Computational Surgery offers a software platform that fuses pre‑operative CT/MRI data with live endoscopic video to deliver real‑time 3D reconstruction, sub‑millimeter spatial registration, and augmented‑reality overlays for minimally invasive procedures. The system generates deterministic tool trajectories and integrates via API with existing robotic and endoscopic hardware, enabling hospitals and device OEMs to improve navigation accuracy and reduce operative time.
Funding
Funding not disclosed
Founders
Product
Problem
Surgeons performing robotic or endoscopic procedures often lack precise, intra‑operative guidance that integrates pre‑operative imaging with real‑time anatomy, leading to longer operative times and increased risk of mis‑placement during laparoscopic, percutaneous, and needle‑insertion surgeries.
Solution
Computational Surgery delivers a mathematics‑driven software platform that fuses pre‑operative CT/MRI data with live endoscopic video to provide automated surgical planning, navigation, and augmented‑reality (AR) visualization. The system uses computer‑vision pipelines to reconstruct 3D anatomy, performs spatial registration to align patient models with the operative field, and projects AR overlays directly onto the surgeon’s display. Real‑time trajectory calculations guide robotic arms and flexible needles, reducing reliance on manual alignment and improving accuracy. The platform is offered through licensing agreements to hospitals and medical‑device manufacturers, enabling integration with existing robotic and endoscopic hardware. Continuous updates incorporate advanced imaging modalities such as hyperspectral data to enhance tissue discrimination during procedures.
Target Audience
Primary customers are tertiary hospitals and surgical centers that perform minimally invasive procedures, as well as OEMs of robotic surgery systems and endoscopic equipment seeking to embed advanced navigation and AR capabilities.
Features
- Deterministic trajectory‑planning algorithms that generate optimal entry points and paths for laparoscopic tools, percutaneous needles, and robotic manipulators.
- Real‑time 3D reconstruction from monocular endoscopic video using deep‑learning‑based depth estimation and SLAM techniques.
- Rigid and deformable spatial registration engine that aligns intra‑operative video with pre‑operative CT/MRI datasets within sub‑millimeter accuracy.
- Augmented‑reality overlay module that renders anatomical models, planned trajectories, and safety margins directly onto the endoscopic feed with low latency (<30 ms).
- API and SDK for seamless integration with commercial robotic platforms, endoscope manufacturers, and OR information systems (FHIR‑compatible).
- Support for hyperspectral imaging streams to provide intra‑operative tissue classification and perfusion mapping.
- Compliance‑ready software architecture meeting IEC 62304 and ISO 14971 standards, with configurable data‑encryption and audit‑logging for clinical use.