Computer analysis of foetal monitoring signals

Intrapartum electronic fetal heart rate monitoring to predict acidemia at birth with the use of deep learning

BACKGROUND

Electronic fetal monitoring is used in most US hospital births but has significant limitations in achieving its intended goal of preventing intrapartum hypoxic-ischemic injury. Novel deep learning techniques can improve complex data processing and pattern recognition in medicine.

OBJECTIVE

This study aimed to apply deep learning approaches to develop and validate a model to predict fetal acidemia from electronic fetal monitoring data.

STUDY DESIGN

The database was created using intrapartum electronic fetal monitoring data from 2006 to 2020 from a large, multisite academic health system. Data were divided into training and testing sets with equal distribution of acidemic cases. Several different deep learning architectures were explored. The primary outcome was umbilical artery acidemia, which was investigated at 4 clinically meaningful thresholds: 7.20, 7.15, 7.10, and 7.05, along with base excess. The receiver operating characteristic curves were generated with the area under the receiver operating characteristic assessed to determine the performance of the models. External validation was performed using a publicly available Czech database of electronic fetal monitoring data.

RESULTS

A total of 124,777 electronic fetal monitoring files were available, of which 77,132 had <30% missingness in the last 60 minutes of the electronic fetal monitoring tracing. Of these, 21,041 were matched to a corresponding umbilical cord gas result, of which 10,182 were time-stamped within 30 minutes of the last electronic fetal monitoring reading and composed the final dataset. The prevalence rates of the outcomes in the data were 20.9% with a pH of <7.2, 9.1% with a pH of <7.15, 3.3% with a pH of <7.10, and 1.3% with a pH of <7.05. The best performing model achieved an area under the receiver operating characteristic of 0.85 at a pH threshold of <7.05. When predicting the joint outcome of both pH of <7.05 and base excess of less than -10 meq/L, an area under the receiver operating characteristic of 0.89 was achieved. When predicting both pH of <7.20 and base excess of less than -10 meq/L, an area under the receiver operating characteristic of 0.87 was achieved. At a pH of <7.15 and a positive predictive value of 30%, the model achieved a sensitivity of 90% and a specificity of 48%.

CONCLUSION

The application of deep learning methods to intrapartum electronic fetal monitoring analysis achieves promising performance in predicting fetal acidemia. This technology could help improve the accuracy and consistency of electronic fetal monitoring interpretation.

Introducing Artificial Intelligence in Interpretation of Foetal Cardiotocography: Medical Dataset Curation and Preliminary Coding-An Interdisciplinary Project

Artificial intelligence (AI) is gaining increasing interest in the field of medicine because of its capacity to process big data and pattern recognition. Cardiotocography (CTG) is widely used for the assessment of foetal well-being and uterine contractions during pregnancy and labour. It is characterised by inter- and intraobserver variability in interpretation, which depends on the observers' experience. Artificial intelligence (AI)-assisted interpretation could improve its quality and, thus, intrapartal care. Cardiotocography (CTG) raw signals from labouring women were extracted from the database at the University Hospital of Bern between 2006 and 2019. Later, they were matched with the corresponding foetal outcomes, namely arterial umbilical cord pH and 5-min APGAR score. Excluded were deliveries where data were incomplete, as well as multiple births. Clinical data were grouped regarding foetal pH and APGAR score at 5 min after delivery. Physiological foetal pH was defined as 7.15 and above, and a 5-min APGAR score was considered physiologic when reaching ≥7. With these groups, the algorithm was trained to predict foetal hypoxia. Raw data from 19,399 CTG recordings could be exported. This was accomplished by manually searching the patient's identification numbers (PIDs) and extracting the corresponding raw data from each episode. For some patients, only one episode per pregnancy could be found, whereas for others, up to ten episodes were available. Initially, 3400 corresponding clinical outcomes were found for the 19,399 CTGs (17.52%). Due to the small size, this dataset was rejected, and a new search strategy was elaborated. After further matching and curation, 6141 (31.65%) paired data samples could be extracted (cardiotocography raw data and corresponding maternal and foetal outcomes). Of these, half will be used to train artificial intelligence (AI) algorithms, whereas the other half will be used for analysis of efficacy. Complete data could only be found for one-third of the available population. Yet, to our knowledge, this is the most exhaustive and second-largest cardiotocography database worldwide, which can be used for computer analysis and programming. A further enrichment of the database is planned.

A CNN-RNN unified framework for intrapartum cardiotocograph classification

BACKGROUND AND OBJECTIVE

Prenatal fetal monitoring, which can monitor the growth and health of the fetus, is very vital for pregnant women before delivery. During pregnancy, it is crucial to judge whether the fetus is abnormal, which helps obstetricians carry out early intervention to avoid fetal hypoxia and even death. At present, clinical fetal monitoring widely used fetal heart rate monitoring equipment. Fetal heart rate and uterine contraction signals obtained by fetal heart monitoring equipment are important information to evaluate fetal health status.

METHODS

This paper is based on 1D-CNN (One Dimension Convolutional Neural Network) and GRU (Gate Recurrent Unit). We preprocess the obtained data and enhances them, to make the proportion of number of instances in different class in the training set is same.

RESULTS

In model performance evaluation, standard evaluation indicators are used, such as accuracy, sensitivity, specificity, and ROC (receiver operating characteristic). Finally, the accuracy of our model in the test set is 95.15%, the sensitivity is 96.20%, and the specificity is 94.09%.

CONCLUSIONS

In fetal heart rate monitoring, this paper proposes a 1D-CNN and bidirectional GRU hybrid models, and the fetal heart rate and uterine contraction signals given by monitoring are used as input feature to classify the fetal health status. The results show that our approach is effective in evaluating fetal health status and can assists obstetricians in clinical decision-making. And provide a baseline for the introduction of 1D-CNN and bidirectional GRU hybrid models into the evaluation of fetal health status.

Computerized cardiotocography analysis during labor - A state-of-the-art review

Cardiotocography is defined as the recording of fetal heart rate and uterine contractions and is widely used during labor as a screening tool to determine fetal wellbeing. The visual interpretation of the cardiotocography signals by the practitioners, following common guidelines, is subject to a high interobserver variability, and the efficiency of cardiotocography monitoring is still debated. Since the 1990s, researchers and practitioners work on designing reliable computer-aided systems to assist practitioners in cardiotocography interpretation during labor. Several systems are integrated in the monitoring devices, mostly based on the guidelines, but they have not clearly demonstrated yet their usefulness. In the last decade, the availability of large clinical databases as well as the emergence of machine learning and deep learning methods in healthcare has led to a surge of studies applying those methods to cardiotocography signals analysis. The state-of-the-art systems perform well to detect fetal hypoxia when evaluated on retrospective cohorts, but several challenges remain to be tackled before they can be used in clinical practice. First, the development and sharing of large, open and anonymized multicentric databases of perinatal and cardiotocography data during labor is required to build more accurate systems. Also, the systems must produce interpretable indicators along with the prediction of the risk of fetal hypoxia in order to be appropriated and trusted by practitioners. Finally, common standards should be built and agreed on to evaluate and compare those systems on retrospective cohorts and to validate their use in clinical practice.

A deep feature fusion network for fetal state assessment

CTG (cardiotocography) has consistently been used to diagnose fetal hypoxia. It is susceptible to identifying the average fetal acid-base balance but lacks specificity in recognizing prenatal acidosis and neurological impairment. CTG plays a vital role in intrapartum fetal state assessment, which can prevent severe organ damage if fetal hypoxia is detected earlier. In this paper, we propose a novel deep feature fusion network (DFFN) for fetal state assessment. First, we extract spatial and temporal information from the fetal heart rate (FHR) signal using a multiscale CNN-BiLSTM network, increasing the features' diversity. Second, the multiscale CNN-BiLSM network and frequently used features are integrated into the deep learning model. The proposed DFFN model combines different features to improve classification accuracy. The multiscale convolutional kernels can identify specific essential information and consider signal's temporal information. The proposed method achieves 61.97%, 73.82%, and 66.93% of sensitivity, specificity, and quality index, respectively, on the public CTU-UHB database. The proposed method achieves the highest QI on the private database, verifying the proposed method's effectiveness and generalization. The proposed DFFN combines the advantages of feature engineering and deep learning models and achieves competitive accuracy in fetal state assessment compared with related works.

On designing a biosignal-based fetal state assessment system: A systematic mapping study

BACKGROUND AND OBJECTIVE

The patterns present in biosignals, such as fetal heart rate (FHR), are valuable indicators of fetal well-being. In designing biosignal analysis systems, the variety of approaches and technology usage impairs the decision-making for the fundamental units of the systems. There is a need for an updated overview of studies encompassing the biosignal-based fetal state assessment systems. Therefore, we propose a systematic mapping study to identify and synthesize recent research regarding the building blocks that compose these systems.

METHODS

We followed well-established guidelines to perform a systematic mapping of studies regarding the building-blocks that compose the fetal state assessment systems and published between January 2016 and January 2021. A search string was determined based on the mapping questions and the PI (population and intervention) divisions. The search string was applied in digital libraries covering the fields of computer science, engineering, and medical informatics. Then, we applied the forward snowballing technique to complement the resulting set. This process resulted in 75 primary studies selected from a total of 871 papers.

RESULTS

Selected studies were classified according to the publication types, systems design stages, datasets, and predictive capabilities. The results revealed that (i) The majority of the selected studies refer to the method as a type of publication and there is a lack of validation studies; (ii) The CTU-UHB was the most frequent biosignal-based dataset and UCI-CTG was the most frequent feature-based data; (iii) The selected studies made use of the system design stages alone or in a mixed-mode.

CONCLUSION

The results indicated that the well-established classification models achieved competitive results compared with the state-of-the-art methods in data-constrained system designs. Moreover, we identified the need for validation studies in the clinical environment.

Machine Learning Algorithm to Predict Acidemia Using Electronic Fetal Monitoring Recording Parameters

Background: Electronic fetal monitoring (EFM) is the universal method for the surveillance of fetal well-being in intrapartum. Our objective was to predict acidemia from fetal heart signal features using machine learning algorithms. Methods: A case–control 1:2 study was carried out compromising 378 infants, born in the Miguel Servet University Hospital, Spain. Neonatal acidemia was defined as pH < 7.10. Using EFM recording logistic regression, random forest and neural networks models were built to predict acidemia. Validation of models was performed by means of discrimination, calibration, and clinical utility. Results: Best performance was attained using a random forest model built with 100 trees. The discrimination ability was good, with an area under the Receiver Operating Characteristic curve (AUC) of 0.865. The calibration showed a slight overestimation of acidemia occurrence for probabilities above 0.4. The clinical utility showed that for 33% cutoff point, missing 5% of acidotic cases, 46% of unnecessary cesarean sections could be prevented. Logistic regression and neural networks showed similar discrimination ability but with worse calibration and clinical utility. Conclusions: The combination of the variables extracted from EFM recording provided a predictive model of acidemia that showed good accuracy and provides a practical tool to prevent unnecessary cesarean sections.

Fetal Heart Rate Fragmentation

Intrapartum fetal monitoring's primary goal is to avoid adverse perinatal outcomes related to hypoxia/acidosis without increasing unnecessary interventions. Recently, a set of indices were proposed as new biomarkers to analyze heart rate (HR), termed HR fragmentation (HRF). In this work, the HRF indices were applied to intrapartum fetal heart rate (FHR) traces to evaluate fetal acidemia. The fragmentation method produces four indices: PIP-Percentage of inflection points; IALS-Inverse of the average length of acceleration/deceleration segments; PSS-Percentage of short segments; PAS-Percentage of alternating segments. On the other hand, the symbolic approach studied the existence of different patterns of length four. We applied the measures to 246 selected FHR recordings sampled at 4 and 2 Hz, where 39 presented umbilical artery's pH ≤ 7.15. When applied to the 4 Hz FHR, the PIP, IASL, and PSS showed significantly higher values in the traces from acidemic fetuses. In comparison, the percentage of "words"W 1 h andW 2 s showed lower values for those traces. Furthermore, when using the 2 Hz, only IASL, W 0, andW 2 m achieved significant differences between traces from both acidemic and normal fetuses. Notwithstanding, the ideal sampling frequency is yet to be established. The fragmentation indices correlated with Sisporto variability measures, especially short-term variability. Accordingly, the fragmentation indices seem to be able to detect pathological patterns in FHR tracings. These indices have the advantage of being suitable and straightforward to apply in real-time analysis. Future studies should combine these indexes with others used successfully to detect fetal hypoxia, improving the power of discrimination in a larger dataset.

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

CONTEXT

The 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.

OBJECTIVE

This 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.

METHODS

Article 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.

RESULTS

After 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).

DISCUSSION

This 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.

CONCLUSIONS

This 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.

A randomized controlled trial studying the effect of maternal hyperoxygenation on fetal heart rate in suspected fetal distress

OBJECTIVE

To investigate the effect of maternal hyperoxygenation on fetal heart rate (FHR) when applied for suspected fetal distress during the second stage of term labor.

APPROACH

A single-center randomized controlled trial was conducted in a tertiary care hospital in The Netherlands. Participants were included during the second stage of labor in case of an intermediary or abnormal FHR pattern. Patients were randomized to receive either 100% oxygen at 10 l/min until delivery, or conventional care without additional oxygen. The primary outcome was the change in FHR pattern before and after the onset of the study, measured as the change in depth and duration of FHR decelerations. Secondary outcome measures were features based on phase-rectified signal averaging (PRSA), baseline assignability, and deceleration characteristics of the FHR pattern.

MAIN RESULTS

Between March 2016 and April 2018, 117 women were included. The FHR pattern could be analyzed for 71 participants, the other 46 women delivered before the end of the post time-frame. A 2.3% reduction in depth and duration of FHR decelerations was found after maternal hyperoxygenation, compared to a 10% increase in the control group (p = 0.24). Maternal hyperoxygenation had a significantly positive effect on PRSA metrics, with a decrease in PRSA-acceleration capacity (p = 0.03) and PRSA-deceleration capacity (p = 0.02) in the intervention group compared to the control group.

SIGNIFICANCE

The difference in depth and duration of decelerations after the start of the study was not significantly different between both study groups. A statistically significant positive effect on PRSA-deceleration capacity and PRSA-acceleration capacity was found after maternal hyperoxygenation, which might be associated with a positive effect on neonatal outcome.

DeepFHR: intelligent prediction of fetal Acidemia using fetal heart rate signals based on convolutional neural network

Background

Fetal heart rate (FHR) monitoring is a screening tool used by obstetricians to evaluate the fetal state. Because of the complexity and non-linearity, a visual interpretation of FHR signals using common guidelines usually results in significant subjective inter-observer and intra-observer variability. Objective: Therefore, computer aided diagnosis (CAD) systems based on advanced artificial intelligence (AI) technology have recently been developed to assist obstetricians in making objective medical decisions.

Methods

In this work, we present an 8-layer deep convolutional neural network (CNN) framework to automatically predict fetal acidemia. After signal preprocessing, the input 2-dimensional (2D) images are obtained using the continuous wavelet transform (CWT), which provides a better way to observe and capture the hidden characteristic information of the FHR signals in both the time and frequency domains. Unlike the conventional machine learning (ML) approaches, this work does not require the execution of complex feature engineering, i.e., feature extraction and selection. In fact, 2D CNN model can self-learn useful features from the input data with the prerequisite of not losing informative features, representing the tremendous advantage of deep learning (DL) over ML.

Results

Based on the test open-access database (CTU-UHB), after comprehensive experimentation, we achieved better classification performance using the optimal CNN configuration compared to other state-of-the-art methods: the averaged ten-fold cross-validation of the accuracy, sensitivity, specificity, quality index defined as the geometric mean of the sensitivity and specificity, and the area under the curve yielded results of 98.34, 98.22, 94.87, 96.53 and 97.82%, respectively

Conclusions

Once the proposed CNN model is successfully trained, the corresponding CAD system can be served as an effective tool to predict fetal asphyxia objectively and accurately.

Independent Analysis of Decelerations and Resting Periods through CEEMDAN and Spectral-Based Feature Extraction Improves Cardiotocographic Assessment

Fetal monitoring is commonly based on the joint recording of the fetal heart rate (FHR) and uterine contraction signals obtained with a cardiotocograph (CTG). Unfortunately, CTG analysis is difficult, and the interpretation problems are mainly associated with the analysis of FHR decelerations. From that perspective, several approaches have been proposed to improve its analysis; however, the results obtained are not satisfactory enough for their implementation in clinical practice. Current clinical research indicates that a correct CTG assessment requires a good understanding of the fetal compensatory mechanisms. In previous works, we have shown that the complete ensemble empirical mode decomposition with adaptive noise, in combination with time-varying autoregressive modeling, may be useful for the analysis of those characteristics. In this work, based on this methodology, we propose to analyze the FHR deceleration episodes separately. The main hypothesis is that the proposed feature extraction strategy applied separately to the complete signal, deceleration episodes, and resting periods (between contractions), improves the CTG classification performance compared with the analysis of only the complete signal. Results reveal that by considering the complete signal, the classification performance achieved 81.7% quality. Then, including information extracted from resting periods, it improved to 83.2%.

Validation of a computerized algorithm to quantify fetal heart rate deceleration area

INTRODUCTION

Reliability in visual cardiotocography interpretation is unsatisfying, which has led to the development of computerized cardiotocography. Computerized analysis is well established for antenatal fetal surveillance but has yet not performed sufficiently during labor. We aimed to investigate the capacity of a new computerized algorithm compared with visual assessment in identifying intrapartum fetal heart rate baseline and decelerations.

MATERIAL AND METHODS

In all, 312 intrapartum cardiotocography tracings with variable decelerations were analyzed by the computerized algorithm and visually examined by two observers, blinded to each other and the computer analysis. The width, depth and area of each deceleration was measured. Four cases (>100 variable decelerations) were subjected to in-depth detailed analysis. The outcome measures were bias in seconds (width), beats per minute (depth), and beats (area) between computer and observers using Bland-Altman analysis. Interobserver reliability was determined by calculating intraclass correlation and Spearman rank analysis.

RESULTS

The analysis (312 cases) showed excellent intraclass correlation (0.89-0.95) and very strong Spearman correlation (0.82-0.91). The detailed analysis of >100 decelerations in four cases revealed low bias between the computer and the two observers; width 1.4 and 1.4 seconds, depth 5.1 and 0.7 beats per minute, and area 0.1 and -1.7 beats. This was comparable to the bias between the two observers: 0.3 seconds (width), 4.4 beats per minute (depth) and 1.7 beats (area). The intraclass correlation was excellent (0.90-.98).

CONCLUSION

A novel computerized algorithm for intrapartum cardiotocography analysis is as accurate as gold standard visual assessment, with high correlation and low bias.

Machine learning ensemble modelling to classify caesarean section and vaginal delivery types using Cardiotocography traces

Human visual inspection of Cardiotocography traces is used to monitor the foetus during labour and avoid neonatal mortality and morbidity. The problem, however, is that visual interpretation of Cardiotocography traces is subject to high inter and intra observer variability. Incorrect decisions, caused by miss-interpretation, can lead to adverse perinatal outcomes and in severe cases death. This study presents a review of human Cardiotocography trace interpretation and argues that machine learning, used as a decision support system by obstetricians and midwives, may provide an objective measure alongside normal practices. This will help to increase predictive capacity and reduce negative outcomes. A robust methodology is presented for feature set engineering using an open database comprising 552 intrapartum recordings. State-of-the-art in signal processing techniques is applied to raw Cardiotocography foetal heart rate traces to extract 13 features. Those with low discriminative capacity are removed using Recursive Feature Elimination. The dataset is imbalanced with significant differences between the prior probabilities of both normal deliveries and those delivered by caesarean section. This issue is addressed by oversampling the training instances using a synthetic minority oversampling technique to provide a balanced class distribution. Several simple, yet powerful, machine-learning algorithms are trained, using the feature set, and their performance is evaluated with real test data. The results are encouraging using an ensemble classifier comprising Fishers Linear Discriminant Analysis, Random Forest and Support Vector Machine classifiers, with 87% (95% Confidence Interval: 86%, 88%) for Sensitivity, 90% (95% CI: 89%, 91%) for Specificity, and 96% (95% CI: 96%, 97%) for the Area Under the Curve, with a 9% (95% CI: 9%, 10%) Mean Square Error.

SisPorto 4.0 - computer analysis following the 2015 FIGO Guidelines for intrapartum fetal monitoring

SisPorto 4.0 is the most recent version of a program for the computer analysis of cardiotocographic (CTG) signals and ST events, which has been adapted to the 2015 International Federation of Gynaecology and Obstetrics (FIGO) guidelines for intrapartum foetal monitoring. This paper provides a detailed description of the analysis performed by the system, including the signal-processing algorithms involved in identification of basic CTG features and the resulting real-time alerts.