Project Proposal - EML-Labs/PPG-Peak-Detection-on-FPGA GitHub Wiki

FPGA Project Proposal – EML Labs

PPG Peak Detection for Heart Rate Variability Monitoring on FPGA

Introduction

Heart Rate Variability (HRV) is a key physiological marker to measure the non-linear dynamics of the heart. It serves as an indicator for the autonomic nervous system. Monitoring HRV provides valuable insight into overall cardiovascular health and is widely recognized as a reliable predictor of cardiac arrhythmias, particularly Atrial Fibrillation (AFib).

Photoplethysmography (PPG) is a non-invasive optical method that captures blood volume variations in peripheral tissues and provides insights on cardiac cycles from pulse waveforms. Accurate detection of peaks of these waveforms and precise calculation of time intervals between peaks provide a strong basis for HRV measurement.

Traditional HRV monitoring often depends on software-based filters and algorithms, which can be inefficient for wearable devices with limited computational power and strict energy constraints. This project aims to implement application-specific hardware for PPG peak detection and HRV estimation on FPGA. By leveraging inherent parallelism and low power consumption, the goal is efficient and continuous physiological monitoring.

Problem Statement

Conventional PPG-based HRV monitoring systems suffer from:

  • Processing delays
  • Inaccurate measurements due to heavy software-based filters (e.g., complex digital filters, adaptive algorithms)

👉 Therefore, there is a need for a real-time, hardware-accelerated solution that can efficiently process PPG signals and compute HRV metrics with minimal latency and high accuracy.

Objectives

  • Design, synthesize, and analyze a real-time digital signal processing pipeline for PPG peak detection on FPGA devices.
  • Analyze the timing behavior of peak detection and inter-beat interval calculation modules.
  • Design and implement sequential digital circuits with feedback mechanisms for Peak-to-Peak Interval (PPI) calculation and HRV metrics extraction.
  • Simulate and validate the integrated system, optimizing for accuracy and throughput.
  • Evaluate the feasibility of continuous monitoring in wearable devices.

Literature Review

Research on HRV monitoring using PPG signals processed on microcontrollers has shown reliability and clinical viability (Lin et al., 2025). Real-time edge computing approaches enhance privacy (Bohrium, 2024; IJRPR, 2023).

FPGAs, however, remain underutilized in PPG-based HRV monitoring, despite offering:

  • Parallelism → faster and efficient DSP
  • Low power consumption → critical for wearables

Most FPGA biomedical signal processing research has focused on ECG analysis, proving improved throughput and latency compared to microcontrollers (Rezaei et al., 2013). Exploring FPGA-based PPG HRV systems presents a powerful opportunity for next-gen wearable health monitoring.

Proposed Methodology

  1. Signal Acquisition – Input raw PPG signal via a PPG sensor
  2. Preprocessing – Filtering to remove noise
  3. Peak Detection – Use Waveform Envelope Peak Detection (WEPD) to minimize false positives/negatives
  4. PPI & HRV Computation – Calculate peak-to-peak intervals and derive time-domain HRV metrics
  5. FPGA Implementation – Develop modules using Verilog/VHDL
  6. Simulation & Testing – Use Xilinx Vivado for verification, timing analysis, and visualization of heart rate

Deliverables

  • HDL code for preprocessing, peak detection, and HRV computation
  • Functional simulation results and waveform plots
  • Synthesized FPGA design with resource utilization report
  • Timing and power analysis from Vivado
  • On-board FPGA test results with real/simulated PPG input
  • Accuracy comparison with software-based HRV calculations
  • Source code repository + user manual

Timeline

  • Week 1–2: Literature survey & environment setup
  • Week 3–4: Preprocessing filter design & simulation
  • Week 5–6: Peak detection module implementation
  • Week 7–8: PPI calculation & HRV extraction
  • Week 9–10: Integration & FPGA synthesis
  • Week 11–12: Testing with real PPG input
  • Week 13–14: Optimization, documentation & final report

Group Members

  • Niroshana H.K.Y. – 210436E
  • Wimalasiri W.M. – 210730B

References

  1. Colak, A. M. et al. Peak detection implementation for real-time signal analysis based on FPGA. Circuits and Systems 9.10 (2018): 148-167.
  2. Tsai, Y.-Y. et al. Photoplethysmography-based HRV analysis and machine learning for real-time stress quantification. APL Bioengineering, 2025.
  3. Kirti, H., Sohal, S. J. (2022). FPGA Implementation of Low Power Pre-processor Design for Biomedical Signal Processing. Springer.
  4. Rezaei, S. et al. Implementation of HRV signal processing into FPGA: System-on-chip design. Computing in Cardiology, 2013.
  5. Vaithianathan, M. FPGA-Based Smart Health Monitoring Systems for Wearable Devices. Int. J. of Intelligent Systems and Applications in Engineering, 2024.
  6. Edge Impulse Documentation. HR/HRV features. 2025.
  7. Gu, X. et al. A Real-Time FPGA-Based Accelerator for ECG Analysis and Diagnosis Using Association-Rule Mining. ACM TECS, 2016.