• 388 S Abbott Ave., Milpitas, CA 95035
  • 1.510.396.6880
  • eho@acuemax.com
Success Story | Medical AI & Surgical Imaging

Real-Time Polyp Detection in Surgical Endoscopy

A medical device OEM migrated to a medical-grade Edge AI platform to achieve sub-10ms latency and pass IEC/EN 60601-1 (4th Edition) safety requirements—enabling real-time overlays in surgery.

<10ms
Inference latency
60 FPS
Jitter-free overlay
1st Try
Passed safety tests
-3 mo
Time-to-market
Enabling Real-Time Polyp Detection in Surgical Endoscopy

Executive Summary

A leading manufacturer of surgical endoscopy systems approached Acuemax because their polyp detection AI was too slow on standard medical PCs. By migrating to a Lambortech-optimized Edge AI platform featuring NVIDIA RTX graphics and IEC 60601-1 certification, the client achieved sub-10ms latency—enabling real-time overlays without compromising patient safety.

The Challenge

  • Latency: Streaming 4K video for remote processing introduced ~200ms delay, risking overlay misalignment.
  • Thermal management: OR hygiene required sealed design, but GPU workloads generate significant heat.
  • Certification: Must meet IEC/EN 60601-1 (4th Edition) medical safety standards.

The Solution

  • Local inference power: NVIDIA RTX A4500 MXM integrated into WMP-24S all-in-one panel.
  • Custom thermal design: Rear-chassis heat dissipation supports sealed, MRSA-resistant housing.
  • Safety integration: 4KV isolation on LAN/COM/USB to protect patient and connected devices.

The Result

  • <10ms latency for a smooth, 60 FPS surgical workflow.
  • Passed IEC 60601-1 safety tests on the first try; accelerated launch by ~3 months.
  • Sealed antibacterial housing supports wipe-down with medical-grade disinfectants.
“Acuemax didn't just sell us a computer; they solved the heat and latency problems that were holding our product back.” — Chief Engineer, Medical Device Partner