A Machine Learning Approach to Monitor the Emergence of Late Intrauterine Growth Restriction

Texture-Based Classification of Fetal Growth Restriction From Intrauterine Neurosonographic Image

OBJECTIVE

Fetal growth restriction (FGR) is a condition where fetuses fail to reach their genetic potential for growth, posing a significant health challenge for newborns. The aim of this research was to explore the efficacy of texture-based analysis of neurosonographic images in identifying FGR in fetuses, which may provide a promising tool for early assessment of FGR.

METHODS

A retrospective analysis collected 100 intrauterine neurosonographic images from 50 FGR and 50 gestational age-appropriate fetuses. Using MaZda software, approximately 300 texture features were extracted from occipital white matter (OWM) and cerebellum of intrauterine neurosonographic images, respectively. Then 10 optimal features were separately selected by 3 algorithms, including the Fisher coefficient method, the method of minimizing classification error probability and average correlation coefficients, and the mutual information coefficient method. Further, the 10 statistically most significant features were selected from these sets to form the mixed feature set. After nonlinear discriminant analysis was performed to reduce feature dimensionality, the artificial neural network (ANN) classifier was conducted, respectively.

RESULTS

For OWM and cerebellum, a total of 11 and 14 statistically significant features were selected. When the mixed feature sets of OWM and cerebellum were applied to ANN classifier, classification accuracy were 90.00% (κ = 0.800; P < .001) and 93.00% (κ = 0.860; P < .001), and the receiver operating characteristic curve for identifying FGR showed an area under the curve of 0.82 and 0.87.

CONCLUSIONS

Texture analysis of fetal intrauterine neurosonographic images is a feasible and noninvasive strategy for evaluating FGR fetuses.

Prediction of low birth weight from fetal ultrasound and clinical characteristics: a comparative study between a low- and middle-income and a high-income country

INTRODUCTION

Adverse perinatal outcomes (APO) pose a significant global challenge, particularly in low- and middle-income countries (LMICs). This study aims to analyse two cohorts of high-risk pregnant women for APO to comprehend risk factors and improve prediction accuracy.

METHODS

We considered an LMIC and a high-income country (HIC) population to derive XGBoost classifiers to predict low birth weight (LBW) from a comprehensive set of maternal and fetal characteristics including socio-demographic, past and current pregnancy information, fetal biometry and fetoplacental Doppler measurements. Data were sourced from the FeDoC (Fetal Doppler Collaborative) study (Pakistan, LMIC) and theIMPACT (Improving Mothers for a Better PrenAtal Care Trial) study (Spain, HIC), and included 520 and 746 pregnancies assessed from 28 weeks gestation, respectively. The models were trained on varying subsets of the mentioned characteristics to evaluate their contribution in predicting LBW cases. For external validation, and to highlight potential differential risk factors for LBW, we investigated the generalisation of these models across cohorts. Models' performance was evaluated through the area under the curve (AUC), and their interpretability was assessed using SHapley Additive exPlanations.

RESULTS

In FeDoC, Doppler variables demonstrated the highest value at predicting LBW compared with biometry and maternal clinical data (AUCDoppler, 0.67; AUCClinical, 0.65; AUCBiometry, 0.63), and its combination with maternal clinical data yielded the best prediction (AUCClinical+Doppler, 0.71). In IMPACT, fetal biometry emerged as the most predictive set (AUCBiometry, 0.75; AUCDoppler, 0.70; AUCClinical, 0.69) and its combination with Doppler and maternal clinical data achieved the highest accuracy (AUCClinical+Biometry+Doppler, 0.81). External validation consistently indicated that biometry combined with Doppler data yielded the best prediction.

CONCLUSIONS

Our findings provide new insights into the predictive role of different clinical and ultrasound descriptors in two populations at high risk for APO, highlighting that different approaches are required for different populations. However, Doppler data improves prediction capabilities in both settings, underscoring the value of standardising ultrasound data acquisition, as practiced in HIC, to enhance LBW prediction in LMIC. This alignment contributes to bridging the health equity gap.

Fetal heart rate spectral analysis in raw signals and PRSA-derived curve: normal and pathological fetuses discrimination

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.

Hybrid-FHR: a multi-modal AI approach for automated fetal acidosis diagnosis

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.

Intelligent antepartum fetal monitoring via deep learning and fusion of cardiotocographic signals and clinical data

Purpose

Cardiotocography (CTG), which measures uterine contraction (UC) and fetal heart rate (FHR), is a crucial tool for assessing fetal health during pregnancy. However, traditional computerized cardiotocography (cCTG) approaches have non-negligible calibration errors in feature extraction and heavily rely on the expertise and prior experience to define diagnostic features from CTG or FHR signals. Although previous works have studied deep learning methods for extracting CTG or FHR features, these methods still neglect the clinical information of pregnant women.

Methods

In this paper, we proposed a multimodal deep learning architecture (MMDLA) for intelligent antepartum fetal monitoring that is capable of performing automatic CTG feature extraction, fusion with clinical data and classification. The multimodal feature fusion was achieved by concatenating high-level CTG features, which were extracted from preprocessed CTG signals via a convolution neural network (CNN) with six convolution layers and five fully connected layers, and the clinical data of pregnant women. Eventually, light gradient boosting machine (LGBM) was implemented as fetal status assessment classifier. The effectiveness of MMDLA was evaluated using a dataset of 16,355 cases, each of which includes FHR signal, UC signal and pertinent clinical data like maternal age and gestational age.

Results

With an accuracy of 90.77% and an area under the curve (AUC) value of 0.9201, the multimodal features performed admirably. The data imbalance issue was also effectively resolved by the LGBM classifier, with a normal-F1 value of 0.9376 and an abnormal-F1 value of 0.8223.

Conclusion

In summary, the proposed MMDLA is conducive to the realization of intelligent antepartum fetal monitoring.

Prediction Models for Intrauterine Growth Restriction Using Artificial Intelligence and Machine Learning: A Systematic Review and Meta-Analysis

BACKGROUND

IntraUterine Growth Restriction (IUGR) is a global public health concern and has major implications for neonatal health. The early diagnosis of this condition is crucial for obtaining positive outcomes for the newborn. In recent years Artificial intelligence (AI) and machine learning (ML) techniques are being used to identify risk factors and provide early prediction of IUGR. We performed a systematic review (SR) and meta-analysis (MA) aimed to evaluate the use and performance of AI/ML models in detecting fetuses at risk of IUGR.

METHODS

We conducted a systematic review according to the PRISMA checklist. We searched for studies in all the principal medical databases (MEDLINE, EMBASE, CINAHL, Scopus, Web of Science, and Cochrane). To assess the quality of the studies we used the JBI and CASP tools. We performed a meta-analysis of the diagnostic test accuracy, along with the calculation of the pooled principal measures.

RESULTS

We included 20 studies reporting the use of AI/ML models for the prediction of IUGR. Out of these, 10 studies were used for the quantitative meta-analysis. The most common input variable to predict IUGR was the fetal heart rate variability (n = 8, 40%), followed by the biochemical or biological markers (n = 5, 25%), DNA profiling data (n = 2, 10%), Doppler indices (n = 3, 15%), MRI data (n = 1, 5%), and physiological, clinical, or socioeconomic data (n = 1, 5%). Overall, we found that AI/ML techniques could be effective in predicting and identifying fetuses at risk for IUGR during pregnancy with the following pooled overall diagnostic performance: sensitivity = 0.84 (95% CI 0.80-0.88), specificity = 0.87 (95% CI 0.83-0.90), positive predictive value = 0.78 (95% CI 0.68-0.86), negative predictive value = 0.91 (95% CI 0.86-0.94) and diagnostic odds ratio = 30.97 (95% CI 19.34-49.59). In detail, the RF-SVM (Random Forest-Support Vector Machine) model (with 97% accuracy) showed the best results in predicting IUGR from FHR parameters derived from CTG.

CONCLUSIONS

our findings showed that AI/ML could be part of a more accurate and cost-effective screening method for IUGR and be of help in optimizing pregnancy outcomes. However, before the introduction into clinical daily practice, an appropriate algorithmic improvement and refinement is needed, and the importance of quality assessment and uniform diagnostic criteria should be further emphasized.

RT-TRAQ: An algorithm for real-time tracking of faint quasi-periodic signals in noisy time series

We present an algorithm for live tracking of quasi-periodic faint signals in non-stationary, noisy, and phase-desynchronized time series measurements that commonly arise in embedded applications, such as wearable health monitoring. The first step of Rt-Traq is to continuously select fixed-length windows based on the rise or fall of data values in the stream. Subsequently, Rt-Traq calculates an averaged representative window, and its spectrum, whose frequency peaks reveal the underlying quasi-periodic signals. As each new data sample comes in, Rt-Traq incrementally updates the spectrum, to continuously track the signals through time. We develop several alternate implementations of the proposed algorithm. We evaluate their performance in tracking maternal and fetal heart rate using non-invasive photoplethysmography (PPG) data collected by a wearable device from animal experiments as well as a number of pregnant women who participated in our study. Our empirical results demonstrate improvements compared to competing approaches. We also analyze the memory requirement and complexity trade-offs between the implementations, which impact their demand on platform resources for real-time operation.

Fetal growth restriction and stillbirth: Biomarkers for identifying at risk fetuses

Fetal growth restriction (FGR) is a major cause of stillbirth, prematurity and impaired neurodevelopment. Its etiology is multifactorial, but many cases are related to impaired placental development and dysfunction, with reduced nutrient and oxygen supply. The fetus has a remarkable ability to respond to hypoxic challenges and mounts protective adaptations to match growth to reduced nutrient availability. However, with progressive placental dysfunction, chronic hypoxia may progress to a level where fetus can no longer adapt, or there may be superimposed acute hypoxic events. Improving detection and effective monitoring of progression is critical for the management of complicated pregnancies to balance the risk of worsening fetal oxygen deprivation in utero, against the consequences of iatrogenic preterm birth. Current surveillance modalities include frequent fetal Doppler ultrasound, and fetal heart rate monitoring. However, nearly half of FGR cases are not detected in utero, and conventional surveillance does not prevent a high proportion of stillbirths. We review diagnostic challenges and limitations in current screening and monitoring practices and discuss potential ways to better identify FGR, and, critically, to identify the “tipping point” when a chronically hypoxic fetus is at risk of progressive acidosis and stillbirth.

Explainable Computational Intelligence Model for Antepartum Fetal Monitoring to Predict the Risk of IUGR

Intrauterine Growth Restriction (IUGR) is a restriction of the fetus that involves the abnormal growth rate of the fetus, and it has a huge impact on the new-born’s health. Machine learning (ML) algorithms can help in early prediction and discrimination of the abnormality of the fetus’ health to assist in reducing the risk during the antepartum period. Therefore, in this study, Random Forest (RF), Support Vector Machine (SVM), K Nearest Neighbor (KNN) and Gradient Boosting (GB) was utilized to discriminate whether a fetus was healthy or suffering from IUGR based on the fetal heart rate (FHR). The Recursive Feature Elimination (RFE) method was used to select the significant feature for the classification of fetus. Furthermore, the study Explainable Artificial Intelligence (EAI) was implemented using LIME and SHAP to generate the explanation and to add comprehensibility in the proposed models. The experimental results indicate that RF achieved the highest accuracy (0.97) and F1-score (0.98) with the reduced set of features. However, the SVM outperformed it in terms of Positive Predictive Value (PPV) and specificity (SP). The performance of the model was further validated using another dataset and found that it outperformed the baseline studies for both the datasets. The proposed model can aid doctors in monitoring fetal health and enhancing the prediction process.

Artificial intelligence in obstetrics

This study reviews recent advances on the application of artificial intelligence for the early diagnosis of various maternal-fetal conditions such as preterm birth and abnormal fetal growth. It is found in this study that various machine learning methods have been successfully employed for different kinds of data capture with regard to early diagnosis of maternal-fetal conditions. With the more popular use of artificial intelligence, ethical issues should also be considered accordingly.

Effects of Prenatal Exposure to Alcohol and Smoking on Fetal Heart Rate and Movement Regulation

Negative associations of prenatal tobacco and alcohol exposure (PTE and PAE) on birth outcomes and childhood development have been well documented, but less is known about underlying mechanisms. A possible pathway for the adverse fetal outcomes associated with PTE and PAE is the alteration of fetal autonomic nervous system development. This study assessed PTE and PAE effects on measures of fetal autonomic regulation, as quantified by heart rate (HR), heart rate variability (SD-HR), movement, and HR-movement coupling in a population of fetuses at ≥ 34 weeks gestational age. Participants are a subset of the Safe Passage Study, a prospective cohort study that enrolled pregnant women from clinical sites in Cape Town, South Africa, and the Northern Plains region, United States. PAE was defined by six levels: no alcohol, low quit early, high quit early, low continuous, moderate continuous, and high continuous; while PTE by 4 levels: no smoking, quit early, low continuous, and moderate/high continuous. Linear regression analyses of autonomic measures were employed controlling for fetal sex, gestational age at assessment, site, maternal education, household crowding, and depression. Analyses were also stratified by sleep state (1F and 2F) and site (South Africa, N = 4025, Northern Plains, N = 2466). The final sample included 6491 maternal-fetal-dyad assessed in the third trimester [35.21 ± 1.26 (mean ± SD) weeks gestation]. PTE was associated with a decrease in mean HR in state 2F, in a dose dependent fashion, only for fetuses of mothers who continued smoking after the first trimester. In state 1F, there was a significant increase in mean HR in fetuses whose mother quit during the first trimester. This effect was driven by the Norther Plains cohort. PTE was also associated with a significant reduction in fetal movement in the most highly exposed group. In South Africa a significant increase in mean HR both for the high quit early and the high continuous group was observed. In conclusion, this investigation addresses a critical knowledge gap regarding the relationship between PTE and PAE and fetal autonomic regulation. We believe these results can contribute to elucidating mechanisms underlying risk for adverse outcomes.