Unsupervised hidden semi-Markov model for automatic beat onset detection in 1D Doppler ultrasound

Abstract

Objective: One dimensional (1D) Doppler ultrasound (DUS) is commonly used for fetal health assessment, during both regular prenatal visits and labor. It is used in preference to ECG and other modalities because of its simplicity and cost. To date, all analysis of such data has been confined to a smoothed, windowed heart rate estimation derived from the 1D DUS signal, reducing the potential of short-term variability information. A first step in improving the assessment of short-term variability of the fetal heart rate (FHR) is through implementing an accurate beat detector for 1D DUS signals. Approach: This work presents an unsupervised probabilistic segmentation method enabled by a hidden semi-Markov model (HSMM). The proposed method employs envelope and spectral features for an online segmentation of fetal 1D DUS signal. The beat onsets and fetal cardiac beat-to-beat intervals are then estimated from the segmentations. For this work, two data sets were used, including 1D DUS recordings from five fetuses recorded in Germany, comprising 6521 beats and 45.06 minutes of data (dataset 1). Simultaneous fetal ECG (fECG) was used as the reference for beat timing. Dataset 2, comprising 4044 beats captured from 17 subjects in the UK was hand scored for beat location and was used as an independent held-out test set. Leave-one-out subject cross-validation was used for parameter tuning on dataset 1. No retraining was performed for dataset 2. To assess the performance of the beat onset detection, the root mean square error (RMSE), F1 score, sensitivity, positive predictivity (PPV) and the error in several standard common heart rate variability metrics were used. These metrics were evaluated on three fiducial points: (1) beat onset, (2) beat offset, and (3) middle of beat interval. Main results: In dataset 1, the proposed method provided an RMSE of 20 ms, F1 score of 97.5 %, a Se of 97.6%, and a PPV of 97.3%. In dataset 2, the proposed method achieved an RMSE of 26 ms, an F1 score of 98.5 %, a Se of 98.0 % and a PPV of 98.9 %. It was also determined that the best beat-to-beat interval was derived from the onset of each beat. For the dataset 2, significant correlations were found in all short term heart rate variability metrics tested, both in the time and frequency domain. Only the proportion of successive normal-to-normal interval differences greater than 20 ms (pNN20) exhibited a significant absolute difference. Significance: This work presents the first-ever description of an algorithm to identify cardiac beats with 1D DUS, closely matching the fetal ECG-derived beats, to enable short-term heart rate variability analysis. The novel algorithm proposed requires no human labeling of data, and could have applicability beyond 1D DUS to other similar highly variable time series.

Estimation and Discriminability of Doppler Ultrasound Fetal Heart Rate Variability Measures

“…We are aware that a number of more sophisticated methods have been proposed to improve the accuracy of the beat-to-beat estimation of the FHR. These methods are based on the AC method with adaptive parameters, or utilize ML models to extract the beat-to-beat sequence from the DUS signal ( Peters et al, 2004 ; Jezewski et al, 2011 ; Alnuaimi et al, 2017 ; Valderrama et al, 2019 ; Katebi et al, 2020 ). The effect of those methods on fHRV as function of their parameters is an important question to be explored but is beyond the scope of the present paper.…”

“… Valderrama et al (2019) developed an open-source AC method that optimizes the peak search parameters using Bayesian optimization. Another approach by Katebi et al (2020) applied unsupervised hidden semi-Markov models to segment the DUS signal for FHR estimation. This approach was able to recover HF features that were very close to those of fECG ( Katebi et al, 2020 ).…”

“…Another approach by Katebi et al (2020) applied unsupervised hidden semi-Markov models to segment the DUS signal for FHR estimation. This approach was able to recover HF features that were very close to those of fECG ( Katebi et al, 2020 ). Despite their improvements, none of these sophisticated methods have yet been applied in bedside monitors ( Jezewski et al, 2017 ).…”

A review of fetal cardiac monitoring, with a focus on low- and middle-income countries

“…In the event a risk factor was identified, the app connected the TBA to appropriate (local or remote) medical care through a voice call, to provide decisional support and onward referral to appropriate healthcare, if needed. The Doppler signal and maternal blood pressure recorded with the mHealth system has allowed the development of different modules for providing estimates of fHRV, gestational age, and hypertension (Valderrama et al 2017, Valderrama et al 2018, Valderrama et al 2019, Valderrama et al 2020a, Valderrama et al 2020b, Katebi et al 2020.…”

A Novel Ultrasonic Doppler Fetal Heart Rate Detection System Using Windowed Digital Demodulation

“…Subsequently, Katebi et al proposed a probabilistic segmentation method enabled by a hidden semi-Markov model, which was regarded as an unsupervised approach along with the use of spectral and temporal features as an input to a clustering algorithm, significantly improved DUS segmentation and heart rate variability assessment. They presented one of the most accurate approaches for the beat-tobeat monitoring of fetuses using 1D-DUS signals [22]. The above methods are based on limited data sets.…”