PhysioWave: A Multi-Scale Wavelet-Transformer for Physiological Signal Representation

<div align="center">

</div> <div align="center">

Official PyTorch implementation of PhysioWave, accepted at NeurIPS 2025

A novel wavelet-based architecture for physiological signal processing that leverages adaptive multi-scale decomposition and frequency-guided masking to advance self-supervised learning

</div>

---

🌟 Key Features

<table> <tr> <td width="50%">

✨ Learnable Wavelet Decomposition

• Adaptive multi-resolution analysis
• Soft gating mechanism for optimal wavelet selection

📊 Frequency-Guided Masking

• Novel masking strategy prioritizing high-energy components
• Superior to random masking for signal representation

</td> <td width="50%">

🔗 Cross-Scale Feature Fusion

• Attention-based fusion across decomposition levels
• Hierarchical feature integration

🧠 Multi-Modal Support

• Unified framework for ECG and EMG signals
• Extensible to other physiological signals

</td> </tr> </table> <div align="center">

📈 Large-Scale Pretraining: Models trained on 182GB of ECG and 823GB of EMG data

</div>

---

🏗️ Model Architecture

<div align="center"> <img src="fig/model.png" alt="PhysioWave Architecture" width="90%"> </div>

Pipeline Overview

The PhysioWave pretraining pipeline consists of five key stages:

1. Wavelet Initialization: Standard wavelet functions (e.g., 'db6', 'sym4') generate learnable low-pass and high-pass filters
2. Multi-Scale Decomposition: Adaptive wavelet decomposition produces multi-scale frequency-band representations
3. Patch Embedding: Decomposed features are processed into spatio-temporal patches with FFT-based importance scoring
4. Masked Encoding: High-scoring patches are masked and processed through Transformer layers with rotary position embeddings
5. Reconstruction: Lightweight decoder reconstructs masked patches for self-supervised learning

Core Components

| Component | Description | |-----------|-------------| | 🌊 Learnable Wavelet Decomposition | Adaptively selects optimal wavelet bases for input signals | | 📐 Multi-Scale Feature Reconstruction | Hierarchical decomposition with soft gating between scales | | 🎯 Frequency-Guided Masking | Identifies and masks high-energy patches for self-supervised learning | | 🔄 Transformer Encoder/Decoder | Processes masked patches with rotary position embeddings |

---

📊 Performance Highlights

Benchmark Results

<div align="center">

| Task | Dataset | Metric | Performance | |------|---------|--------|-------------| | ECG Arrhythmia | PTB-XL | Accuracy | 73.1% | | ECG Multi-Label | CPSC 2018 | F1-Micro | 77.1% | | ECG Multi-Label | Shaoxing | F1-Micro | 94.6% | | EMG Gesture | EPN-612 | Accuracy | 94.5% |

</div>

Multi-Label Classification Detailed Metrics

<details> <summary><b>CPSC 2018 Dataset (9-Class Multi-Label)</b></summary> <div align="center">

| Metric | Micro-Average | Macro-Average | |--------|---------------|---------------| | Precision | 0.7389 | 0.6173 | | Recall | 0.8059 | 0.6883 | | F1-Score | 0.7709 | 0.6500 | | AUROC | 0.9584 | 0.9280 |

</div>

Dataset Details:

• 9 official diagnostic classes (SNR, AF, IAVB, LBBB, RBBB, PAC, PVC, STD, STE)
• 12-lead ECG signals at 500 Hz
• Record-level split to prevent data leakage

</details> <details> <summary><b>Chapman-Shaoxing Dataset (4-Class Multi-Label)</b></summary> <div align="center">

| Metric | Micro-Average | Macro-Average | |--------|---------------|---------------| | Precision | 0.9389 | 0.9361 | | Recall | 0.9536 | 0.9470 | | F1-Score | 0.9462 | 0.9413 | | AUROC | 0.9949 | 0.9930 |

</div>

Dataset Details:

• 4 merged diagnostic classes (SB, AFIB, GSVT, SR)
• 12-lead ECG signals at 500 Hz
• Balanced multi-label distribution

</details>

---

💾 Pretrained Models

<div align="center">

📥 Download Pretrained Models

</div>

| Model | Parameters | Training Data | Description | |-------|------------|---------------|-------------| | ecg.pth | 14M | 182GB ECG | ECG pretrained model | | emg.pth | 5M | 823GB EMG | EMG pretrained model |

Usage:

# Load pretrained model
checkpoint = torch.load('ecg.pth')
model.load_state_dict(checkpoint['model_state_dict'])

---

🚀 Quick Start

Prerequisites

# Clone repository
git clone https://github.com/ForeverBlue816/PhysioWave.git
cd PhysioWave

# Create conda environment
conda create -n physiowave python=3.11
conda activate physiowave

# Install PyTorch (CUDA 12.1)
pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu121

# Install requirements
pip install -r requirements.txt

📦 Data Preparation

<details> <summary><b>Dataset Download Links</b></summary>

ECG Datasets

• PTB-XL Database - 21,837 clinical ECG records
• MIMIC-IV-ECG - 800K+ ECG recordings
• PhysioNet Challenge 2021 - Multi-database ECG
• CPSC 2018 - Arrhythmia classification challenge
• Chapman-Shaoxing - Large-scale 12-lead ECG

EMG Datasets

• EPN-612 Dataset - 612 hand gestures
• NinaPro Database DB6 - HD-sEMG recordings

</details> <details> <summary><b>Data Format Specifications</b></summary>

HDF5 Structure

# Single-label classification
{
    'data': (N, C, T),   # Signal data: float32
    'label': (N,)        # Labels: int64
}

# Multi-label classification
{
    'data': (N, C, T),   # Signal data: float32
    'label': (N, K)      # Multi-hot labels: float32
}

Dimensions:

• N = Number of samples
• C = Number of channels
• T = Time points
• K = Number of classes (multi-label only)

Signal Specifications

| Signal | Channels | Length | Sampling Rate | Normalization | |--------|----------|--------|---------------|---------------| | ECG | 12 | 2048 | 500 Hz | MinMax [-1,1] or Z-score | | EMG | 8 | 1024 | 200-2000 Hz | Max-abs or Z-score |

</details>

🔄 Preprocessing Examples

<details> <summary><b>ECG Preprocessing (PTB-XL - Single-Label)</b></summary>

# Download PTB-XL dataset
wget -r -N -c -np https://physionet.org/files/ptb-xl/1.0.3/

# Preprocess for single-label classification
python ECG/ptbxl_finetune.py

Output files:

• train.h5 - Training data with shape (N, 12, 2048)
• val.h5 - Validation data
• test.h5 - Test data

Label format: (N,) with 5 superclasses (NORM, MI, STTC, CD, HYP)

</details> <details> <summary><b>ECG Preprocessing (CPSC 2018 - Multi-Label)</b></summary>

# Preprocess CPSC 2018 dataset
python ECG/cpsc_multilabel.py

Output files:

• cpsc_9class_train.h5 - Training data
• cpsc_9class_val.h5 - Validation data
• cpsc_9class_test.h5 - Test data
• cpsc_9class_info.json - Dataset metadata
• label_map.json - Class mappings
• record_splits.json - Record-level split info

Label format: (N, 9) with 9 official CPSC classes

</details> <details> <summary><b>ECG Preprocessing (Chapman-Shaoxing - Multi-Label)</b></summary>

# Preprocess Chapman-Shaoxing dataset
python ECG/shaoxing_multilabel.py

Output files:

• train.h5 - Training data
• val.h5 - Validation data
• test.h5 - Test data
• dataset_info.json - Metadata
• record_splits.json - Split information

Label format: (N, 4) with 4 merged classes (SB, AFIB, GSVT, SR)

</details> <details> <summary><b>EMG Preprocessing (EPN-612)</b></summary>

# Download from Zenodo and preprocess
python EMG/epn_finetune.py

Output files:

• epn612_train_set.h5 - Training set (N, 8, 1024)
• epn612_val_set.h5 - Validation set
• epn612_test_set.h5 - Test set

Label format: (N,) with 6 gesture classes

</details>

---

🎯 Training

Pretraining

<details> <summary><b>ECG Pretraining</b></summary>

# Edit ECG/pretrain_ecg.sh to set data paths
bash ECG/pretrain_ecg.sh

Key parameters:

--mask_ratio 0.7                    # Mask 70% of patches
--masking_strategy frequency_guided # Use frequency-guided masking
--importance_ratio 0.7              # Balance importance vs randomness
--epochs 100                        # Pretraining epochs

</details> <details> <summary><b>EMG Pretraining</b></summary>

# Edit EMG/pretrain_emg.sh to set data paths
bash EMG/pretrain_emg.sh

Key parameters:

--mask_ratio 0.6                    # Mask 60% of patches
--in_channels 8                     # 8-channel EMG
--wave_kernel_size 16               # Smaller kernel for EMG

</details>

---

Fine-tuning

Single-Label Classification

<details> <summary><b>Standard Fine-tuning (ECG/EMG)</b></summary>

# ECG fine-tuning (PTB-XL)
bash ECG/finetune_ecg.sh

# EMG fine-tuning (EPN-612)
bash EMG/finetune_emg.sh

Example command:

torchrun --nproc_per_node=4 finetune.py \
  --train_file path/to/train.h5 \
  --val_file path/to/val.h5 \
  --test_file path/to/test.h5 \
  --pretrained_path pretrained/ecg.pth \
  -