MedApp

MedApp S.A.

Improving Healthcare Through Imaging Innovation

Computer Vision

Three areas of business

From a diagnostic image to a decision at the bedside.

MedApp builds software that turns a flat imaging study into a spatial model, guides the clinician's hand during a procedure, and teaches machines to understand what they see. Three products, one technology — visualising and analysing medical data in three dimensions.

Surgical team in mixed-reality headsets with a holographic heart above the operating field
FDA · CE
01CarnaLife Holo

The patient as a hologram, not a flat screen

A standard CT or MR study in DICOM format is reconstructed into a volumetric model the clinician views in a mixed-reality headset or on a monitor. It can be rotated, scaled, cut with planes and measured — at true scale, in the space around them.

  • 3D reconstruction straight from DICOM data, with no extra hardware in the imaging suite
  • Works in a mixed-reality headset, on a monitor and in a shared multi-user session
  • Clinical, teaching and patient-communication use

FDA

cleared for the United States market

CE

medical device in the European Union

3D

from any CT or MR study

Explore CarnaLife Holo
Needle trajectory planning in CarnaLife Holo MedNav — a 3D chest model beside a sagittal CT slice
Clinical research
02CarnaLife Holo MedNav

Navigation that guides the needle in real time

The CT or MR model is superimposed on the patient inside the headset, while the needle and the ultrasound probe are tracked with no markers attached to either. The clinician plans the trajectory before the puncture and sees how far the hand has to be corrected — confirmed by two peer-reviewed papers written with AGH University.

  • Trajectory planning, verified in the 3D view and on 2D cross-sections
  • Fusion of live ultrasound with CT data — two modalities in one scene
  • Instrument tracking from the headset cameras alone — no markers, no external tracker

87%

target hit rate, against 33% with ultrasound alone

-53%

shorter preparation and procedure time

3.41 mm

upper bound of the US–CT fusion error (95% CI)

See how MedNav works
A robotic arm whose vision system analyses a procedural scene in three dimensions
Physical AI
03Computer Vision Services

Vision systems that do not only see - they understand

Our engineers bring the latest generation of vision models — including the NVIDIA Cosmos and Isaac GR00T families — into medical devices and robotic systems. Instead of a single object detection you get a layer that understands the scene and can propose a safe course of action.

  • 3D perception, vision-language and Vision-Language-Action models tuned to your machine
  • Training data generated in Isaac Sim wherever real data is scarce
  • On-device deployment — TensorRT, ONNX, NVIDIA Jetson

VLA

vision and language turned into action

Simulation

training in a virtual world before deployment

Edge

inference on the device, without the cloud

Explore our Computer Vision services

Looking for a technology partner in medical imaging or computer vision?

Talk to us

Contact

Address

Aleja Juliusza Słowackiego 6/10

30-037 Kraków

Poland

Follow us

NIP: 6631868308

REGON: 260637552

KRS: 0000443282

MedApp Resources

Login
2026 © Medapp S.A.Terms & ConditionsPrivacy Policy
EU banner