Estimation of Missing Data in Fetal Heart Rate Signals Using Shift-Invariant Dictionary

Signal Processing for Newborn Survival from labour to resuscitation

“…The threshold c is therefore set to 30 beats per minute. The threshold T k is set to 25 seconds based on Barzideh et al [73].…”

“…This can be caused by sensor movement, suboptimal placement of the sensor, maternal heart rate, doubling and halving of the FHR signal caused by the Doppler principle. Missing data can be estimated to resemble the measured data using dictionary learning [73,93]. Artefacts due to noise may affect the interpretability and should be removed for both visual interpretation and further digital analysis.…”

Reconstruction of Missing Samples in Antepartum and Intrapartum FHR Measurements Via Mini-Batch-Based Minimized Sparse Dictionary Learning

“…The miniMSDL can be seen in Algorithm 2 (Table II) where Condition 2 satisfies the constraints shown in equation (22), which is a variant of formula (19) and calculates the difference degree of W in the iterative update process. Using a mini-batch solution without adapting the entire data set would mean using the learned dictionary from the previous iteration as the initial dictionary for the current iteration.…”

“…On the one hand, most previous approaches are actually designed for an offline mode since the reconstruction relies on tens of thousands of training samples and is time-consuming and sometimes quite repetitive [10]- [12]. But on the other, traditional dictionary learning is usually computationally expensive to train as well as to use [18], [19], especially for an inpainting problem with long missing sample length. To reduce the training load, we adopted a model optimization strategy of the mini-batch version, and simultaneously used a conjugate gradient to solve the complex dictionary and sparse coefficient matrix.…”

Imputation Method for Fetal Heart Rate Signal Evaluation Based on Optimal Transport Theory

“…Take for example the application of SVT, LTV, and entropy. The preprocessing step is vital in FHR evaluation, because its quality influences the values of features and, consequently, the evaluation's performance [5][6][7]. For instance, [8] explored the quality of several STV and LTV functions when 0-50% of samples are missing; the functions were randomly picked in a 5 min FHR segment within the initial stage of labor.…”