SurgVis AR Healthcare Pvt. Ltd. — Lucknow, India
Projecting the anatomy surgeons can't see, directly onto the patient.
SurgNetra is a holographic surgical navigation platform: a patient's own CT/MRI is segmented into a 3D model and registered in real time onto the operative field through a HoloLens 2 headset — so a surgeon locates a perforator vessel or nerve before the first incision, not by probing for it.
Live demo
Scan to hologram, in one continuous pass
CT segmentation and 3D reconstruction in SurgNetra, followed by real-time AR tracking and overlay on the patient in the operating room.
The problem
Reconstructive microsurgery still finds vessels by feel.
In free-flap reconstruction, locating a perforator artery reliably determines how the flap survives. Today that means a handheld Doppler probe swept over the skin, or the surgeon's own memory of the CT scan reviewed that morning — both translate a 3D vascular map into a 2D image, then back into the surgeon's hands, freehand.
Every translation step costs time on the table and margin for error. Longer flap harvest time means longer anaesthesia, more blood loss, and a narrower safety margin if the anatomy doesn't match the textbook — which, in a meaningful share of patients, it doesn't.
Why AR, not another screen
A monitor still asks the surgeon to look away from the field and mentally re-project a flat image onto a 3D patient. SurgNetra removes that step: the reconstructed vessel or nerve is rendered as a hologram registered directly onto the anatomy, in the surgeon's own line of sight, tracked as the limb or flap moves.
The platform
SurgNetra: one pipeline, from scan to hologram
SurgNetra now covers the full path from a patient's DICOM imaging to an intraoperative AR overlay — including the segmentation and 3D-reconstruction tooling built for this purpose.
DICOM import
CT/MRI pulled directly from PACS, no manual export step required.
Segmentation & 3D reconstruction
SurgNetra's segmentation suite — axial/coronal/sagittal + 3D views, Otsu auto-threshold, region-growing from seed points — converts imaging into printable, AR-ready STL models of vessels, nerves and bone.
Registration
QR/fiducial marker tracking with SLAM anchors the 3D model to the live patient anatomy in the operating field.
Intraoperative overlay
The registered hologram is projected through HoloLens 2, tracked continuously as the surgical field moves.
Clinical evidence
Validated on the table, not just on the bench
SurgNetra's AR guidance has been through a published randomised controlled trial, an earlier RCT cohort, and a prospective pilot in a second surgical indication.
| Outcome measure | AR-assisted (n=20) | Conventional (n=20) | Significance |
|---|---|---|---|
| Vascular visualization score | 75 ± 6.4 | 49 ± 3.6 | p < 0.001 |
| Flap harvest time (min) | 126 ± 4.5 | 154 ± 5.9 | p < 0.001 |
| Ease of use / assistance required | Higher ease, less assistance | — | p < 0.001 |
| Flap survival | Comparable between groups | — | |
Earlier cohort — ICOMS 2025 (n=30)
Vascular display time 70 → 50 sec; flap harvest time 155 → 129 min with AR guidance (both p<0.001). An independent cohort from an earlier stage of the same programme.
Second indication — JPRAS pilot (n=10)
Prospective observational study in craniofacial / head & neck post-ablative free-flap reconstruction using HoloLens 2 with QR-fiducial registration. Accurate alignment reported in all 10 cases.
IP & regulatory
Protected, and on a defined path to approval
AR navigation patent
Complete Specification and a PCT application filed, covering the intraoperative projection and registration system.
CDSCO test license — Form MD-12
Filed via the ICMR MedTech Mitra initiative and currently under review with India's drug and device regulator.
SaMD classification
CDSCO Class B (moderate risk) software-as-a-medical-device.
Ethics committee (IEC) approval
Institutional ethics approval secured for the completed pilot RCT, ahead of multi-centre expansion.
ICMR MedTech Mitra
Programme support for the regulatory and clinical validation pathway.
Roadmap
Five phases, from projection to teaching
The build sequence, in order — each phase depends on the one before it.
AR projection setup
Core HoloLens 2 projection pipeline for a fixed 3D model, built with PreviewLabs.
DICOM-to-STL reconstruction
SurgNetra's segmentation suite for converting patient imaging into AR-ready 3D models.
Real-time tracking & drift control
Moving from locked-marker QR tracking to continuous marker-based tracking, so the hologram holds registration as a limb or flap moves.
AI-assisted segmentation & GPU pipeline
Automating the point-picking and segmentation steps that are performed manually today.
Anatomy teaching mode
Repurposing the same AR pipeline for anatomy education, using patient-derived and reference 3D models.
Team
A small team, working from the same model
Founders & leadership
Get in touch
Building toward UK & EU expansion, and always keen to talk clinical partnerships
For clinical & investment enquiries
Reach the founder directly, or the general SurgVis AR inbox.
aakashkohli22@gmail.com surgvisar@gmail.comCompany