1
|
Li J, Li J, Guo C, Chen Q, Liu G, Li L, Luo X, Wei H. Multicentric intelligent cardiotocography signal interpretation using deep semi-supervised domain adaptation via minimax entropy and domain invariance. COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE 2024; 249:108145. [PMID: 38582038 DOI: 10.1016/j.cmpb.2024.108145] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Received: 11/09/2023] [Revised: 03/22/2024] [Accepted: 03/24/2024] [Indexed: 04/08/2024]
Abstract
BACKGROUND AND OBJECTIVE Obstetricians use Cardiotocography (CTG), which is the continuous recording of fetal heart rate and uterine contraction, to assess fetal health status. Deep learning models for intelligent fetal monitoring trained on extensively labeled and identically distributed CTG records have achieved excellent performance. However, creation of these training sets requires excessive time and specialist labor for the collection and annotation of CTG signals. Previous research has demonstrated that multicenter studies can improve model performance. However, models trained on cross-domain data may not generalize well to target domains due to variance in distribution among datasets. Hence, this paper conducted a multicenter study with Deep Semi-Supervised Domain Adaptation (DSSDA) for intelligent interpretation of antenatal CTG signals. This approach helps to align cross-domain distribution and transfer knowledge from a label-rich source domain to a label-scarce target domain. METHODS We proposed a DSSDA framework that integrated Minimax Entropy and Domain Invariance (DSSDA-MMEDI) to reduce inter-domain gaps and thus achieve domain invariance. The networks were developed using GoogLeNet to extract features from CTG signals, with fully connected, softmax layers for classification. We designed a Dynamic Gradient-driven strategy based on Mutual Information (DGMI) to unify the losses from Minimax Entropy (MME), Domain Invariance (DI), and supervised cross-entropy during iterative learning. RESULTS We validated our DSSDA model on two datasets collected from collaborating healthcare institutions and mobile terminals as the source and target domains, which contained 16,355 and 3,351 CTG signals, respectively. Compared to the results achieved with deep learning networks without DSSDA, DSSDA-MMEDI significantly improved sensitivity and F1-score by over 6%. DSSDA-MMEDI also outperformed other state-of-the-art DSSDA approaches for CTG signal interpretation. Ablation studies were performed to determine the unique contribution of each component in our DSSDA mechanism. CONCLUSIONS The proposed DSSDA-MMEDI is feasible and effective for alignment of cross-domain data and automated interpretation of multicentric antenatal CTG signals with minimal annotation cost.
Collapse
Affiliation(s)
- Jialu Li
- School of Medical Information Engineering, Guangzhou University of Chinese Medicine, Guangzhou, 510006, China
| | - Jun Li
- School of Medical Information Engineering, Guangzhou University of Chinese Medicine, Guangzhou, 510006, China; College of Computer Science and Software Engineering, Shenzhen University, Shenzhen, 518060, China
| | - Chenshuo Guo
- School of Medical Information Engineering, Guangzhou University of Chinese Medicine, Guangzhou, 510006, China
| | - Qinqun Chen
- School of Medical Information Engineering, Guangzhou University of Chinese Medicine, Guangzhou, 510006, China
| | - Guiqing Liu
- The First Affiliated Hospital of Guangzhou University of Chinese Medicine, Guangzhou, 510405, China
| | - Li Li
- Guangzhou Sunray Medical Apparatus Co. Ltd, Guangzhou, 510620, China; Tianhe District People's Hospital, First Affiliated Hospital of Jinan University, Guangzhou, 510630, China
| | - Xiaomu Luo
- School of Medical Information Engineering, Guangzhou University of Chinese Medicine, Guangzhou, 510006, China
| | - Hang Wei
- School of Medical Information Engineering, Guangzhou University of Chinese Medicine, Guangzhou, 510006, China; Intelligent Chinese Medicine Research Institute, Guangzhou University of Chinese Medicine, Guangzhou, 510006, China.
| |
Collapse
|
2
|
Bai J, Lu Y, Liu H, He F, Guo X. Editorial: New technologies improve maternal and newborn safety. FRONTIERS IN MEDICAL TECHNOLOGY 2024; 6:1372358. [PMID: 38872737 PMCID: PMC11169838 DOI: 10.3389/fmedt.2024.1372358] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Key Words] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/17/2024] [Accepted: 05/17/2024] [Indexed: 06/15/2024] Open
Affiliation(s)
- Jieyun Bai
- Guangdong Provincial Key Laboratory of Traditional Chinese Medicine Information Technology, Jinan University, Guangzhou, China
- Auckland Bioengineering Institute, University of Auckland, Auckland, New Zealand
| | - Yaosheng Lu
- Guangdong Provincial Key Laboratory of Traditional Chinese Medicine Information Technology, Jinan University, Guangzhou, China
| | - Huishu Liu
- Guangzhou Women and Children’s Medical Center, Guangzhou Medical University, Guangzhou, China
| | - Fang He
- Department of Obstetrics and Gynecology, Third Affiliated Hospital of Guangzhou Medical University, Guangzhou, China
| | - Xiaohui Guo
- Department of Obstetrics, Shenzhen People’s Hospital, Shenzhen, China
| |
Collapse
|
3
|
Cao Z, Wang G, Xu L, Li C, Hao Y, Chen Q, Li X, Liu G, Wei H. Intelligent antepartum fetal monitoring via deep learning and fusion of cardiotocographic signals and clinical data. Health Inf Sci Syst 2023; 11:16. [PMID: 36950107 PMCID: PMC10025176 DOI: 10.1007/s13755-023-00219-w] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/26/2022] [Accepted: 02/27/2023] [Indexed: 03/21/2023] Open
Abstract
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.
Collapse
Affiliation(s)
- Zhen Cao
- School of Medical Information Engineering, Guangzhou University of Chinese Medicine, Guangzhou, 510006 China
| | - Guoqiang Wang
- School of Medical Information Engineering, Guangzhou University of Chinese Medicine, Guangzhou, 510006 China
| | - Ling Xu
- School of Medical Information Engineering, Guangzhou University of Chinese Medicine, Guangzhou, 510006 China
| | - Chaowei Li
- School of Medical Information Engineering, Guangzhou University of Chinese Medicine, Guangzhou, 510006 China
- Nvogene Co., Ltd., Tianjing, China
| | - Yuexing Hao
- Department of Human Centered Design, Cornell University, Ithaca, NY USA
| | - Qinqun Chen
- School of Medical Information Engineering, Guangzhou University of Chinese Medicine, Guangzhou, 510006 China
| | - Xia Li
- Guangzhou Medical University Second Affiliated Hospital, Guangzhou, China
| | - Guiqing Liu
- The First Affiliated Hospital, Guangzhou University of Chinese Medicine, Guangzhou, China
| | - Hang Wei
- School of Medical Information Engineering, Guangzhou University of Chinese Medicine, Guangzhou, 510006 China
| |
Collapse
|
4
|
Kapila R, Saleti S. Optimizing fetal health prediction: Ensemble modeling with fusion of feature selection and extraction techniques for cardiotocography data. Comput Biol Chem 2023; 107:107973. [PMID: 37926049 DOI: 10.1016/j.compbiolchem.2023.107973] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/25/2023] [Revised: 09/12/2023] [Accepted: 10/19/2023] [Indexed: 11/07/2023]
Abstract
Cardiotocography (CTG) captured the fetal heart rate and the timing of uterine contractions. Throughout pregnancy, CTG intelligent categorization is crucial for monitoring fetal health and preserving proper fetal growth and development. Since CTG provides information on the fetal heartbeat and uterus contractions, which helps determine if the fetus is pathologic or not, obstetricians frequently use it to evaluate a child's physical health during pregnancy. In the past, obstetricians have artificially analyzed CTG data, which is time-consuming and inaccurate. So, developing a fetal health categorization model is crucial as it may help to speed up the diagnosis and treatment and conserve medical resources. The CTG dataset is used in this study. To diagnose the illness, 7 machine learning models are employed, as well as ensemble strategies including voting and stacking classifiers. In order to choose and extract the most significant and critical attributes from the dataset, Feature Selection (FS) techniques like ANOVA and Chi-square, as well as Feature Extraction (FE) strategies like Principal Component Analysis (PCA) and Independent Component Analysis (ICA), are being used. We used the Synthetic Minority Oversampling Technique (SMOTE) approach to balance the dataset because it is unbalanced. In order to forecast the illness, the top 5 models are selected, and these 5 models are used in ensemble methods such as voting and stacking classifiers. The utilization of Stacking Classifiers (SC), which involve Adaboost and Random Forest (RF) as meta-classifiers for disease detection. The performance of the proposed SC with meta-classifier as RF model, which incorporates Chi-square with PCA, outperformed all other state-of-the-art models, achieving scores of 98.79%,98.88%,98.69%,96.32%, and 98.77% for accuracy, precision, recall, specificity, and f1-score respectively.
Collapse
Affiliation(s)
- Ramdas Kapila
- Data Science Laboratory, Computer Science and Engineering, SRM University - AP, India.
| | - Sumalatha Saleti
- Data Science Laboratory, Computer Science and Engineering, SRM University - AP, India.
| |
Collapse
|
5
|
Feng J, Liang J, Qiang Z, Hao Y, Li X, Li L, Chen Q, Liu G, Wei H. A hybrid stacked ensemble and Kernel SHAP-based model for intelligent cardiotocography classification and interpretability. BMC Med Inform Decis Mak 2023; 23:273. [PMID: 38017460 PMCID: PMC10685618 DOI: 10.1186/s12911-023-02378-y] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/31/2022] [Accepted: 11/16/2023] [Indexed: 11/30/2023] Open
Abstract
BACKGROUND Intelligent 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. METHODS Aiming 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. RESULTS For 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. CONCLUSIONS The proposed model has high classification performance and reasonable interpretability for intelligent fetal monitoring.
Collapse
Affiliation(s)
- Junyuan Feng
- School of Medical Information Engineering, Guangzhou University of Chinese Medicine, Guangzhou, China
| | - Jincheng Liang
- School of Medical Information Engineering, Guangzhou University of Chinese Medicine, Guangzhou, China
| | - Zihan Qiang
- School of The Fifth Clinical Medical School, Guangzhou University of Chinese Medicine, Guangzhou, China
| | - Yuexing Hao
- Department of Human Centered Design, Cornell University, Ithaca, NY, USA
| | - Xia Li
- Third Affiliated Hospital, Guangzhou University of Chinese Medicine, Guangzhou, China
| | - Li Li
- Tianhe District People's Hospital, First Affiliated Hospital of Jinan University, Guangzhou, China
- Guangzhou Sunray Medical Apparatus Co. Ltd, Guangzhou, China
| | - Qinqun Chen
- School of Medical Information Engineering, Guangzhou University of Chinese Medicine, Guangzhou, China
| | - Guiqing Liu
- First Affiliated Hospital, Guangzhou University of Chinese Medicine, Guangzhou, China
| | - Hang Wei
- School of Medical Information Engineering, Guangzhou University of Chinese Medicine, Guangzhou, China.
- Intelligent Chinese Medicine Research Institute, Guangzhou University of Chinese Medicine, Guangzhou, China.
| |
Collapse
|
6
|
Asfaw D, Jordanov I, Impey L, Namburete A, Lee R, Georgieva A. Multimodal Deep Learning for Predicting Adverse Birth Outcomes Based on Early Labour Data. Bioengineering (Basel) 2023; 10:730. [PMID: 37370663 DOI: 10.3390/bioengineering10060730] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/02/2023] [Revised: 05/29/2023] [Accepted: 06/07/2023] [Indexed: 06/29/2023] Open
Abstract
Cardiotocography (CTG) is a widely used technique to monitor fetal heart rate (FHR) during labour and assess the health of the baby. However, visual interpretation of CTG signals is subjective and prone to error. Automated methods that mimic clinical guidelines have been developed, but they failed to improve detection of abnormal traces. This study aims to classify CTGs with and without severe compromise at birth using routinely collected CTGs from 51,449 births at term from the first 20 min of FHR recordings. Three 1D-CNN and LSTM based architectures are compared. We also transform the FHR signal into 2D images using time-frequency representation with a spectrogram and scalogram analysis, and subsequently, the 2D images are analysed using a 2D-CNNs. In the proposed multi-modal architecture, the 2D-CNN and the 1D-CNN-LSTM are connected in parallel. The models are evaluated in terms of partial area under the curve (PAUC) between 0-10% false-positive rate; and sensitivity at 95% specificity. The 1D-CNN-LSTM parallel architecture outperformed the other models, achieving a PAUC of 0.20 and sensitivity of 20% at 95% specificity. Our future work will focus on improving the classification performance by employing a larger dataset, analysing longer FHR traces, and incorporating clinical risk factors.
Collapse
Affiliation(s)
- Daniel Asfaw
- School of Computing, University of Portsmouth, Portsmouth PO1 3HE, UK
- Nuffield Department of Women's & Reproductive Health, University of Oxford, Oxford OX1 2JD, UK
| | - Ivan Jordanov
- School of Computing, University of Portsmouth, Portsmouth PO1 3HE, UK
| | - Lawrence Impey
- Nuffield Department of Women's & Reproductive Health, University of Oxford, Oxford OX1 2JD, UK
| | - Ana Namburete
- Department of Computer Science, University of Oxford, Oxford OX1 3QG, UK
| | - Raymond Lee
- Faculty of Technology, University of Portsmouth, Portsmouth PO1 2UP, UK
| | - Antoniya Georgieva
- Nuffield Department of Women's & Reproductive Health, University of Oxford, Oxford OX1 2JD, UK
| |
Collapse
|
7
|
Zhang Y, Deng Y, Zhou Z, Zhang X, Jiao P, Zhao Z. Multimodal learning for fetal distress diagnosis using a multimodal medical information fusion framework. Front Physiol 2022; 13:1021400. [DOI: 10.3389/fphys.2022.1021400] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/17/2022] [Accepted: 10/25/2022] [Indexed: 11/09/2022] Open
Abstract
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.
Collapse
|
8
|
Accessing Artificial Intelligence for Fetus Health Status Using Hybrid Deep Learning Algorithm (AlexNet-SVM) on Cardiotocographic Data. SENSORS 2022; 22:s22145103. [PMID: 35890783 PMCID: PMC9319518 DOI: 10.3390/s22145103] [Citation(s) in RCA: 12] [Impact Index Per Article: 6.0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Received: 04/13/2022] [Revised: 06/25/2022] [Accepted: 07/04/2022] [Indexed: 12/22/2022]
Abstract
Artificial intelligence is serving as an impetus in digital health, clinical support, and health informatics for an informed patient’s outcome. Previous studies only consider classification accuracies of cardiotocographic (CTG) datasets and disregard computational time, which is a relevant parameter in a clinical environment. This paper proposes a modified deep neural algorithm to classify untapped pathological and suspicious CTG recordings with the desired time complexity. In our newly developed classification algorithm, AlexNet architecture is merged with support vector machines (SVMs) at the fully connected layers to reduce time complexity. We used an open-source UCI (Machine Learning Repository) dataset of cardiotocographic (CTG) recordings. We divided 2126 CTG recordings into 3 classes (Normal, Pathological, and Suspected), including 23 attributes that were dynamically programmed and fed to our algorithm. We employed a deep transfer learning (TL) mechanism to transfer prelearned features to our model. To reduce time complexity, we implemented a strategy wherein layers in the convolutional base were partially trained to leave others in the frozen states. We used an ADAM optimizer for the optimization of hyperparameters. The presented algorithm also outperforms the leading architectures (RCNNs, ResNet, DenseNet, and GoogleNet) with respect to real-time accuracies, sensitivities, and specificities of 99.72%, 96.67%, and 99.6%, respectively, making it a viable candidate for clinical settings after real-time validation.
Collapse
|
9
|
Chen M, Yin Z. Classification of Cardiotocography Based on the Apriori Algorithm and Multi-Model Ensemble Classifier. Front Cell Dev Biol 2022; 10:888859. [PMID: 35646917 PMCID: PMC9130474 DOI: 10.3389/fcell.2022.888859] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/03/2022] [Accepted: 03/31/2022] [Indexed: 12/04/2022] Open
Abstract
Cardiotocography (CTG) recorded fetal heart rate and its temporal relationship with uterine contractions. CTG intelligent classification plays an important role in evaluating fetal health and protecting fetal normal growth and development throughout pregnancy. At the feature selection level, this study uses the Apriori algorithm to search frequent item sets for feature extraction. At the level of the classification model, the combination model of AdaBoost and random forest with the highest classification accuracy is finally selected by comparing various models. The suspicious class data in the CTG data set affect the overall classification accuracy. The number of suspicious class data is predicted by the multi-model ensemble method. Finally, the data set is fused from three classifications to two classifications. The classification accuracy is 0.976, and the AUC is 0.98, which significantly improves the classification effect. In conclusion, the method used in this study has high accuracy in model classification, which is helpful to improve the accuracy of fetal abnormality detection.
Collapse
Affiliation(s)
| | - Zhixiang Yin
- School of Mathematics, Physics and Statistics, Shanghai University of Engineering Science, Shanghai, China
| |
Collapse
|
10
|
FHRGAN: Generative adversarial networks for synthetic fetal heart rate signal generation in low-resource settings. Inf Sci (N Y) 2022. [DOI: 10.1016/j.ins.2022.01.070] [Citation(s) in RCA: 2] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 12/21/2022]
|
11
|
Qian X, Zhou Z, Hu J, Zhu J, Huang H, Dai Y. A comparative study of kernel-based vector machines with probabilistic outputs for medical diagnosis. Biocybern Biomed Eng 2021. [DOI: 10.1016/j.bbe.2021.09.003] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/20/2022]
|