Fetal Pulsed-Wave Doppler Atrioventricular Activity Detection by Envelope Extraction and Processing

Abstract

Pulsed-Wave Doppler (PWD) is a diagnostic ultrasound technique widely used for fetal heart rate monitoring. Fetal PWD is particularly challenging since, beyond being intrinsically operator-dependent, different issues related to the fetal heart size, the fetal movements and the ultrasound artifacts appear. In long PWD recordings, the signal segments completely meaningful for a morphological analysis, i.e. including a readable atrial and ventricular activity, are then limited in number and duration. In this work, an approach for the automatic detection of the meaningful fetal cardiac activity from PWD video recordings is presented and evaluated, using the annotations made by an expert cardiologist. It consists of the video pre-processing for image thresholding, filtering and envelope extraction by edge detection, and a supervised classification stage. A dataset including 30 signals from 17 pregnant women was adopted, extracting from it multiple segments, including different quality recordings. A supervised classification approach for the detection of the signal segments completely meaningful for a morphological analysis was then applied, revealing an accuracy greater than 99%.

Impact of pulsed-wave-Doppler velocity-envelope tracing techniques on classification of complete fetal cardiac cycles

“…Moreover, we we used the Wilcoxon signed rank test to make a pairwise comparison of all the distributions with that of the SOC. The results of all the comparisons, except that for the chain proposed by Sulas et al [ 16 ], revealed statistically significant differences ( p < 0.05). This is not surprising because the chain proposed by Sulas et al [ 16 ] and the SOC adopted the same methods for image binarization and envelope extraction, which were identified as the most important steps in the work chain.…”

“…The chain proposed by Magagnin et al [ 21 ] uses the same envelope extraction adopted by Sulas et al [ 16 ] and the SOC, but not the same binarization technique (adaptive thresholding rather than the Otsu 2D). If we look at Fig 8 , the adaptive threshold (used in chain number 7 ( Table 2 ), whereby the MC was modified to introduce an adaptive threshold,) and the Otsu 2D (binarization algorithm used in the MC) have similar distributions.…”

“…White-region contour method . This method, used by the authors in [ 16 , 23 ], traces the boundary pixels of all white regions to obtain contours as follows: where y b is the baseline position, i.e., the horizontal axis line that divides the image into two parts. Depending on the direction of the blood flow away from or toward the transducer, one part is characterized by positive waves and the other by negative waves.…”

A non-invasive multimodal foetal ECG–Doppler dataset for antenatal cardiology research

“…As the dataset was not conceived for foetal heart rate (fHR) variability studies (typically requiring at least five minutes of continuous recording), some of the records are relatively short and not recommended for signal processing algorithms requiring a long training period. Some of the cases for this database were previously used for studies on PWD [27][28][29] and fECG [30][31][32][33] .…”

“…Image binarization: based on 2D Otsu's method, we adopted a global threshold from a grey-level-median histogram 53 . After a test phase, this threshold was selected as the most efficient and robust regarding noise and information preservation 28 .…”