Cardiotocography Classification Research Articles

DANNMCTG: Domain-Adversarial Training of Neural Network for multicenter antenatal cardiotocography signal classification

Intelligent classification of cardiotocography (CTG) based on machine learning (ML), a useful tool to improve the accuracy of fetal abnormality detection, can assist obstetricians with clinical decisions. With the advancement of information technologies and medical devices, there are development opportunities for multicenter clinical research and obtaining more digital CTG signals. However, most of the existing clinical multicenter CTG datasets are partially annotated and have discrepancies which do not satisfy the ML condition of independent identical distribution. Therefore, this paper focuses on an unsupervised domain adaptation (UDA) algorithm to realize cross-domain intelligent classification of multimodal CTG signals. We propose a method dubbed domain adversarial training of neural network for multicenter CTG (DANNMCTG), which mainly consists of a label classifier, a feature extractor and a domain discriminator. To match different distribution of fetal heart rate (FHR), uterine contraction (UC) and fetal movement (FetMov) signals, we condition the domain alignment on label predictions by defining the multi-linear map. For analysis, two datasets from the hospital central station and home monitoring devices were considered as the source and target domains. The results showed that the accuracy value, F1 value and area under the curve (AUC) value of the DANNMCTG were 71.25%, 76.08% and 0.7705, respectively. This method significantly improved the performance of the deep learning models without exploiting any information in the target domain, and outperformed the state-of-the-art UDA algorithms for CTG classification. In summary, the DANNMCTG can effectively mitigate the influence of domain shift for multicenter intelligent prenatal fetal monitoring.

A hybrid stacked ensemble and Kernel SHAP-based model for intelligent cardiotocography classification and interpretability

BackgroundIntelligent cardiotocography (CTG) classification can assist obstetricians in evaluating fetal health. However, high classification performance is often achieved by complex machine learning (ML)-based models, which causes interpretability concerns. The trade-off between accuracy and interpretability makes it challenging for most existing ML-based CTG classification models to popularize in prenatal clinical applications.MethodsAiming to improve CTG classification performance and prediction interpretability, a hybrid model was proposed using a stacked ensemble strategy with mixed features and Kernel SHapley Additive exPlanations (SHAP) framework. Firstly, the stacked ensemble classifier was established by employing support vector machines (SVM), extreme gradient boosting (XGB), and random forests (RF) as base learners, and backpropagation (BP) as a meta learner whose input was mixed with the CTG features and the probability value of each category output by base learners. Then, the public and private CTG datasets were used to verify the discriminative performance. Furthermore, Kernel SHAP was applied to estimate the contribution values of features and their relationships to the fetal states.ResultsFor intelligent CTG classification using 10-fold cross-validation, the accuracy and average F1 score were 0.9539 and 0.9249 in the public dataset, respectively; and those were 0.9201 and 0.8926 in the private dataset, respectively. For interpretability, the explanation results indicated that accelerations (AC) and the percentage of time with abnormal short-term variability (ASTV) were the key determinants. Specifically, the probability of abnormality increased and that of the normal state decreased as the value of ASTV grew. In addition, the likelihood of the normal status rose with the increase of AC.ConclusionsThe proposed model has high classification performance and reasonable interpretability for intelligent fetal monitoring.

Identifying fetal status with fetal heart rate: Deep learning approach based on long convolution

CTG (Cardiotocography) is an effective tool for fetal status assessment. Clinically, doctors mainly evaluate the health of fetus by observing FHR (fetal heart rate). The rapid development of Artificial Intelligence has led realization of computer-aided CTG technology, Intelligent CTG classification based on FHR is a fundamental component of these technologies. Its implementation can provide doctors with auxiliary decisions. Most of existing FHR classification methods are based on combing different deep learning models, such as CNN (Convolutional Neural Network), LSTM (Long short-term memory) and Transformer. However, these studies ignore the balance of positive and negative samples in dataset and the matching degree between model and FHR classification task, which reduces the classification accuracy. In this paper, we mainly discuss two major problems in previous FHR classification studies: reduce class imbalance and select appropriate convolution kernel. To address above two problems, we propose a data augmentation method based on ECMN (Edge Clipping and Multiscale Noise) to resolve class imbalance. Subsequently, we introduce a one-dimensional long convolutional layer, which use trend area to calculate the appropriate convolution kernel. Based on appropriate convolution kernel, an improved residual structure with attention mechanism named TGLCN (Trend-Guided Long Convolution Network) is proposed to improve FHR classification accuracy. Finally, horizontal and longitudinal experiments show that the TGLCN obtains high classification accuracy and speed of parameter adjustment.

Detection of Suspicious Cardiotocographic Recordings by Means of a Machine Learning Classifier.

Cardiotocography (CTG) is one of the fundamental prenatal diagnostic methods for both antepartum and intrapartum fetal surveillance. Although it has allowed a significant reduction in intrapartum and neonatal mortality and morbidity, its diagnostic accuracy is, however, still far from being fully satisfactory. In particular, the identification of uncertain and suspicious CTG traces remains a challenging task for gynecologists. The introduction of computerized analysis systems has enabled more objective evaluations, possibly leading to more accurate diagnoses. In this work, the problem of classifying suspicious CTG recordings was addressed through a machine learning approach. A machine-based labeling was proposed, and a binary classification was carried out using a support vector machine (SVM) classifier to distinguish between suspicious and normal CTG traces. The best classification metrics showed accuracy, sensitivity, and specificity values of 92%, 92%, and 90%, respectively. The main results were compared both with results obtained by considering a more unbalanced dataset and with relevant literature studies in the field. The use of the SVM proved to be promising in the field of CTG classification. However, appropriate feature selection and dataset balancing are crucial to achieve satisfactory performance of the classifier.

Multimodal learning for fetal distress diagnosis using a multimodal medical information fusion framework

Cardiotocography (CTG) monitoring is an important medical diagnostic tool for fetal well-being evaluation in late pregnancy. In this regard, intelligent CTG classification based on Fetal Heart Rate (FHR) signals is a challenging research area that can assist obstetricians in making clinical decisions, thereby improving the efficiency and accuracy of pregnancy management. Most existing methods focus on one specific modality, that is, they only detect one type of modality and inevitably have limitations such as incomplete or redundant source domain feature extraction, and poor repeatability. This study focuses on modeling multimodal learning for Fetal Distress Diagnosis (FDD); however, exists three major challenges: unaligned multimodalities; failure to learn and fuse the causality and inclusion between multimodal biomedical data; modality sensitivity, that is, difficulty in implementing a task in the absence of modalities. To address these three issues, we propose a Multimodal Medical Information Fusion framework named MMIF, where the Category Constrained-Parallel ViT model (CCPViT) was first proposed to explore multimodal learning tasks and address the misalignment between multimodalities. Based on CCPViT, a cross-attention-based image-text joint component is introduced to establish a Multimodal Representation Alignment Network model (MRAN), explore the deep-level interactive representation between cross-modal data, and assist multimodal learning. Furthermore, we designed a simple-structured FDD test model based on the highly modal alignment MMIF, realizing task delegation from multimodal model training (image and text) to unimodal pathological diagnosis (image). Extensive experiments, including model parameter sensitivity analysis, cross-modal alignment assessment, and pathological diagnostic accuracy evaluation, were conducted to show our models' superior performance and effectiveness.

Implementation of a revised classification for intrapartum fetal heart rate monitoring and association to birth outcome: Anational cohort study.

A revised intrapartum cardiotocography (CTG) classification was introduced in Sweden in 2017. The aims of the revision were to adapt to the international guideline published in 2015 and to adjust the classification of CTG patterns to current evidence regarding intrapartum fetal physiology. This study aimed to investigate adverse neonatal outcomes before and after implementation of the revised CTG classification. A before-and-after design was used. Cohort I (n=160210) included births from June 1, 2014 through May 31, 2016 using the former CTG classification, and cohort II (n=166558) included births from June 1, 2018 through May 31, 2020 with the revised classification. Data were collected from the Swedish Pregnancy and Neonatal Registers. The primary outcome was moderate to severe neonatal hypoxic ischemic encephalopathy (HIE 2-3). Secondary outcomes were birth acidemia (umbilical artery pH <7.05 and base excess < -12mmol/L or pH <7.00), A-criteria for neonatal hypothermia treatment, 5-min Apgar scores <4 and <7, neonatal seizures, meconium aspiration, neonatal mortality and delivery mode. Logistic regression was used (period II vs period I), and results are presented as adjusted odds ratios (aORs) with 95% confidence intervals (95% CIs). There were no statistically significant differences in HIE 2-3 (aOR 1.27; 95% CI 0.97-1.66), proportion of neonates meeting A-criteria for hypothermia treatment (aOR 0.96; 95% CI 0.89-1.04) or neonatal mortality (aOR 0.68; 95% CI 0.39-1.18) between the cohorts. Birth acidemia (aOR 1.36; 95% CI 1.25-1.48), 5-min Apgar scores <7 (aOR 1.27; 95% CI 1.18-1.36) and <4 (aOR 1.40; 95% CI 1.17-1.66) occurred more often in cohort II. The absolute risk difference for HIE 2-3 was 0.02% (95% CI 0.00-0.04). Operative delivery (vacuum or cesarean) rates were lower in cohort II (aOR 0.82; 95% CI 0.80-0.85 and aOR 0.94; 95% CI 0.91-0.97, respectively). Although not statistically significant, a small increase in the incidence of HIE 2-3 after implementation of the revised CTG classification cannot be excluded. Operative deliveries were fewer but incidences of acidemia and low Apgar scores were higher in the latter cohort. This warrants further in-depth analyses before a full re-evaluation of the revised classification can be made.

A systematic review of automated pre-processing, feature extraction and classification of cardiotocography.

ContextThe interpretations of cardiotocography (CTG) tracings are indeed vital to monitor fetal well-being both during pregnancy and childbirth. Currently, many studies are focusing on feature extraction and CTG classification using computer vision approach in determining the most accurate diagnosis as well as monitoring the fetal well-being during pregnancy. Additionally, a fetal monitoring system would be able to perform detection and precise quantification of fetal heart rate patterns.ObjectiveThis study aimed to perform a systematic review to describe the achievements made by the researchers, summarizing findings that have been found by previous researchers in feature extraction and CTG classification, to determine criteria and evaluation methods to the taxonomies of the proposed literature in the CTG field and to distinguish aspects from relevant research in the field of CTG.MethodsArticle search was done systematically using three databases: IEEE Xplore digital library, Science Direct, and Web of Science over a period of 5 years. The literature in the medical sciences and engineering was included in the search selection to provide a broader understanding for researchers.ResultsAfter screening 372 articles, and based on our protocol of exclusion and inclusion criteria, for the final set of articles, 50 articles were obtained. The research literature taxonomy was divided into four stages. The first stage discussed the proposed method which presented steps and algorithms in the pre-processing stage, feature extraction and classification as well as their use in CTG (20/50 papers). The second stage included the development of a system specifically on automatic feature extraction and CTG classification (7/50 papers). The third stage consisted of reviews and survey articles on automatic feature extraction and CTG classification (3/50 papers). The last stage discussed evaluation and comparative studies to determine the best method for extracting and classifying features with comparisons based on a set of criteria (20/50 articles).DiscussionThis study focused more on literature compared to techniques or methods. Also, this study conducts research and identification of various types of datasets used in surveys from publicly available, private, and commercial datasets. To analyze the results, researchers evaluated independent datasets using different techniques.ConclusionsThis systematic review contributes to understand and have insight into the relevant research in the field of CTG by surveying and classifying pertinent research efforts. This review will help to address the current research opportunities, problems and challenges, motivations, recommendations related to feature extraction and CTG classification, as well as the measurement of various performance and various data sets used by other researchers.

Intelligent classification of antepartum cardiotocography model based on deep forest

Intelligent classification of antepartum cardiotocography (CTG) can assist obstetricians to make clinical decisions, which is helpful to improve the accuracy of fetal abnormality detection. However, most of the existing machine learning based classification models fail to meet the clinical requirements. In this work, a method is proposed to improve the classification accuracy by using Deep Forest (DF) algorithm. Principal component analysis and visualization were adopted to mine the distribution characteristics of the CTG data. After data preprocessing, deep forest multi-granularity scanning phase was used to explore the connection between the case characteristic. Then the cascade forest phase, which was designed to integrate Random Forest (RF), Weighted Random Forest (WRF), Completely Random Forest (CRF) and Gradient Boosting Decision Tree (GBDT) as the basic classifiers, performed deep iterations and finally got the best performance model. Compared with the traditional machine learning models, deep neural network and the state-of-art CTG classification models, the results show that the accuracy value, average F1 value and Area Under the Curve (AUC) value were 92.64 %, 92.01 % and 0.990 respectively in the external public data set, and were 91.64 %, 88.92 % and 0.9493 respectively in the internal private data set, which were the most excellent among all comparison models. In conclusion, the proposed DF model is effective and feasible, and has a good application prospect in the intelligent assessment of antenatal fetal health status.

Quantification of Posterior Risk Related to Intrapartum FIGO 2015 Criteria for Cardiotocography in the Second Stage of Labor

Introduction: Intrapartum cardiotocography (CTG) was used for several decades to detect a stressed fetus so that delivery can be expedited to prevent birth asphyxia. The main aim of the study was to calculate the risk of neonatal acidemia (pH ≤ 7.10) according to duration of the 2nd stage of labor and occurrence of the International Federation of Gynecology and Obstetrics (FIGO) 2015 CTG classification parameters. Materials and Methods: This was a retrospective case-control study on 552 pregnancies receiving continuous CTG monitoring in labor and immediate hemogasanalysis at birth. Cases with umbilical artery (UA) pH ≤ 7.10 and controls with UA pH ≥ 7.10 were matched for parity and gestational age at delivery, with ratio 1:5. Logistic regression analysis, adjusted for the expected risk in the general population, was used to calculate the baseline risk of UA pH ≤ 7.10 in the absence of any CTG pathological feature and those associated with pathological CTG patterns occurring in the 2nd stage according to FIGO 2015. Results: Seventy-three cases and 387 controls reached 2nd stage and were included in the analysis. For those reaching 2nd stage, the mean adjusted risk of acidemia associated with nonpathological CTG was 1.6%. Stratification of risk according to duration of the 2nd stage yielded risks of neonatal acidemia of 1.23, 2.08, 5.81, and 15.22% at 30, 60, 120, and 180 min, respectively. Bradycardia >10 min was associated with risk of neonatal acidemia of 9.9 and 15.8% for 2nd-stage durations of 30 and 60 min, respectively. Risks associated with 1 prolonged deceleration >5 min were 6.80, 11.08, 27.0, and 51.0% at 30, 60, 120, and 180 min, respectively. Repetitive late or prolonged decelerations >30 min were associated with risk of neonatal acidemia of 2.43, 4.14, 11.17, and 26.45% at 30, 60, 120, and 180 min, respectively. Conclusion: The risk of neonatal acidemia is directly proportional to duration of the 2nd stage, irrespective of the presence of CTG abnormalities, increasing 12-fold (1.2–15.3%) from 30 to 180 min. Occurrence of FIGO 2015 pathological CTG patterns showed a decreasing impact from bradycardia >10 min to decelerations >5 min, recurrent later or prolonged decelerations >30 min, and nonpathological CTG.

Impaired validity of the new FIGO and Swedish CTG classification templates to identify fetal acidosis in the first stage of labor

Introduction Cardiotocography (CTG) is the main method of intrapartum fetal surveillance. In 2015 a new guideline was introduced by the International Federation of Gynecology and Obstetrics (FIGO), FIGO-15. In Sweden it was adjusted to SWE-17, replacing the previous national template, SWE-09. This study, conducted at one university hospital and one regional hospital in southern Sweden, evaluated the diagnostic validity of these three templates to detect fetal acidosis during the first stage of labor. Material and methods A total of 73 neonates with pH <7.1 in umbilical cord artery or vein at cesarean delivery during the first stage of labor were identified retrospectively. For each acidotic neonate, three non-acidemic neonates, with a pH ≥7.2 in cord artery and vein, and Apgar scores ≥9 at five and ten minutes, in all 219 neonates, were selected. The CTG tracings before birth in acidemic neonates, and tracings at the same cervical dilatation in the non-acidemic neonates, were independently assessed by three professionals from the obstetric staff, blinded to group and clinical data. Based on their categorizations of the included variables (baseline, variability, accelerations, decelerations and contraction rate), each CTG tracing was systematically classified according to the three templates. The sensitivity and specificity to identify acidemia by the classification pathological were determined for each template. Interobserver agreement in the assessments of tracings as pathological or not was analyzed, using free-marginal Kappa index. Results The sensitivity for patterns classified as pathological to identify acidemia was similar for FIGO-15 (71%) and SWE-17 (77%, p = .13), and the specificity was 97% for both. SWE-09 had a significantly higher sensitivity (95%, p < .001) albeit with a lower specificity (90%, p < .001) than the other two templates. Among acidemic neonates, the fraction of tracings classified as normal was higher with SWE-17 (9.6%) than with SWE-09 (0%; p = .01) and FIGO-15 (1.4%; p = .06). For tracings from neonates with acidemia, agreement for three independent assessors was strong (κ 0.85) with SWE-09, and weak for FIGO-15 (κ 0.47), and SWE-17 (κ 0.51). For tracings from neonates without acidemia, the agreement was almost perfect for FIGO-15 (κ 0.91), strong withSWE-17 (κ 0.90) and moderate with SWE-09 (κ 0.78). Conclusions The ability of FIGO-15 and SWE-17 to identify fetal acidosis is considered insufficient. The combination of a high sensitivity and a high specificity makes SWE-09 the most discriminatory template during the first stage of labor.

Cardiotocography alone is outdated and ST analysis is the way forward in fetal monitoring: FOR: Does the use of ST analysis in conjunction with cardiotocography improve perinatal outcome and/or reduce interventions for fetal distress?

Well, according to European randomised controlled trials (RCTs) the answer should be – at least partly – yes: there is a lower rate of instrumental deliveries and lower need for fetal blood sampling, whereas several meta-analyses are inconclusive regarding the effect on umbilical artery metabolic acidosis (Olofsson et al. Acta Obstet Gynecol Scand 2014;93:571–86). A recent large RCT in the USA did not show any differences between the ST arm and the conventially monitored arm of the trial, however (Belfort et al. NEJM 2015;373:632–41). Differences between the US trial and the European trials include a three-tier instead of a four-tier cardiotocography (CTG) classification and an ST-action system, absence of fetal scalp sampling, and a very low institutional recruitment rate, despite the overall large size of the trial (on average nine inclusions per hospital per month, in contrast with 15–135 in the five European trials). The low inclusion rate of only two per week in large hospitals with many obstetricians and midwives working in the labour ward may have hampered adequate exposure to the new monitoring technique. Data from the European trials has shown that results in favour of ST analysis were more pronounced in the second half of the trials, which suggests a continuing learning process (Schuit AJOG 2013;209:394–5). Moreover, longitudinal data from several hospitals has shown that after the introduction of ST technology metabolic acidosis at birth fell over the course of several years by 50–75%, without an increase in instrumental deliveries (Visser et al. Acta Obstet Gynecol Scand 2014;93:539–43). The latter data suggest that an introduction of ST analysis in the labour ward may have resulted in a considerably better outcome than could have been expected from the results of the trials. The introduction of the new device may have resulted in more attention towards CTG interpretation, especially as structured CTG classification and training are essential parts of the ST package. The much better outcome in the conventional arm of the Dutch RCT than was expected from earlier data may also point towards an improved CTG interpretation (i.e. the ‘Hawthorne effect’). Finally, the automatic detection and depiction of (significant) ST changes forces the clinician to assess the CTG pattern and ST events to take action or otherwise. In other words, in the case of a possible deterioration of the fetal condition the clinician is obliged to make a documented decision. From medical legal cases it is known that inadequate fetal monitoring is the main factor leading to financial compensation. The Royal College of Obstetricians and Gynaecologists (RCOG) project ‘Each baby counts’ is currently reviewing cases that resulted in a poor fetal outcome during labour. It is our opinion that the prevention of poor intrapartum outcomes will primarily be achieved by prioritising the 24/7 presence of experienced clinicians, case discussions, and continuous training on the labour ward. And what about ST technology? We think it is likely to be of additional help, mainly because it requires structured CTG assessment and decision-making, with some additional benefits of the technology itself. None declared. Completed disclosure of interests form available to view online as supporting information. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Inter‐ and intra‐observer agreement of non‐reassuring cardiotocography analysis and subsequent clinical management

To quantify inter- and intra-observer agreement of non-reassuring intrapartum cardiotocography (CTG) patterns and subsequent clinical management. Methodological study. University Medical Center. CTG patterns of 79 women beyond 37weeks of gestation with a singleton fetus in vertex position in first stage of labor in whom fetal blood sampling (FBS) had been performed. Nine observers assessed CTG patterns, which were formerly clinically classified as non-reassuring and indicative for FBS, according to the guidelines of the International Federation of Gynecology and Obstetrics modified for ST analysis. They also proposed clinical management strategies without and with insight into clinical parameters. Weighted kappa values (κw ) and proportions of agreement (Pa ) were calculated. Agreement on CTG classification and clinical management. Inter-observer agreement on CTG classification and on clinical management were poor for most observer categories (κw range 0.31-0.50 and 0.20-0.45, respectively). Observers agreed best on abnormal CTG patterns (Pa range 0.28-0.36) and on the clinical management option "continue monitoring" (Pa range 0.32-0.40). Intra-observer agreement was fair to good for most observers (κw 0.33-0.70). Insight into clinical parameters resulted in similar inter- and intra-observer agreement. There was poor inter-observer agreement and fair to good intra-observer agreement on classification and clinical management of intrapartum CTG patterns, which had been classified as non-reassuring and indicative for FBS during birth.

ST Analysis of the Fetal ECG, as an Adjunct to Fetal Heart Rate Monitoring in Labour: A Review.

ST Analysis of the Fetal ECG, as an Adjunct to Fetal Heart Rate Monitoring in Labour: A Review.

Interobserver agreement in the assessment of intrapartum automated fetal electrocardiography in singleton pregnancies

To study interobserver agreement in the assessment of intrapartum automated fetal electrocardiogram ST interval analysis (STAN). Observational study. Labor ward in tertiary level university hospital. Sample. Two hundred (140 reassuring and 60 non-reassuring) STAN recordings on non-selected women with singleton, vertex, term pregnancies were selected from our archive. Samples of 60-min were analysed from the end of each recording, excluding the last 30 min before delivery. Three consultants, who had undergone STAN training and had clinical experience in using STAN, reviewed the recordings using cardiotocography (CTG) and ST information with no clinical data. The reviewers were asked to follow STAN guidelines and 1 to classify the CTG as normal, intermediary, abnormal, or preterminal, and 2 to make a clinical decision on labor management. Interobserver agreement evaluated by weighted kappa (kappa(w)) values and the proportion of agreement. In CTG classification, the interobserver agreement between three observers was moderate (kappa(w), 0.47-0.48). The proportion of agreement was 56-59%. In clinical decision-making, kappa(w) values varied from 0.47 to 0.60, and the proportion of agreement was 80-86%. In non-selected term pregnancies, the interobserver agreement among experienced obstetricians in the classification of CTG and clinical decision-making according to STAN guidelines is moderate at best.

Utilisation en routine de l’analyse du segment ST de l’électrocardiogramme fœtal pour la surveillance du travail. Une année d’expérience (résultats préliminaires)

Objectives. – Two randomized trials found that monitoring labor by analyzing the fetal electrocardiogram (STAN™) appears to offer better sensitivity and specificity in screening for metabolic acidosis than cardiotocography (CTG) does. We report here the results of a descriptive study of 173 cases between 1 June 2001 and 31 May 2002, the objectives of which were twofold: to determine the number of possible false negatives (defined by neonatal metabolic acidosis with a pH less than 7.05 and base deficit (BD) greater than 12 mmol/l, in the absence of a STAN™ decisions event more than 30 min before birth), and to assess whether STAN™ indicated the need for operative intervention too late in cases of fetal distress. Patients and methods. – Our study was conducted on an at-risk population with CTG changes. STAN™ was used with a scalp electrode to monitor labor according to a protocol based upon the FIGO CTG classification and the clinical instructions used in the Swedish trial. Accordingly, we compared two groups: in group 1, a STAN™ decision event required a rapid conclusion to labor, while in group 2, the absence of any STAN™ decision event meant that labor continued, despite the changes in the tracing. Results. – We found seven cases with a neonatal pH less than 7.05, but only one may involve a false negative by this method. The mean arterial pH and the number of fetuses born with an arterial pH less than 7.10 did not differ significantly between the two groups, nor were there any significant differences in the mean BD at birth, or the number of children with a BD greater than 12 mmol/l. Conclusion. – The results appear to indicate that use of STAN™ with the usual CTG enables an appropriate response to cases of fetal distress, but it remains difficult to determine the real benefits that can be expected from this method in France. Longer assessment in our population is required before we can assess its actual advantages.