Deep Learning-based ECG Modeling for ANS Monitoring
[NRF] Autonomic nervous system pathological change monitoring via deep learning-based ECG modeling for stress detection and depression prediction.
Period: 2021.09 – 2022.05
Funding: National Research Foundation of Korea (NRF)
Summary: Developed technology to detect autonomic nervous system changes through deep learning modeling of ECG signals, enabling monitoring of pathological changes caused by diseases such as major depressive disorder and panic disorder.
Key Responsibilities:
- Extracted RR intervals (RRI) from ECG signals and built database with QC
- Designed data preprocessing pipelines per model architecture (time-series, 1D CNN)
- Developed and validated Mental Stress Detection and Depression Prediction models using Residual 1D CNN, CRNN, and LSTM architectures, along with SOTA model implementations
Outcome: 2 SCI publications