curated Signal edition
The state of surgical AI
A field map for surgeons: what Surgical AI can currently observe, predict, or support, and where the evidence still stops short of clinical use.
Sep 2Read Briefing
Operative Signal
The data network for surgical AI.
Hospitals connect the surgical data they already have. Companies and researchers request the cohorts they need. Operative Signal coordinates discovery, permissions, and multicentre validation without requiring hospitals to hand over their entire data archive.
Viewing as Neutral Overview
The Data Network for Surgical AI
Operative Signal provides the neutral data and validation infrastructure layer connecting clinical sites, researchers, and AI teams.
Step 01 · Connect
Surgical data is already everywhere.
Operating rooms record thousands of procedures daily onto departmental NAS, PACS archives, robotic consoles, and endoscopy towers. It is trapped in local silos. OS Node links this existing storage in minutes—indexing metadata locally while original video never leaves the hospital firewall.
On-Premises Data Boundary
How OS Node Keeps Video at its Source
Zero Cloud Video Upload
INSIDE HOSPITAL FIREWALL
Your Hospital Storage
NAS · PACS · Workstations · Towers
Original Video Never Moves
High-resolution recordings, patient records, and local file paths remain on your existing storage drives.
LOCAL DAEMON
OS Node
Software running on your server
Extracts Metadata Only
Indexes procedure type, duration, optics vendor, and local rights. Assigns a pseudonymous OS Video ID.
OPERATIVE SIGNAL
Discovery & Validation
Governed Network Coordination
Zero-Knowledge Feasibility
Evaluates cohort availability in the aggregate. No external party can browse or access your files without approval.
No Manual Video Uploads
Connect folders once. OS Node keeps catalog records synchronized automatically.
Keep Existing Storage
Works seamlessly with your current NAS, PACS, local drives, and endoscopy recording towers.
Institutional Sovereignty
Your hospital decides who may query, analyze, or validate models against your data.
Step 02 · Find
One request. Multiple institutions.
Define the exact procedure, case volume, camera hardware, and clinical outcomes your project needs. OS Exchange evaluates feasibility across active hospital nodes in the aggregate, so you see whether the right cases exist before any private data is disclosed.
Interactive Cohort Request Simulator
Surgical Data Request Builder
Select Procedure Scenario:
Request SpecificationRobotic Radical Prostatectomy
Target Case Count:500 cases
Minimum Multicentre Sites:≥4 institutions
Required Modalities:
Endoscopic Video Surgical Kinematics Pathology Reports 90-Day Outcomes
Camera Vendor Inclusions:
Intuitive da Vinci XiKarl Storz 3D
Intended Purpose:Multicentre External Validation
Aggregate Feasibility MatchFEASIBLE
Potential Matches640 cases
Site Readiness4 nodes ready
Identified Data Gaps
90-day urinary continence available on 82% of cohort
Governance Pre-Clearance
Pre-cleared DUA Matrix across 4 Academic Nodes
Step 03 · Approve
Hospitals decide what happens next.
Nothing moves without explicit institutional consent. OS Governance translates IRB approvals, data use agreements, and department permissions into a machine-enforceable Rights Ledger. You determine exactly who can access what, for what purpose, and for how long.
Interactive Rights Ledger
Clear Permissions. Absolute Institutional Control.
Deterministic Governance
Who owns & controls the data?
Site A Health System
Authorized Project:
Consortium Study: Generalization of Laparoscopic Workflow Models
What can this project do with the data?
External Validation
Remote Analysis
Model Retraining
Download Raw VideoFORBIDDEN
Ethics Approval:IRB-2026-088 (Exempt / Secondary Research)
Legal Agreement:Standard Multicentre DUA v2
Valid Until:January 31, 2028
Step 04 · Validate
Test the model where it matters.
An algorithm that achieves 94% accuracy at your home hospital might drop to 68% at another center simply due to a different camera vendor or lighting setup. OS Validate runs multi-center evaluation protocols across external hospitals to identify performance drops before clinical deployment.
Multicentre Generalization Analysis
Why External Validation Matters
4 Independent Hospital Nodes
The Generalization Gap: High Single-Center Accuracy Hides Multi-Site Failure
The model scored 94.2% at Site A where it was trained. When evaluated across independent hospitals without code modifications, performance plummeted to 68.4% at Site C due to differences in camera sensors, color grading, and surgical lighting.
Inspecting Validation Telemetry:
84 procedures evaluatedSite C (External Hospital)
Optical Tower & Hardware:
Olympus VISERA Elite II
Root Cause / Telemetry Finding:
Critical generalization failure: Optical color balance and smoke scattering degrade vessel segmentation.
Ecosystem
Built for everyone shaping surgical AI.
For Hospitals
Connect surgical video archives behind your firewall. Maintain complete control over data rights, ethics approvals, and research participation.
For AI Companies
Find multi-center surgical cohorts for model validation. Test generalization across camera manufacturers and hospital sites in weeks, not years.
For Researchers
Form multi-institutional consortia, share structured datasets under standardized DUAs, and run reproducible multicentre validation studies.
For Surgeons
Understand surgical AI without the hype. Access source-linked weekly briefings, structured 8-minute lessons, and peer surgical video discussions.
Clinical AI Intelligence
Latest from The Signal
curated Signal edition
Seeing the operation
Surgical video is becoming a structured source of workflow, anatomy, action, and skill information. This edition separates useful capabilities from benchmark performance alone.
Sep 2Read Briefing
curated Signal edition
Evidence beyond the benchmark
A strong internal score is not the same as clinical evidence. These sources focus attention on external validation, prospective evaluation, and comparisons that matter in practice.
Sep 2Read Briefing