VitalSense – Part 1: System Concept and Architecture

Project Overview and Objectives

As part of my project for the element14 Smart Home Healthcare Challenge, I am investigating how high-accuracy sensing, digital signal processing, and embedded computing can be combined to develop flexible systems for physiological monitoring.

The project is built around two evaluation platforms from Analog Devices:

• MAX30208EVSYS — a temperature-sensing evaluation system based on the MAX30208 digital temperature sensor, specified for ±0.1°C accuracy over the 30°C to 50°C range.
• MAXREFDES117 — an optical heart-rate and pulse-oximetry reference design based on the MAX30102 sensor, featuring red and infrared LEDs and example algorithms for heart-rate and SpO₂ estimation.

My first objective is to develop a PC-based diagnostic research station that acquires temperature and photoplethysmography (PPG) data through two independent USB-connected interfaces. A dedicated application written in C++/Qt or National Instruments LabVIEW will support signal acquisition, digital filtering, real-time waveform visualization, data logging, and algorithm evaluation.

The second stage explores three complementary embedded architectures: an nRF52840-based Bluetooth Low Energy telemedicine node, an ultra-low-power TI MSP430-based measurement device, and a Seeed Studio Wio Terminal providing a standalone graphical interface and Wi-Fi connectivity.

By comparing these platforms, I aim to evaluate their measurement performance, signal-processing capabilities, energy consumption, connectivity, and suitability for different physiological-monitoring scenarios. The project will establish an experimental foundation for future portable healthcare prototypes, with measurement accuracy and system performance assessed through documented validation procedures.

Technical Architecture, Implementation Plan, and Validation Strategy

1. Main Architecture - Diagnostic Station (PC + C++/Qt / LabVIEW)

1.1. element14 Kits and Modules

• Temperature measurement kit: MAX30208EVSYS# evaluation system, connected to a PC via USB using its supported evaluation interface and software or firmware configuration.

• Heart rate and SpO2 measurement kit: MAXREFDES117# optical reference design, interfaced with an STM32 microcontroller that acquires sensor data and streams it to the PC via USB using a Virtual COM Port (VCP) or USB-to-UART bridge.

1.2 PC Connection Method and Software Layer

The workstation is designed to receive data through two independent USB-connected interfaces:

• Channel 1 — Temperature: MAX30208EVSYS# → USB → evaluation firmware/software → temperature measurements.

• Channel 2 — PPG and SpO₂: MAXREFDES117# → STM32 → USB (VCP/UART) → optical samples and calculated parameters, where supported by the firmware.

Alternative A: C++ / Qt Application (Cross-Platform)

A multi-threaded application using QThread, QSerialPort, and a plotting library such as QCustomPlot will provide a cross-platform software architecture for Windows and Linux, including Raspberry Pi OS, subject to the availability of compatible drivers and libraries.

The proposed application will include:

• Independent data-acquisition workers for the temperature and PPG channels.
• Real-time visualization of temperature and optical waveforms.
• Digital filtering and signal-quality assessment.
• Heart-rate estimation and SpO₂ estimation, where the required algorithm and calibration are available.
• Timestamped data logging and CSV export.
• Connection-status indicators and measurement-error reporting.

Alternative B: National Instruments LabVIEW Application

LabVIEW provides an established graphical programming environment for measurement engineering, data acquisition, signal processing, and real-time visualization.

Communication layer — NI-VISA

• Configure independent communication sessions for the two data sources.

• Use VISA functions such as VISA Configure Serial Port, VISA Read, and VISA Write where the devices expose compatible serial interfaces.

• Implement appropriate framing, parsing, timeout handling, and communication-error detection.

Software architecture — Producer/Consumer design pattern

• Producer loops: Independent acquisition loops read incoming data from the two interfaces without blocking the user interface.

• Queues: Buffered data is transferred to processing and display loops.

• Consumer loops: Validate, timestamp, process, visualize, and store measurements.

Signal processing and digital filters

• Apply suitable digital filtering, such as a Butterworth band-pass filter, to reduce selected noise components in the PPG signal.

• Use peak detection and inter-beat interval analysis to estimate heart rate.

• Assess signal quality and identify motion-related artifacts.

• Estimate SpO₂ using a suitable algorithm based on red and infrared PPG signals and an appropriate calibration model.

User interface — Front Panel

• Waveform Chart and Waveform Graph elements for live temperature and PPG visualization.

• Numeric indicators, gauges, and status indicators for measurement results.

• Configurable sampling and display settings.

• Measurement-session management and communication diagnostics.

• Data export to TDMS (Technical Data Management Streaming) or CSV for subsequent research analysis.

1.3. Comparison of Similar Solutions in Portable Medical Diagnostics

2. Portable Autonomous Device Concepts (Architecture Triad)

The second part of VitalSense explores three complementary embedded architectures. Each concept prioritizes a different combination of energy efficiency, connectivity, processing capability, and user interaction.

• Concept 1: Personal Telemedicine Node with Cloud Connectivity (nRF52840)

  • Platform: Nordic Semiconductor nRF52840, initially using the nRF52840-DK development kit. The microcontroller integrates a 64 MHz Arm Cortex-M4F processor with FPU, 256 KB RAM, and 1 MB Flash.

  • Role: Acquire physiological sensor data, perform local signal processing, estimate heart rate, assess PPG signal quality, and support SpO₂ estimation using an appropriate algorithm.

  • Communication: Bluetooth Low Energy (Bluetooth 5.0) connects the device to a smartphone application.

    Potential standard GATT services include:

    • Heart Rate Service (HRS): For heart-rate measurements.

    • Health Thermometer Service (HTS): For temperature measurements, with the measurement type and units appropriately defined.

  • Planned safety and notification features

    The smartphone application could evaluate configurable thresholds, indicate abnormal readings, and notify the user or designated caregiver. If supported by the phone, operating system, permissions, and network availability, it could also initiate an SMS or other emergency notification.

• Concept 2: Ultra-Low-Power Field Rescue Device (TI MSP430)

  • Platform: Texas Instruments MSP430G2553, using the MSP-EXP430G2ET development platform. The MCU has a 16-bit architecture, 512 bytes of RAM, and 16 KB of Flash memory.

  • Design Philosophy: Prioritizing hardware simplicity, reliability, and ultra-low power consumption. Fully autonomous device intended for emergency field missions and long-term home healthcare.

  • Power management: The microcontroller will spend most of its idle time in a low-power state, with LPM3 considered for suitable clock and peripheral configurations.

  • Proposed operation

    1. The patient activates the device using a push-button.

    2. The controller wakes and initializes the required peripherals.

    3. Temperature and PPG data are acquired through compatible sensor interfaces.

    4. Heart rate and other supported parameters are estimated.

    5. The results are displayed locally, if a suitable display is implemented.

    6. The controller returns to a low-power state after the measurement session.

• Concept 3: Autonomous Diagnostic Monitor with Advanced HMI (Wio Terminal)

  • Platform: Seeed Studio Wio Terminal, based on the Microchip ATSAMD51P19 microcontroller: Arm Cortex-M4F at 120 MHz, 512 KB Flash, and 192 KB RAM. The platform includes a 2.4-inch, 320 × 240-pixel color TFT display, navigation buttons, and an RTL8720DN wireless module supporting Wi-Fi and Bluetooth Low Energy.

  • Design philosophy: Combine local visualization and interactive controls with portable embedded computing and wireless data communication.

  • Role: Acquire and display temperature and PPG data through compatible sensor interfaces. The processing capacity of the ATSAMD51P19 will be evaluated for local filtering, parameter estimation, and live waveform rendering.

  • HMI and connectivity

    • Standalone mode: Display current measurements, signal quality, connection status, and stored temperature and heart-rate trends using the integrated screen and navigation controls.

    • IoT mode: Send selected measurements to a local gateway or cloud service over Wi-Fi using MQTT, subject to implementation of the required network and application protocols.

Comparative Summary of the Three Portable Concepts

3. Project Schedule and Milestones

1. Phase I: Hardware Integration and Sensor Validation

   • Set up the MAX30208EVSYS# evaluation system and verify its USB communication path to the PC.

   • Interface the MAXREFDES117# reference design with an STM32 microcontroller.

   • Implement USB data transmission using a suitable VCP/UART interface.

   • Verify sensor readings, sample formats, timing, and communication stability.

2. Phase II: Diagnostic Station Application Development (C++/Qt or LabVIEW)

   • Configure the required USB and serial communication sessions.

   • Develop the graphical user interface for temperature and PPG visualization.

   • Implement data buffering, timestamps, error handling, and measurement logging.

   • Develop and evaluate digital filtering, heart-rate estimation, and SpO₂ estimation algorithms where supported.

3. Phase III: Portable Concept Implementation

   • Develop the nRF52840-based BLE monitoring prototype.

   • Evaluate the feasibility of the MSP430-based low-power implementation.

   • Develop the Wio Terminal graphical interface and wireless communication functions.

   • Document the differences in hardware integration, software complexity, and energy requirements.

4. Phase IV: Final Testing and Validation

   • Evaluate acquisition stability, missing samples, and communication errors.

   • Compare temperature readings against an appropriate reference instrument under documented test conditions.

   • Compare heart-rate and SpO₂ estimates with suitable reference equipment, recognizing the limitations of non-clinical validation.

   • Measure power consumption during standby, acquisition, processing, display operation, and wireless transmission.

   • Evaluate PPG signal quality and the effect of motion and sensor placement.

   • Document limitations, repeatability, and the conditions under which the results were obtained.

VitalSense is an experimental research and development project. The results will help determine the practical capabilities and limitations of each hardware platform and guide the development of future portable physiological-monitoring prototypes. VitalSense explores a modular approach to physiological monitoring, combining temperature sensing, optical PPG acquisition, digital signal processing, and embedded computing.

The PC-based diagnostic station will provide a flexible environment for acquiring and analyzing measurements, while the three portable concepts will enable a comparative study of wireless connectivity, low-power operation, local processing, and user interaction.