Healthcare / Medical Devices
Medical wearable: Remote patient monitoring device for cardiac care
End-to-end development of a wearable ECG monitor engineered to be FDA 510(k) submission-ready, designed for 14-day battery life and on-device arrhythmia detection for remote cardiac monitoring.
What we did: We designed the ECG analog front end and nRF52840 electronics, the rigid-flex PCB, IEC 62304 Class B firmware, the companion apps and cloud platform, and prepared the design-control and 510(k) documentation package for the client to file. FDA status: 510(k) submission-ready (not cleared).
US-based digital health startup18 months
Published Last reviewed:
Illustrative imageThe challenge
A US-based digital health startup needed to develop a medical-grade wearable ECG monitor that could continuously monitor cardiac patients for up to 14 days, detect arrhythmias in real-time, and be ready for an FDA 510(k) submission for the US market.
Regulatory compliance
The device needed to meet stringent FDA requirements including IEC 62304 software lifecycle, ISO 13485 quality management, and IEC 60601-1 electrical safety standards.
Battery life requirements
Continuous ECG monitoring with BLE transmission required 14 days of operation on a single charge while maintaining a form factor acceptable for patient comfort.
Clinical-grade accuracy
The device needed to match or exceed Holter monitor accuracy for arrhythmia detection, with sensitivity and specificity targets above 95%.
Cost constraints
To enable wide deployment, the BOM cost needed to be below $35 at volumes of 100,000 units while meeting all medical-grade requirements.
Impact figures describe the situation before the project, as the client reported it to us; we did not measure them.
Our solution
We developed a complete medical-grade wearable ECG monitoring system including custom hardware, embedded firmware, mobile app, cloud platform, and the design-control documentation needed for an FDA 510(k) submission.
System architecture
End-to-end medical IoT architecture designed for regulatory compliance, reliability, and scalability.
Wearable device
- Medical-grade analog front-end (AFE)
- Ultra-low-power MCU with DSP
- On-device arrhythmia detection
- Bluetooth Low Energy 5.0 connectivity
- Skin-friendly medical adhesive
Mobile gateway
- iOS and Android companion apps
- Real-time ECG visualization
- Offline data buffering
- Patient symptom logging
- Automatic cloud sync
Clinical platform
- Cloud infrastructure designed to support HIPAA Security Rule safeguards
- AI-powered ECG analysis
- Clinician dashboard
- Alert management system
- EHR integration (HL7 FHIR)
Custom hardware design
- MCU
- nRF52840 (BLE 5.0, DSP capable)
- AFE
- ADS1292R (24-bit, medical-grade)
- Electrodes
- Ag/AgCl hydrogel snap electrodes
- Battery
- 180mAh Li-Poly (rechargeable)
- Dimensions
- 55mm x 25mm x 8mm
- Weight
- 18g (with adhesive)
- Protection
- IP67 (design target); designed to IEC 60601-1
Firmware architecture (IEC 62304 Class B)
- FreeRTOS-based architecture with watchdog supervision
- Real-time 250 SPS ECG sampling with DMA
- On-device R-peak detection and HRV calculation
- QRS complex classification (normal, PVC, PAC)
- BLE GATT service for ECG streaming
- Encrypted firmware OTA updates
- Comprehensive fault detection and logging
- Battery management with 1% accuracy
On-device AI for arrhythmia detection
We implemented a lightweight neural network on the device for real-time arrhythmia screening, with cloud-based analysis for detailed diagnosis.
Beat classification
1D CNN (Convolutional Neural Network)
99.2% VEB sensitivity (ANSI/AAMI EC57, MIT-BIH)
Design targetAtrial fibrillation detection
LSTM with attention mechanism
97.8% sensitivity, 98.5% specificity
Design target< 2 second detection delay
Signal quality assessment
Random Forest classifier
96% noise/artifact detection
Design targetImplementation timeline
Phase 1: Requirements & architecture
8 weeks- User needs analysis and clinical workflow study
- Regulatory strategy and predicate device analysis
- System architecture and risk analysis
- Design input documentation
Phase 2: Hardware development
16 weeks- Analog front-end design and simulation
- PCB layout (4-layer, flex-rigid)
- 3 prototype iterations
- IEC 60601-1 design verification
Phase 3: Firmware development
20 weeks- IEC 62304 software architecture
- Real-time ECG acquisition and processing
- BLE stack implementation
- On-device ML model integration
- Unit and integration testing (98% coverage)
Phase 4: Software development
16 weeks- iOS and Android app development
- Cloud platform development
- Clinician dashboard
- Cybersecurity implementation
Phase 5: Verification & validation
12 weeks- Bench testing per IEC 60601-2-47
- Clinical validation study support
- Usability testing
- Biocompatibility test support (materials selected to ISO 10993)
Phase 6: Submission package preparation
6 weeks- 510(k) submission compilation
- Design verification & validation closeout
- Submission package finalized for client filing
- CE technical file preparation
Results & impact
The device was engineered to be FDA 510(k) submission-ready; the submission package was handed to the client for filing. It is not FDA-cleared. Performance figures below are design targets.
How to read the labels: Design target = an engineering target, not a field measurement; Client-reported baseline = the client's own figure for the situation before the project; we did not measure it. Figures without a label are specifications or project scope.
VEB detection sensitivity
Design target99.2%
Ventricular ectopic beats, ANSI/AAMI EC57 on the MIT-BIH database
Battery life
Design target14 days
Continuous ECG with BLE streaming
Development timeline
18 months
Requirements to submission-ready package
Estimated BOM cost
Design target$32
Estimated BOM at 100,000 units, below the $35 target
Technologies used
- nRF52840
- ADS1292R
- FreeRTOS
- TensorFlow Lite
- BLE 5.0
- React Native
- AWS HealthLake
- HL7 FHIR
- Python
- PostgreSQL
Related case studies
Healthcare / Medical Devices
Smart urinometer: Automated ICU urine output monitoring system with real-time analytics
Public Safety / Government