Filling of Fetal Heart Rate Signal: Diffusion Model Based on Dimension Construction and Period Segmentation
- 10 July 2024
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Consumer Electronics
- Vol. 70 (3) , 5453-5460
Abstract
Fetal Heart Rate (FHR) signal is widely used in doppler fetal heart monitors. However, incomplete FHR signals reduce the effectiveness of fetal heart rate monitoring. Filling missing data is a key technique to improve FHR quality, but existing filling algorithms lack consideration the correlation of FHR signals. Therefore, we focus on two correlations related to FHR, and propose a filling algorithm called Diffusion model for Missing Data Imputation in FHR (DMDI-FHR). Firstly, we construct the Dual-dimensional Sample Construction (DDSC) method that finds two FHR signals with maximizing the difference to form correlations between them. Secondly, the Multi-period Decomposition (MPD) method is introduced to obtain the internal correlation of FHR signal. Finally, DMDI-FHR algorithm controls the filling process based on the diffusion model. Experimental results demonstrate the performance of DMDI-FHR algorithm, which provides an effective way to improve the quality of FHR signal.Keywords
Funding Information
- Zhejiang Provincial Natural Science Foundation of China (LDT23F01012F01, LDT23F01015F01)
- National Natural Science Foundation of China (62071162, 62071327, 62301205)
This publication has 35 references indexed in Scilit:
- Double-sided asymmetric method for automated fetal heart rate baseline calculationPhysical and Engineering Sciences in Medicine, 2023
- SO-slope entropy coupled with SVMD: A novel adaptive feature extraction method for ship-radiated noiseOcean Engineering, 2023
- Digital signal extraction approach for cardiotocography imageComputer Methods and Programs in Biomedicine, 2022
- Imputation Method for Fetal Heart Rate Signal Evaluation Based on Optimal Transport TheorySN Computer Science, 2021
- Missing data estimation method for time series data in structure health monitoring systems by probability principal component analysisAdvances in Engineering Software, 2020
- Predicting fetal hypoxia using common spatial pattern and machine learning from cardiotocography signalsApplied Acoustics, 2020
- Smoothed LSTM-AE: A spatio-temporal deep model for multiple time-series missing imputationNeurocomputing, 2020
- Co-regularized nonnegative matrix factorization for evolving community detection in dynamic networksInformation Sciences, 2020
- A data imputation method for multivariate time series based on generative adversarial networkNeurocomputing, 2019
- A Recovery of FHR Signal in the Embedded Space2014