Fig 4 - uploaded by Giovanni Magenes
Content may be subject to copyright.
(Left) ROC curves of APRS (blue thin line) and DPRS (red thick line) computed using 150 intervals. Cut-off value is 0.1327 for APRS and –0.1426 for DPRS. (Right) ROC curve for STV computed using 150 intervals. Cut-off value = 5.4658.
Source publication
Since the 1980s, cardiotocography (CTG) has been the most diffused technique to monitor fetal well-being during pregnancy. CTG consists of the simultaneous recording of fetal heart rate (FHR) signal and uterine contractions and its interpretation is usually performed through visual inspection by trained obstetric personnel. To reduce inter- and int...
Contexts in source publication
Context 1
... the results for the two populations. The boxplots confirm the statistics described above, since Delta, STV, APRS, and DPRS are the parameters that behave most efficiently in discriminating healthy and IUGR patients. In order to determine the optimal discrimination threshold, we computed the ROC curves for the most efficient parame- ters. Fig. 4 shows the ROC curves computed for the APRS, DPRS, and STV. For APRS, we obtained a cut-off point of 0.1327, which guarantees a sensitivity of 0.7413 and specificity of 0.8033. The corresponding AUC is 0.8235. For DPRS, we obtained a cut-off value of −0.1426, corresponding to a sensi- tivity of 0.7705 and specificity of 0.7541 (AUC = ...
Context 2
... eters (ApEn), and 3) other parameters computed on the PRSA curve. The results on a population of 122 subjects (61 normal and 61 IUGRs, carefully selected) show that the parameter we are proposing performs better than any other considered in the comparison. This is confirmed both by the t-tests on the two populations and by the ROC curves shown in Fig. 4. According to [14], the PRSA curve describes the main in- crease and decrease episodes (or patterns) in the FHR time series under analysis. Although these events do not correspond exactly to the definitions, usually employed in the clinical routine, of "acceleration" and "deceleration" of FHR signal, they provide almost the same ...
Similar publications
The purpose of this work is to present a new mathematical model for fetal monitoring simulation. It involves the simultaneous generation of fetal heart rate and maternal uterine contraction signals through a parametrical model. This model allows the generation of the main fetal monitoring dynamics including fetal movements, acceleration and deceler...
The purpose of this work is to present a new mathematical model for fetal monitoring simulation. It involves the generation of fetal heart rate and maternal uterine contraction signals simultaneously through a parametrical model. This model allows the generation of the main fetal monitoring dynamics including fetal movements, acceleration and decel...
Citations
... Prenatal monitoring and timely diagnosis are imperative for effectively addressing these conditions and minimizing potential complications, thereby decreasing fetal morbidity and mortality [3]. Cardiotocography (CTG) is currently the most widely used electronic fetal monitoring (EFM) device in clinical practice [4]. This method involves the transmission of ultrasonic waves through ultrasonic probes, followed by the reception of frequency-shift echo signals. ...
The fetal electrocardiogram (FECG) records changes in the graph of fetal cardiac action potential during conduction, reflecting the developmental status of the fetus in utero and its physiological cardiac activity. Morphological alterations in the FECG can indicate intrauterine hypoxia, fetal distress, and neonatal asphyxia early on, enhancing maternal and fetal safety through prompt clinical intervention, thereby reducing neonatal morbidity and mortality. To reconstruct FECG signals with clear morphological information, this paper proposes a novel deep learning model, CBLS-CycleGAN. The model’s generator combines spatial features extracted by the CNN with temporal features extracted by the BiLSTM network, thus ensuring that the reconstructed signals possess combined features with spatial and temporal dependencies. The model’s discriminator utilizes PatchGAN, employing small segments of the signal as discriminative inputs to concentrate the training process on capturing signal details. Evaluating the model using two real FECG signal databases, namely “Abdominal and Direct Fetal ECG Database” and “Fetal Electrocardiograms, Direct and Abdominal with Reference Heartbeat Annotations”, resulted in a mean MSE and MAE of 0.019 and 0.006, respectively. It detects the FQRS compound wave with a sensitivity, positive predictive value, and F1 of 99.51%, 99.57%, and 99.54%, respectively. This paper’s model effectively preserves the morphological information of FECG signals, capturing not only the FQRS compound wave but also the fetal P-wave, T-wave, P-R interval, and ST segment information, providing clinicians with crucial diagnostic insights and a scientific foundation for developing rational treatment protocols.
... Cardiotocography (CTG), also known as electronic fetal monitoring (EFM), is a common monitoring technique wherein clinicians assess the fetal health by analyzing signals related to the Fetal Heart Rate (FHR) and uterine contractions (UC) obtained from CTG. While CTG has become the most widely employed fetal monitoring method [2], its utility remains a subject of debate due to high interobserver (different specialists at the same time) and intraobserver (same specialist at different times) variability. Furthermore, CTG may lead to an increase in false positives and a higher rate of planned deliveries [3,4]. ...
Background
In clinical medicine, fetal heart rate (FHR) monitoring using cardiotocography (CTG) is one of the most commonly used methods for assessing fetal acidosis. However, as the visual interpretation of CTG depends on the subjective judgment of the clinician, this has led to high inter-observer and intra-observer variability, making it necessary to introduce automated diagnostic techniques.
Methods
In this study, we propose a computer-aided diagnostic algorithm (Hybrid-FHR) for fetal acidosis to assist physicians in making objective decisions and taking timely interventions. Hybrid-FHR uses multi-modal features, including one-dimensional FHR signals and three types of expert features designed based on prior knowledge (morphological time domain, frequency domain, and nonlinear). To extract the spatiotemporal feature representation of one-dimensional FHR signals, we designed a multi-scale squeeze and excitation temporal convolutional network (SE-TCN) backbone model based on dilated causal convolution, which can effectively capture the long-term dependence of FHR signals by expanding the receptive field of each layer’s convolution kernel while maintaining a relatively small parameter size. In addition, we proposed a cross-modal feature fusion (CMFF) method that uses multi-head attention mechanisms to explore the relationships between different modalities, obtaining more informative feature representations and improving diagnostic accuracy.
Results
Our ablation experiments show that the Hybrid-FHR outperforms traditional previous methods, with average accuracy, specificity, sensitivity, precision, and F1 score of 96.8, 97.5, 96, 97.5, and 96.7%, respectively.
Conclusions
Our algorithm enables automated CTG analysis, assisting healthcare professionals in the early identification of fetal acidosis and the prompt implementation of interventions.
... Several works employing measures extracted using PRSA have been applied to FHR analysis: in particular, acceleration and deceleration capacity [19][20][21], acceleration and deceleration phase-rectified slope [22], and deceleration reserve [23]. Some authors postulate that the deceleration capacity is a measure of the vagal control of the heart rate and the acceleration capacity is a measure of sympathetic activity [24]. ...
Cardiotocography (CTG) is the most common technique for electronic fetal monitoring and consists of the simultaneous recording of fetal heart rate (FHR) and uterine contractions. In analogy with the adult case, spectral analysis of the FHR signal can be used to assess the functionality of the autonomic nervous system. To do so, several methods can be employed, each of which has its strengths and limitations. This paper aims at performing a methodological investigation on FHR spectral analysis adopting 4 different spectrum estimators and a novel PRSA-based spectral method. The performances have been evaluated in terms of the ability of the various methods to detect changes in the FHR in two common pregnancy complications: intrauterine growth restriction (IUGR) and gestational diabetes. A balanced dataset containing 2178 recordings distributed between the 32nd and 38th week of gestation was used. The results show that the spectral method derived from the PRSA better differentiates high-risk pregnancies vs. controls compared to the others. Specifically, it more robustly detects an increase in power percentage within the movement frequency band and a decrease in high frequency between pregnancies at high risk in comparison to those at low risk.
Graphical abstract
... We computed the set of parameters listed in Table 1, which also include relevant references for their definition and computation. Regarding PRSA-related parameters, the PRSA signal was calculated from the fHR series expressed in bpm in agreement with [5,6]. When comparing results, it should be noted that [7] and other authors used the RR series expressed in ms instead. ...
... None of the classical time-domain parameters (i.e., STV, DELTA, LTI, II, mean frequency) showed significant differences, confirming the unsuitability of the indices most used in clinic for the management of diabetes in pregnancy. Similar results were also obtained in [5] and [10]. Several significant differences were instead identified in the spectral and non-linear analyses. ...
... These results suggest that gestational diabetes causes a state of fetal hyperactivation. Interestingly, despite our findings are inconsistent with the ones by Lobmaier et al. [5] who observed a significant increase in AAC and ADC, frequency domain parameters allow us to reach similar conclusions regarding the hyperactivation of the sympathetic nervous system in GDMs, which may explain the increased risk of developing hypertension later in life. Among the PRSA-related features, the one that better differentiated between the two populations is the Deceleration Reserve. ...
... Phase Rectified Signal Average (PRSA) [21] consists in the detection and the quantification of quasiperiodic oscillations in nonstationary signals compromised by noise and artifacts, by synchronizing the phase of all the periodic components. From PRSA we derived the Acceleration Phase Rectified Slope (APRS) and the Deceleration Phase Rectified Slope (DPRS) as suggested in [22]. These parameters are respectively global descriptors of rate of increase (APRS) and rate of decrease (DPRS) of the FHR time series. ...
Background:
The clinical diagnosis of late Fetal Growth Restriction (FGR) involves the integration of Doppler ultrasound data and Fetal Heart Rate (FHR) monitoring through computer assisted computerized cardiotocography (cCTG). The aim of the study was to evaluate the diagnostic power of combined Doppler and cCTG parameters by contrasting late FGR -and healthy controls.
Methods:
The study was conducted from January 2018 to May 2020. Only pregnant women who had the last Doppler measurement obtained within 1 week before delivery and cCTG performed within 24 h before delivery were included in the study. Two hundred forty-nine pregnant women fulfilling the inclusion criteria were enrolled in the study; 95 were confirmed as late FGR and 154 were included in the control group.
Results:
Among the extracted cCTG parameters, Delta Index, Short Term Variability (STV), Long Term Variability (LTV), Acceleration and Deceleration Phase Rectified Slope (APRS, DPRS) values were lower in the late FGR participants compared to the control group. In the FGR cohort, Delta, STV, APRS, and DPRS were found different when stratifying by MCA_PI (MCA_PI <5th centile or > 5th centile). STV and DPRS were the only parameters to be found different when stratifying by (UA_PI >95th centile or UA_PI <95th centile). Additionally, we measured the predictive power of cCTG parameters toward the identification of associated Doppler measures using figures of merit extracted from ROC curves. The AUC of ROC curves were accurate for STV (0,70), Delta (0,68), APRS (0,65) and DPRS (0,71) when UA_PI values were > 95th centile while, the accuracy attributable to the prediction of MCA_PI was 0.76, 0.77, 0.73, and 0.76 for STV, Delta, APRS, and DPRS, respectively. An association of UA_PI>95th centile and MCA_PI<5th centile with higher risk for NICU admission, was observed, while CPR < 5th centile resulted not associated with any perinatal outcome. Values of STV, Delta, APRS, DPRS were significantly lower for FGR neonates admitted to NICU, compared with the uncomplicated FGR cohort.
Conclusions:
The results of this study show the contribution of advanced cCTG parameters and fetal Doppler to the identification of late FGR and the association of those parameters with the risk for NICU admission.
Trial registration:
Retrospectively registered.
... Відомі певні результати використання даного методу, що дозволяють поліпшити діагностику дистресу або аритмії плода [6][7][8][9][10][11]. Серед показників функціонального стану плода за даними неінвазивної ЕКГ дуже перспективним є застосування коефіцієнтів акцелерацій АС і децелерацій DC. Використання AC/DC дозволяє не лише поліпшити діагностику загрозливих станів плода, але й проводити дослідження його неврологічного дозрівання [12,13]. ...
Research objective: to study the use of uterine activity and variables of acceleration capacity (AC) and deceleration capacity (DC) in uterine contractile activity and fetal well-being monitoring in women at risk of preterm birth.
Materials and methods. 292 pregnant women were included in the prospective study. All involved women underwent ultrasound cervicometry at 16 weeks. 124 pregnant women with a “short cervix” of the III (main) group were observed in the dynamics, as well as uterine activity and fetal AC/DC at 26, 32 and 38 weeks of gestation, and during labor were studied. In 112 women of group II the variables of fetal AC/DC were detected at these terms of pregnancy. They were monitored via conventional cardiotocography during labor. 56 pregnant women in group I with normal cervicometry were monitored.
Results. Sensitivity and specificity of the diagnosis of the threatened preterm birth in the main group was 97.30% and 94.74% respectively. Diagnostic accuracy in the case of uterine activity according to fetal non-invasive electrocardiography was 96.18%. Sensitivity and specificity of the diagnosis of the threatened preterm birth in the comparison group were significantly lower: 89.29% and 87.80% respectively. The diagnostic accuracy was 88.41%.
Patients in all clinical groups at 26 weeks had low AC/DC values. However, later in women of group I was found highest AC/DC level. AC/DC values in patients with risk of preterm birth were reduced compared with controls (p < 0.05). Sensitivity and specificity of the diagnosis of fetal distress if AC/DC was used were 91.67% and 99.12% respectively. Diagnostic accuracy of the test was 98.41%. In the comparison group the sensitivity was 77.78%, and the specificity was 89.22%. Diagnostic accuracy of intranatal cardiotocography was 86.82%.
Conclusions. Uterine activity extracted from the maternal abdominal signal can significantly improve the diagnosis of the threatenedpreterm delivery. Patients at risk of preterm birth have a delayed fetal neurological, which leads to fetal distress. AC/DC variables obtained via fetal non-invasive electrocardiography allow increasing the accuracy of fetal distress diagnosis.
... For example, Czabanski et al. [21] predict neonatal academia using a weighted fuzzy scoring (WFS) with SVM and obtained 92% accuracy and 88% quality index. Fanelli et al. [22] introduced phase-rectified-signal-average nonlinear-parameter for the quantitative assessment of fetal abnormality and achieved 75% AUC. Comert et al. [26] applied a neural network and achieved 92.40% accuracy, 95.89% sensitivity, and 74.75% specificity. ...
Fetal heart rate (FHR) is used to monitor the fetal state by obstetricians as a screening tool. Common guidelines for visual interpretation of FHR signals results in significant subjective variability due to the fetal physiological dynamics complexity. Automated diagnostic technology can assist obstetricians in medical decisions based on artificial intelligence and also can be an automatic diagnostic tool for primary health care centres and remote areas. This work presents a machine learning-based automated diagnostic tool for classification and diagnosis of Fetal Acidosis using FHR. A 1D-CNN model has been proposed because of its ability to automatically diagnose Fetal Acidosis into healthy or pathological conditions with high accuracy. To make the method robust and to improve accuracy with the artefacts present in the signal, the signal pre-processing is performed before training and classification. The accuracy was evaluated on a comprehensive dataset and achieved 99.09% for the diagnosis of Fetal Acidosis. Low-cost electronic hardware integrated with the proposed methodology can perform in real-time and can achieve high accuracy and reliability. This method can be used to support the expert decision and as an automatic stand-alone diagnostic tool that can assist the obstetricians in the early diagnosis of fetal acidosis.
... The last nonlinear technique is Phase Rectified Signal Averaging (PRSA) (Bauer et al., 2006). Average Acceleration and Deceleration Capacities (AC and DC) are among the various parameters which can be derived from the PRSA curve (Fanelli et al., 2013). More recently, Deceleration Reserve (DR) (Rivolta et al., 2019) was defined as the simple summation of AC and DC and it has been shown to achieve enhanced performance in detecting fetal hypoxia compared to AC and DC standalone parameters in the context of intrapartum FHR recordings. ...
... The impairment in ANS maturation was consistently found in this investigation by various quantitative CTG-derived parameters which have been extensively associated with the fetal ANS modulation throughout pregnancy as standalone features (Signorini et al., 2003;Gonçalves et al., 2018). In comparison with previous machine learning-derived and univariate results by our group in different populations of early IUGR Fanelli et al., 2013;Signorini et al., 2020b), it is possible to observe a consistent discriminative power of features LZC, HF_pow, and LTI. Specifically, the average value of each feature was greater in the control group vs. late IUGR fetuses. ...
... Lastly, in the described late IUGR population, short scale (T s 5 and T s 9) PRSA-extracted features were characterized by a greater discriminative power compared to global ones (T 40 and s 1). The reported findings are in accordance with the univariate results and support the hypothesis of an impaired fetal beat-to-beat responsiveness regulation in the context of nutrient restriction and chronic hypoxemia (Fanelli et al., 2013;Rivolta et al., 2019). Lastly, toward enhancing the general applicability as well as interpretability of the proposed RBF-SVM model, we tested its performance once excluding the information of fetal sex. ...
Late intrauterine growth restriction (IUGR) is a fetal pathological condition characterized by chronic hypoxia secondary to placental insufficiency, resulting in an abnormal rate of fetal growth. This pathology has been associated with increased fetal and neonatal morbidity and mortality. In standard clinical practice, late IUGR diagnosis can only be suspected in the third trimester and ultimately confirmed at birth. This study presents a radial basis function support vector machine (RBF-SVM) classification based on quantitative features extracted from fetal heart rate (FHR) signals acquired using routine cardiotocography (CTG) in a population of 160 healthy and 102 late IUGR fetuses. First, the individual performance of each time, frequency, and nonlinear feature was tested. To improve the unsatisfactory results of univariate analysis we firstly adopted a Recursive Feature Elimination approach to select the best subset of FHR-based parameters contributing to the discrimination of healthy vs. late IUGR fetuses. A fine tuning of the RBF-SVM model parameters resulted in a satisfactory classification performance in the training set (accuracy 0.93, sensitivity 0.93, specificity 0.84). Comparable results were obtained when applying the model on a totally independent testing set. This investigation supports the use of a multivariate approach for the in utero identification of late IUGR condition based on quantitative FHR features encompassing different domains. The proposed model allows describing the relationships among features beyond the traditional linear approaches, thus improving the classification performance. This framework has the potential to be proposed as a screening tool for the identification of late IUGR fetuses.
... Initially, for the feature extraction stage, methods for automatically replicating the guidelines issued by the International Federation of Obstetricians and Gynecologists (FIGO) [4] were employed. Soon after, new features brought in from other mature fields were adopted and tested, e.g., time-domain features, frequency-domain features, time-frequency-domain features, etc. [9][10][11][12][13][14][15][16][17][18]. ...
This paper presents an exploratory approach of the fetal heart rate (FHR) analysis, aiming to highlight potential limitations of the current predictive modeling attempts. To do so, a set of features that are usually encountered in FHR analysis as well as features extracted using a variant of symbolic aggregate approximation were projected onto a lower-dimensional space where patterns can easily be discerned. The results show, both in a qualitative and a quantitative manner, that there is high overlap between the classes that are formed using solely the umbilical cord pH information, irrespective of the selected dimensionality reduction method. These findings suggest that there is probably a limit to the performance expectation of the current pH-based systems and that alternative approaches should be also pursued to enhance the utility of computer-based decision support technologies.
... When ultrasound sensors are used in CTG devices, fetal heart rate values are obtained from the autocorrelation function of the fetal heart movement signal. In autocorrelation, a sliding window of 1.2 s is generally shifted across the signal and the average heart rate values obtained from these windows are stored in a buffer and read out four times per second (every 250 ms, 4Hz) (Signorini, Magenes et al. 2003, Fanelli, Magenes et al. 2013. Established routine CTG characteristics are baseline heart rate, accelerations, and decelerations. ...
Objective:
Fetal heart rate variability (HRV) is predestinated for monitoring fetal developmental disturbances. Only expensive magnetocardiography (fMCG) allows the precise recording of the individual fetal heart beat intervals uncovering also highly frequent vagal modulation. In contrast, fetal transabdominal electrocardiography (fECG) suffers from noise overlaying the fetal cardiac signal. Cardiotocography is the clinical method of choice, however, based on Doppler ultrasound, improper to resolve single beats concisely. The present work addresses the transferability of established electrophysiological HRV indices to CTG recordings during the fetal maturation period of 20 to 40 weeks of gestation.
Approach:
We compared (i) HRV indices obtained from fMCG, CTG and fECG of short- and long-term amplitude fluctuations and complexity, and (ii) their diagnostic value for identifying maturational age, fetal growth restriction (FGR) and small for gestational age (SGA). We used the functional brain age score (fABAS) and categories of long- and short-term regulation and complexity.
Main results:
Integrating all substudies we found: (1) indices related to long-term regulation, and with modified meaning and values of short-term regulation and sympathovagal balance according to electrophysiological HRV standards can be obtained from CTG. (2) Models using HRV indices calculated from CTG allow the identification of maturational age and discriminate FGR from controls with almost similar precision as electrophysiological means. (3) A modified set of HRV parameters containing short- and long-term regulation and long-term/short-term ratio appeared to be most suitable to describe autonomic developmental state when CTG data is used.
Significance:
Whereas the predominantly vagally modulated beat-to-beat precise high frequencies of HRV are not assessable from CTG, we identified relevant related HRV indices and categories for CTG recordings with diagnostic potential. They require further evaluation and confirmation with respect to any issues of fetal developmental and perinatal problems in subsequent studies. This methodology significantly extends the measures of established CTG devices.
