Skip to content

Artificial Intelligence for Congenital Heart Disease Screening

A study on large-scale CHD screening using AI-based heart sound analysis, transmitting data via STM32 and W5100 Ethernet for deep learning classification.

Artificial Intelligence for Congenital Heart Disease Screening

Components

Hardware components

Project description

πŸ“Œ Overview

This project is a study on large-scale screening for congenital heart disease (CHD) using an AI-based heart sound analysis system. Heart sound data collected with an electronic stethoscope are transmitted from an embedded system based on STM32 + Ethernet (W5100) to an upper-level computing platform, where diseases are automatically classified using deep learning models (RNN/CNN). The system aims to provide a low-cost, highly scalable diagnostic support solution that can be deployed even in regions with limited medical infrastructure.

πŸ“Œ Features

  • AI-Based Heart Sound Diagnosis
    • Automatic classification of major CHD types such as ASD, VSD, and PDA
    • Achieves up to approximately 90% accuracy using an RNN-based model
  • Large-Scale Screening Validation
    • Real-world field screening conducted on 518,258 individuals
  • Low-Cost Hardware Configuration
    • STM32 microcontroller combined with an Ethernet network
  • Field-Oriented Design
    • Portable, low-power operation with real-time data acquisition and transmission

πŸ“Œ System Architecture

πŸ“Œ System Architecture
πŸ“Œ System Architecture

The system adopts a clearly separated two-layer architecture, consisting of a lower embedded acquisition system (Lower Computer) and an upper processing and visualization system (Upper Computer).
This structure is designed to appropriately distribute responsibilitiesβ€”ensuring stable medical signal acquisition, reliable data transmission, and high-performance AI processingβ€”according to each layer’s purpose.

Lower Computer

  • Sensor Layer
    • Heart Sound Sensor
    • EMS Sensor
      β†’ Collect heart sounds and other biosignals from the human body in analog form
  • Signal Preprocessing and Conversion
    • Sensor signals are conditioned using a ThinkLabs module
    • Multi-channel analog signals are converted into digital data using the AD7606 ADC
  • Main Control MCU
    • STM32F103ZET6
    • Handles ADC data acquisition and packet construction
    • Manages data processing for Ethernet transmission
  • Network Interface
    • W5100 Ethernet Controller
    • Connected to the STM32 via SPI
    • Performs TCP/IP communication at the hardware level
    • Reliably transmits collected medical data to the upper system

Upper Computer

  • APSoC-Based Processing Platform
    • ZedBoard (Zynq APSoC)
    • Composed of a Processing System (PS) and Programmable Logic (PL)
  • Data Reception and Buffering
    • Receives data from the lower system via Ethernet
    • Stores and shares data in DDR3 memory
  • AI Processing and Application Execution
    • Runs on a Linux (Linaro)–based environment
    • Executes Qt-based UI and data acquisition/analysis applications
    • Performs heart disease classification and diagnostic assistance using AI models
  • Display Output
    • ADV7511 HDMI transmitter
    • Connects to LCD drivers and HDMI displays
    • Visualizes real-time waveforms and analysis results
  • Internal Interfaces
    • Data exchange between PS and PL via AXI interconnects

This architecture clearly separates medical signal acquisition, transmission, and analysis, ensuring scalability and system reliability.

πŸ“Œ Role and Application of the W5100

In this study, the W5100 Ethernet controller plays a core networking role.

  • Handles wired Ethernet data transmission on the STM32-based embedded board
  • Reliably delivers real-time heart sound and ECG data to the upper-level computer
  • Implements TCP/IP communication while minimizing MCU resource usage
  • Contributes to the high reliability and low latency required in medical environments

πŸ“Œ WIZnet Strategic Value

This project is a representative example of an Ethernet-based medical IoT system.
WIZnet’s W5500 hardware TCP/IP Ethernet chip can provide stable and reliable data transmission for STM32- and ESP32-based medical and healthcare devices, and it can serve as a practical reference for AI-assisted diagnostic devices and remote healthcare platforms like the system presented in this study.

πŸ“Œ Summary

This study is a successful example of large-scale screening for congenital heart disease by combining AI-based heart sound analysis with an Ethernet-based embedded system.
The architecture using STM32 and Ethernet (W5100) serves as a strong reference model for low-cost, high-reliability medical IoT design.
WIZnet Ethernet solutions have significant potential to expand as core network infrastructure in next-generation smart healthcare devices.


πŸ“Œ Overview

λ³Έ ν”„λ‘œμ νŠΈλŠ” AI 기반 μ‹¬μŒ(Heart Sound) 뢄석 μ‹œμŠ€ν…œμ„ μ΄μš©ν•΄ **μ„ μ²œμ„± 심μž₯μ§ˆν™˜(CHD)**을 λŒ€κ·œλͺ¨λ‘œ μŠ€ν¬λ¦¬λ‹ν•˜λŠ” 연ꡬ이닀. μ „μž μ²­μ§„κΈ°λ‘œ μˆ˜μ§‘ν•œ μ‹¬μŒ 데이터λ₯Ό STM32 + Ethernet(W5100) 기반 μž„λ² λ””λ“œ μ‹œμŠ€ν…œμ—μ„œ μƒμœ„ μ»΄ν“¨νŒ… ν”Œλž«νΌμœΌλ‘œ μ „μ†‘ν•˜κ³ , λ”₯λŸ¬λ‹(RNN/CNN) λͺ¨λΈμ„ 톡해 μ§ˆν™˜μ„ μžλ™ λΆ„λ₯˜ν•œλ‹€. 이 μ‹œμŠ€ν…œμ€ 의료 인프라가 λΆ€μ‘±ν•œ μ§€μ—­μ—μ„œλ„ ν™œμš© κ°€λŠ₯ν•œ μ €λΉ„μš©Β·κ³ ν™•μž₯μ„± 진단 보쑰 μ†”λ£¨μ…˜μ„ λͺ©ν‘œλ‘œ ν•œλ‹€.

πŸ“Œ Features

  • AI μ‹¬μŒ 진단
    • ASD, VSD, PDA λ“± μ£Όμš” CHD μœ ν˜• μžλ™ λΆ„λ₯˜
    • RNN 기반 λͺ¨λΈλ‘œ μ΅œλŒ€ μ•½ 90% 정확도 달성
  • λŒ€κ·œλͺ¨ μŠ€ν¬λ¦¬λ‹ 검증
    • 518,258λͺ… λŒ€μƒ μ‹€μ œ ν˜„μž₯ μŠ€ν¬λ¦¬λ‹ μˆ˜ν–‰
  • μ €λΉ„μš© ν•˜λ“œμ›¨μ–΄ ꡬ성
    • STM32 마이크둜컨트둀러 + Ethernet λ„€νŠΈμ›Œν¬
  • ν˜„μž₯ μΉœν™”μ  섀계
    • μ΄λ™ν˜•Β·μ €μ „λ ₯Β·μ‹€μ‹œκ°„ 데이터 μˆ˜μ§‘ 및 전솑

πŸ“Œ System Architecture

πŸ“Œ System Architecture
πŸ“Œ System Architecture

ν•΄λ‹Ή μ‹œμŠ€ν…œμ€ ν•˜μœ„ μž„λ² λ””λ“œ μˆ˜μ§‘ μ‹œμŠ€ν…œ(Lower Computer) κ³Ό μƒμœ„ μ²˜λ¦¬Β·ν‘œμ‹œ μ‹œμŠ€ν…œ(Upper Computer) 으둜 λͺ…ν™•νžˆ λΆ„λ¦¬λœ 2-계측 ꡬ쑰λ₯Ό κ°€μ§„λ‹€.
이 κ΅¬μ‘°λŠ” 의료 μ‹ ν˜Έμ˜ μ•ˆμ •μ  μˆ˜μ§‘, μ‹ λ’°μ„± μžˆλŠ” 전솑, κ³ μ„±λŠ₯ AI 처리λ₯Ό λͺ©μ μ— 맞게 λΆ„λ‹΄ν•˜λ„λ‘ μ„€κ³„λ˜μ—ˆλ‹€.

ν•˜μœ„ μ‹œμŠ€ν…œ (Lower Computer)

  • μ„Όμ„œ 계측
    • μ‹¬μŒ μ„Όμ„œ(Heart Sound Sensor)
    • EMS μ„Όμ„œ
      β†’ μΈμ²΄μ—μ„œ λ°œμƒν•˜λŠ” μ‹¬μŒ 및 생체 μ‹ ν˜Έλ₯Ό μ•„λ‚ λ‘œκ·Έ ν˜•νƒœλ‘œ μˆ˜μ§‘
  • μ‹ ν˜Έ μ „μ²˜λ¦¬ 및 λ³€ν™˜
    • ThinkLabs λͺ¨λ“ˆμ„ 톡해 μ„Όμ„œ μ‹ ν˜Έ μ •ν˜•
    • AD7606 ADCλ₯Ό μ‚¬μš©ν•΄ 닀채널 μ•„λ‚ λ‘œκ·Έ μ‹ ν˜Έλ₯Ό λ””μ§€ν„Έ λ°μ΄ν„°λ‘œ λ³€ν™˜
  • 메인 μ œμ–΄ MCU
    • STM32F103ZET6
    • ADC 데이터 μˆ˜μ§‘ 및 νŒ¨ν‚· ꡬ성
    • Ethernet 전솑을 μœ„ν•œ 데이터 처리 λ‹΄λ‹Ή
  • λ„€νŠΈμ›Œν¬ μΈν„°νŽ˜μ΄μŠ€
    • W5100 Ethernet 컨트둀러
    • STM32와 SPI둜 μ—°κ²°
    • TCP/IP 톡신을 ν•˜λ“œμ›¨μ–΄ λ ˆλ²¨μ—μ„œ 처리
    • μˆ˜μ§‘λœ 의료 데이터λ₯Ό μ•ˆμ •μ μœΌλ‘œ μƒμœ„ μ‹œμŠ€ν…œμœΌλ‘œ 전솑

μƒμœ„ μ‹œμŠ€ν…œ (Upper Computer)

  • APSoC 기반 처리 ν”Œλž«νΌ
    • ZedBoard (Zynq APSoC)
    • PS(Processing System)와 PL(Programmable Logic)둜 ꡬ성
  • 데이터 μˆ˜μ‹  및 버퍼링
    • Ethernet을 톡해 ν•˜μœ„ μ‹œμŠ€ν…œμœΌλ‘œλΆ€ν„° 데이터 μˆ˜μ‹ 
    • DDR3 λ©”λͺ¨λ¦¬μ— 데이터 μ €μž₯ 및 곡유
  • AI 처리 및 μ• ν”Œλ¦¬μΌ€μ΄μ…˜ μ‹€ν–‰
    • Linux(Linaro) 기반 ν™˜κ²½
    • QT 기반 UI 및 데이터 μˆ˜μ§‘Β·λΆ„μ„ ν”„λ‘œκ·Έλž¨ μ‹€ν–‰
    • AI λͺ¨λΈμ„ ν†΅ν•œ 심μž₯ μ§ˆν™˜ λΆ„λ₯˜ 및 보쑰 진단 μˆ˜ν–‰
  • λ””μŠ€ν”Œλ ˆμ΄ 좜λ ₯
    • ADV7511 HDMI νŠΈλžœμŠ€λ―Έν„°
    • LCD λ“œλΌμ΄λ²„ 및 HDMI λ””μŠ€ν”Œλ ˆμ΄ μ—°κ²°
    • μ‹€μ‹œκ°„ νŒŒν˜•, 뢄석 κ²°κ³Ό μ‹œκ°ν™”
  • λ‚΄λΆ€ μΈν„°νŽ˜μ΄μŠ€
    • AXI 인터컀λ„₯트λ₯Ό 톡해 PS–PL κ°„ 데이터 κ΅ν™˜

이 κ΅¬μ‘°λŠ” 의료 μ‹ ν˜Έ μˆ˜μ§‘β€“μ „μ†‘β€“λΆ„μ„μ„ λͺ…ν™•νžˆ 뢄리해 ν™•μž₯μ„±κ³Ό μ•ˆμ •μ„±μ„ ν™•λ³΄ν•œλ‹€.

πŸ“Œ Role and Application of the W5100

λ³Έ μ—°κ΅¬μ—μ„œλŠ” W5100 Ethernet μ»¨νŠΈλ‘€λŸ¬κ°€ 핡심 λ„€νŠΈμ›Œν¬ 역할을 μˆ˜ν–‰ν•œλ‹€.

  • STM32 기반 μž„λ² λ””λ“œ λ³΄λ“œμ—μ„œ μœ μ„  Ethernet 데이터 전솑 λ‹΄λ‹Ή
  • μ‹€μ‹œκ°„ μ‹¬μŒΒ·ECG 데이터λ₯Ό μƒμœ„ μ»΄ν“¨ν„°λ‘œ μ•ˆμ •μ μœΌλ‘œ 전달
  • MCU μžμ›μ„ μ΅œμ†Œλ‘œ μ‚¬μš©ν•˜λ©΄μ„œ TCP/IP 톡신 κ΅¬ν˜„
  • 의료 ν™˜κ²½μ—μ„œ μš”κ΅¬λ˜λŠ” μ‹ λ’°μ„±Β·μ§€μ—° μ΅œμ†Œν™”μ— κΈ°μ—¬

πŸ“Œ WIZnet Strategic Value

이 ν”„λ‘œμ νŠΈλŠ” Ethernet 기반 의료 IoT μ‹œμŠ€ν…œμ˜ λŒ€ν‘œ 사둀닀.
WIZnet의 W5500 ν•˜λ“œμ›¨μ–΄ TCP/IP Ethernet 칩은 STM32Β·ESP32 기반 μ˜λ£ŒΒ·ν—¬μŠ€μΌ€μ–΄ μž₯λΉ„μ—μ„œ μ•ˆμ •μ μΈ 데이터 전솑을 μ œκ³΅ν•  수 있으며,
λ³Έ 연ꡬ와 같은 AI 진단 보쑰 μž₯μΉ˜Β·μ›κ²© 의료 ν”Œλž«νΌμ—μ„œ μ‹€μ§ˆμ μΈ 레퍼런슀둜 ν™œμš© κ°€λŠ₯ν•˜λ‹€.

πŸ“Œ Summary

λ³Έ μ—°κ΅¬λŠ” AI μ‹¬μŒ 뢄석 + Ethernet μž„λ² λ””λ“œ μ‹œμŠ€ν…œμ„ κ²°ν•©ν•΄ μ„ μ²œμ„± 심μž₯μ§ˆν™˜μ„ λŒ€κ·œλͺ¨λ‘œ μŠ€ν¬λ¦¬λ‹ν•œ 성곡적인 사둀닀.
STM32와 Ethernet(W5100)을 ν™œμš©ν•œ κ΅¬μ‘°λŠ” μ €λΉ„μš©Β·κ³ μ‹ λ’° 의료 IoT μ„€κ³„μ˜ 쒋은 μ°Έκ³  λͺ¨λΈμ΄ λœλ‹€.
WIZnet Ethernet μ†”λ£¨μ…˜μ€ μ΄λŸ¬ν•œ μ°¨μ„ΈλŒ€ 슀마트 ν—¬μŠ€μΌ€μ–΄ κΈ°κΈ°μ—μ„œ 핡심 λ„€νŠΈμ›Œν¬ μΈν”„λΌλ‘œ ν™•μž₯ κ°€λŠ₯성이 크닀.

 

 

 

Comments

Similar projects you might like

Comments