Skip to content

README.md

Latest commit

 

History

History
91 lines (56 loc) · 2.62 KB

README.md

File metadata and controls

91 lines (56 loc) · 2.62 KB

A Hybrid Multi-Factor Network with Dynamic Sequence Modeling for Early Warning of Intraoperative Hypotension

License
Python Version

This repository provides the implementation of HMF, a Hybrid Multi-Factor framework for dynamic intraoperative hypotension (IOH) prediction.

🔍 Overview

The HMF framework is designed to predict IOH using Mean Arterial Pressure (MAP) as a key indicator. Traditional approaches often use static models that fail to capture the dynamic nature of physiological signals. In contrast, our framework employs advanced techniques to address challenges such as distributional drift and temporal dependencies in physiological data.

Key Features

  • Transformer Encoder: Captures the temporal evolution of MAP through patch-based input representation.
  • Symmetric Normalization and De-Normalization: Mitigates distributional drift in data to ensure robust performance across varying conditions.
  • Sequence Decomposition: Disaggregates input signals into trend and seasonal components for improved sequence modeling.
  • State-of-the-Art Performance: Extensively validated on two real-world datasets, outperforming competitive baselines.

📥 Installation

Prerequisites

  • Python 3.7 or later
  • PyTorch >= 1.10
  • Other dependencies listed in requirements.txt

Steps

  1. Clone the repository:

    git clone https://github.com/yourusername/HMF-IOH.git
cd HMF-IOH

  2. Install dependencies:

    pip install -r requirements.txt

🚀 Usage

Data Preparation

  • Place your physiological signal datasets in the data/ directory.
  • Ensure the data format follows the structure outlined in docs/data_format.md.

Training

To train the model, run:

python train.py --config configs/default.yaml

Evaluation

Evaluate the trained model using:

python evaluate.py --model checkpoints/best_model.pth --data data/test_data.csv

Configuration

Modify configs/default.yaml to customize training parameters, such as learning rate, batch size, and model architecture.

📊 Results

The HMF framework demonstrates state-of-the-art performance on two real-world datasets. See the paper for detailed experimental results.

🤝 Contributing

We welcome contributions! Please read our CONTRIBUTING.md file before submitting pull requests.

📜 License

This project is licensed under the MIT License. See the LICENSE file for details.