guide.
Painting the surface: forehead and the bridge of the nose traced, checked against the isosurface, then registered.
The face captured as a point cloud and verified against the scan — automatic in seconds, no fiducial hunt, and the primary method for ventriculostomy. Surface capture beyond the face is in development; talk to us about your indication.
Four to eight virtual landmarks, then the same points touched on the patient — the error is computed before anything snaps.
Registration from the scan itself — fiducials found automatically and aligned to the tracked array.
Fluoroscopy registered to the volume. Talk to us about your program.
A slow sweep of the tracked probe over the anatomy aligns MRI and CT to live ultrasound.
The plan on the patient, where the surgeon is looking — cranial and spine cleared, body under investigation.
Where our research is heading — an agent that is really a harness, and robots that join a system which already knows where everything is.
Frontier: co-development programs, not cleared products.
Frontier
Maia is a contextual harness for state-of-the-art vision-language models, cloud or local: the right imaging, algorithms and live case state in front of any model, then our validated tools called on request — load, segment, plan, register. New models integrate the week they ship.
Where our research is heading — an agent that is really a harness, and robots that join a system which already knows where everything is.
Frontier: co-development programs, not cleared products.
Maia is a contextual harness for state-of-the-art vision-language models, cloud or local: the right imaging, algorithms and live case state in front of any model, then our validated tools called on request — load, segment, plan, register. New models integrate the week they ship.
Robotic positioning inside a loop that already knows where the tools, the patient and the plan are — Level 2 task autonomy, where the surgeon approves and the robot drives the screw or needle to the plan; Level 3 supervised sequences like drill → tap → insert. A surgeon supervising, always. Research use only.
Founded by physicians and engineers in New York. Everything we make lives in black — the umbra, where the geometry is absolute and ambiguity disappears.
Send us a message — a member of the Medivis team replies within one business day.