Multiclass Fetal Abnormality Detection Using Ensemble Deep Learning Techniques

Abstract

Classifying fetal cardiotocography data is essential in the efficient prenatal risk assessment due to its potential for identifying errors or abnormalities during pregnancy. Traditional fetal heart rate (FHR) analysis frameworks, which unfortunately still rely on manual interpretation of results, subsequently lead to the inefficient use of human resources and sometimes require more time for abnormality detection. With the implementation of Machine Learning (ML) algorithms, automatic analysis and early detection of abnormalities are now possible.  The model’s performance is directly influenced by the retrieval of features and the optimal management of class imbalance in the dataset. In this regard, we introduce a feature-based innovative strategy for multi-class classification in fetal cardiograph datasets based on feature importance analysis. The proposed model utilizes Random Forest (RF) for feature extraction, which employs two distinct target importance analyses: 1. class imbalance, and 2. class weights. In Phase 1, an artificial neural network and an improved TabNet model were utilised for classifying three classes: Normal, Suspect, and Pathology (NSP), with SMOTE balancing. In Phase 2, we identify the features of classes that contribute to NSP classification, and we consider nine additional features based on class weight for various cardiotography features, such as baseline, ASTV, ALTV, etc. In Phase 2, NSP classification is performed by including class 1-9 features (A, B, C, D, E, AD, DE…) and assigning class weights.   Using our proposed ensemble deep learning model, the accuracy of prediction is improved.  The RF model retrieves primary features from the fetal cardiograph, and complex relationships among these features enhance the representation of information. The next step is the classification stage, which applies an attention-based deep learning model, TabNet. Due to the nature of the TABNET model in handling tabular data, it can selectively focus on relevant features while ensuring explainability. The proposed model is evaluated using different performance metrics for two novel feature importance analyses. The RF+TabNet+LSTM achieves a maximum accuracy of 97% with SMOTE in NSP target classification (phase 1), while including Class weight in class1-9 features, the model achieves classification accuracy of 92% (phase II) and proves the importance of features contributing to prediction and classification.  All code and the curated dataset for Multiclass Fetal Abnormality Detection are available at https://github.com/rrramyaresea/Multiclass-Fetal-Abnormality-Detection, enabling the reproducibility of our findings.