FACT-Net
🧠 Multimodal Physiological Signal Acquisition System
The multimodal physiological signal acquisition system developed in this study integrates cutting-edge hardware components, ensuring precise, reliable, and efficient data acquisition for ABP reconstruction tasks in diverse healthcare settings.
🔧 System Components
- Hardware Modules:
- (a) 🛠 Sensor Unit: Collects physiological signals (e.g., ECG, PPG).
- (b) 📊 Signal Processing Circuitry: Processes raw signals into usable data.
- (c) 🖥 Microcontroller Unit (MCU): Coordinates the system’s operations.
- (d) 💾 Data Processing Module: Handles complex computations for signal analysis.
- (e) 🔋 Charging Unit: Powers the system with high-efficiency charging.
- System Design:
- (f) 🖼 Simulation Diagram: Illustrates the functional design of the system.
- (g) 🔲 Front View: Showcases the compact and ergonomic design.
- (h) 🔲 Rear View: Highlights modularity for easy upgrades.
- (i) 🌟 PPG Sensor: Provides high-accuracy signal acquisition.
- (j) 🔋 3D-Printed Casing: Integrated with a lithium battery for enhanced portability and durability.
🔍 FACT-Net Architecture
FACT-Net leverages a two-stage hybrid architecture combining CNNs and Transformers to achieve high-fidelity ABP waveform reconstruction. This architecture enables robust cross-modal feature extraction and integration for optimal performance.
🚀 Stage I: Parallel Cross-Hybrid Architecture
- Objective: Efficiently extract multimodal features and provide constraint information for accurate ABP reconstruction.
- Components:
- 🔲 Multi-Scale CNN Blocks: Capture hierarchical features across varying temporal resolutions.
- 🔀 Mix-T Blocks: Facilitate efficient multimodal feature fusion for improved integration.
🔨 Stage II: Serial Hybrid CNN-Transformer Structure
- Objective: Refine feature representations and ensure high-fidelity ABP waveform reconstruction.
- Components:
- 🌐 Hybrid CNN Layers: Enhance spatial and temporal feature extraction.
- ⚡ Transformer Modules: Improve global context representation for superior signal reconstruction accuracy.
🌐 Cross-Platform Multi-Patient IoT Framework (CPMP-IoT)
The CPMP-IoT framework extends FACT-Net’s capabilities to real-world healthcare applications, offering a scalable and reliable solution for multi-patient health management.
🔑 Framework Features
- 📱 Individual Monitoring APP
- Offline Inference: Real-time, personalized health monitoring without cloud reliance.
- 💻 Host Computer Integration
- Devices connect via LAN, enabling seamless access to health data and reports through a secure web interface.
- 🏥 Ward-Level Multi-Patient Monitoring
- Simultaneous Monitoring: Enables the management of multiple patients in healthcare wards, ensuring scalability and efficient resource use.
📑 Appendices
📐 Appendix I: Circuit Schematic
The circuit schematic provides a comprehensive illustration of the hardware design, detailing the interconnections between key components in the physiological signal acquisition system.
Download Circuit Schematic PDF
🖥 Appendix II: PCB Design
The PCB design outlines the printed circuit board layout, ensuring optimal integration and functionality of the system components.
Download PCB Design PDF
💾 Download Buttons
You can easily download the circuit schematic and PCB design from the following links:
