Synetra
Technology

Ultra-low-power neuromorphic compute, purpose-built for pre-hospital care.

Synetra's signal analysis runs on purpose-built neuromorphic hardware — not a generic embedded GPU — so inference stays at the patient at milliwatt scale, fully offline. Everything below is a concept description subject to development, validation and regulatory processes.

Concept stack

A layered concept — not a black box.

Sensors
ECG · SpO₂ · BP · PPG · Temp · ambient inputs
Edge compute
BrainChip Akida — quantized-CNN inference at milliwatt scale
Local storage
Encrypted patient timeline record
Connectivity
Cellular · Wi-Fi · NFC · optional satellite
Interfaces
Rugged touchscreen · quick-action controls
Integrations
Portable ultrasound · digital stethoscope

Architecture concept — subject to change through development, validation and regulatory processes.

Validation

Accuracy preserved when deployed on-chip.

Balanced accuracy · float vs on-chip (Akida simulator)
ECG arrhythmia
float 0.930 · on-chip 0.930
Float
On-chip
Heart sounds
float 0.865 · on-chip 0.865
Float
On-chip
ECG — myocardial infarction
float 0.790 · on-chip 0.810
Float
On-chip

Across three physiological signals validated on real clinical datasets, deploying the trained model onto the target neuromorphic chip (simulated) preserved accuracy — it does not cost you the model's performance to run it at milliwatt scale.

Architecture
1
Sensor front-ends

ECG analog front-end via SPI · PPG via I²C · MEMS microphone via I²S · capnography via UART

2
Host MCU / SBC

Acquire → window → preprocess / encode → feed Akida → fuse across modalities

3
Akida neuromorphic processor

Quantized-CNN inference on the target neuromorphic chip — milliwatt-scale, fully offline

Done

Software + Akida-simulator validation, three real modalities.

In progress

Sensor-integrated prototype on an off-the-shelf Akida board — no custom silicon or PCB on the critical path.

Next

Sub-milliwatt field device (Akida Pico); power validated on a cloud FPGA.

Built on BrainChip Akida neuromorphic processors.
Feature reference

Every feature, grouped by MARCH.

M

Massive Hemorrhage

Compensated hemorrhage detection
See deterioration before it becomes visible.

Synetra is designed to surface physiological changes associated with occult / compensated hemorrhage before heart rate or blood pressure visibly move — the window most commonly missed in current pre-hospital monitoring.

  • Pulse-wave morphology
  • Autonomic tone shifts
  • Multimodal fusion
  • Trend-first display
A

Airway

Real-time and offline operation
Built for care beyond perfect connectivity.

When a connection is available, the system can support real-time data transmission. When connectivity is interrupted, local operation and storage preserve the patient record until information can be transferred securely.

  • Rural EMS
  • Disaster response
  • Military medicine
  • Remote regions
  • Areas with inconsistent connectivity
R

Respiration

Modular diagnostic integration
One platform, designed to connect with more.

The system is being developed as a central hub for compatible diagnostic tools, reducing the need to operate multiple isolated devices and interfaces.

  • Portable ultrasound
  • Digital stethoscope
  • Future portable diagnostic modules
C

Circulation

Continuous multimodal monitoring
Individual measurements become a continuous story.

Synetra continuously monitors and displays key physiological measurements while placing greater emphasis on how they change over time.

  • 2-lead ECG
  • Blood pressure
  • Heart rate
  • Oxygen saturation
  • Pulse wave
Intelligent early warning
Designed to draw attention to meaningful change.

Synetra analyses incoming physiological information and highlights potentially important changes that may warrant renewed clinical attention.

  • Rapid blood-pressure decline
  • Worsening oxygen saturation
  • Rising heart-rate trend
  • Abnormal ECG changes
  • Increasing physiological instability
On-device neuromorphic compute
Milliwatt-scale intelligence, purpose-built.

Signal analysis runs on ultra-low-power, purpose-built neuromorphic hardware — not a generic embedded GPU — so inference stays at the patient at milliwatt scale, fully offline.

  • BrainChip Akida
  • Quantized-CNN inference
  • Milliwatt-scale power
  • Fully offline
H

Hypothermia / Head

NFC patient timeline transfer
Carry the patient's physiological story into the hospital.

Synetra is intended to create a structured, time-stamped patient timeline that can be transferred rapidly during handover.

  • Vital-sign trends
  • Timestamped clinical events
  • ECG recordings
  • Interventions performed
  • Clinical alerts
  • Future ultrasound images
Roadmap

What exists in the current concept — and what's planned.

Current concept
  • Core physiological monitoring
  • Trend visualisation
  • Intelligent early-warning logic
  • Offline data storage
  • Patient timeline
  • NFC-based handover concept
  • Unified interface
Planned
  • Portable ultrasound integration
  • Digital stethoscope integration
  • Remote data transmission
  • Additional diagnostic modules
  • Receiving-hospital interfaces
Long-term exploration
  • Thermal imaging
  • Respiratory audio
  • Broader interoperability
  • Expanded multimodal analysis
Positioning

What Synetra is — and what it isn't.

Is
  • A portable, ruggedised monitoring platform
  • A central hub for connected diagnostics
  • A structured patient timeline for handover
  • An early-warning support tool for clinicians
Is not
  • An autonomous diagnostic system
  • A replacement for clinician judgement
  • A cloud-dependent service
  • A single-purpose disease detector
Synetra is currently under development and is not presented on this website as a commercially available or regulator-approved medical device. Information is provided for research, evaluation, partnership, and future procurement discussions.
Next step

See Synetra in action.

Book a Demo