PatchCTG: Patch Cardiotocography Transformer for Antepartum Fetal Health Monitoring

TL;DR

PatchCTG is introduced, a transformer-based model specifically designed for CTG analysis, employing patch-based tokenisation, instance normalisation and channel-independent processing to capture essential local and global temporal dependencies within CTG signals.

Abstract

Antepartum Cardiotocography (CTG) is vital for fetal health monitoring, but traditional methods like the Dawes-Redman system are often limited by high inter-observer variability, leading to inconsistent interpretations and potential misdiagnoses. This paper introduces PatchCTG, a transformer-based model specifically designed for CTG analysis, employing patch-based tokenisation, instance normalisation and channel-independent processing to capture essential local and global temporal dependencies within CTG signals. PatchCTG was evaluated on the Oxford Maternity (OXMAT) dataset, comprising over 20,000 CTG traces across diverse clinical outcomes after applying the inclusion and exclusion criteria. With extensive hyperparameter optimisation, PatchCTG achieved an AUC of 77%, with specificity of 88% and sensitivity of 57% at Youden's index threshold, demonstrating adaptability to various clinical needs. Testing across varying temporal thresholds showed robust predictive performance, particularly with finetuning on data closer to delivery, achieving a sensitivity of 52% and specificity of 88% for near-delivery cases. These findings suggest the potential of PatchCTG to enhance clinical decision-making in antepartum care by providing a reliable, objective tool for fetal health assessment. The source code is available at https://github.com/jaleedkhan/PatchCTG.

Figures (6)

• Figure 1: Training and validation results of the PatchCTG model on the complete dataset with optimised hyperparameters, showing convergence to approximately 77% AUC and consistent model performance throughout training.
• Figure 2: The performance of PatchCTG in terms of AUC with varying thresholds of days to delivery (1-7) for the APO cohort, indicating a gradual improvement in AUC as the temporal threshold expands.
• Figure 3: Performance of PatchCTG model trained and tested on the complete dataset with cases recorded up to 7 days prior to delivery, evaluated using sensitivity, specificity, PPV, NPV, F1 score, and accuracy for different classification thresholds.
• Figure 4: Results of PatchCTG model trained on a subset of data with cases recorded 3-7 days prior to delivery and tested on a subset with cases recorded up to 2 days prior to delivery (AUC = 72.5%).
• Figure 5: Results of PatchCTG model pretrained on a subset of cases recorded 3-7 days before delivery, followed by finetuning and testing on cases recorded up to 2 days prior to delivery (AUC = 74.6%).