Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2022 Jul;63(7):692-700.
doi: 10.3349/ymj.2022.63.7.692.

Machine Learning Model for Classifying the Results of Fetal Cardiotocography Conducted in High-Risk Pregnancies

Tae Jun Park #  1 , Hye Jin Chang #  2 , Byung Jin Choi  1 , Jung Ah Jung  2 , Seongwoo Kang  2 , Seokyoung Yoon  2 , Miran Kim  2   3 , Dukyong Yoon  4   5   6
Affiliations

Machine Learning Model for Classifying the Results of Fetal Cardiotocography Conducted in High-Risk Pregnancies

Tae Jun Park et al. Yonsei Med J. 2022 Jul.

Abstract

Purpose: Fetal well-being is usually assessed via fetal heart rate (FHR) monitoring during the antepartum period. However, the interpretation of FHR is a complex and subjective process with low reliability. This study developed a machine learning model that can classify fetal cardiotocography results as normal or abnormal.

Materials and methods: In total, 17492 fetal cardiotocography results were obtained from Ajou University Hospital and 100 fetal cardiotocography results from Czech Technical University and University Hospital in Brno. Board-certified physicians then reviewed the fetal cardiotocography results and labeled 1456 of them as gold-standard; these results were used to train and validate the model. The remaining results were used to validate the clinical effectiveness of the model with the actual outcome.

Results: In a test dataset, our model achieved an area under the receiver operating characteristic curve (AUROC) of 0.89 and area under the precision-recall curve (AUPRC) of 0.73 in an internal validation dataset. An average AUROC of 0.73 and average AUPRC of 0.40 were achieved in the external validation dataset. Fetus abnormality score, as calculated from the continuous fetal cardiotocography results, was significantly associated with actual clinical outcomes [intrauterine growth restriction: odds ratio, 3.626 (p=0.031); Apgar score 1 min: odds ratio, 9.523 (p<0.001), Apgar score 5 min: odds ratio, 11.49 (p=0.001), and fetal distress: odds ratio, 23.09 (p<0.001)].

Conclusion: The machine learning model developed in this study showed precision in classifying FHR signals. This suggests that the model can be applied to medical devices as a screening tool for monitoring fetal status.

Keywords: Cardiotocography; high-risk-pregnancy; machine learning.

PubMed Disclaimer

Conflict of interest statement

MK is the founder and an employee of COSMOSWHALE Inc. DY is the founder and an employee of BUD.on Inc. The other authors declare no conflicts of interests.

Figures

Fig. 1
Fig. 1. Study design. Datasets were gathered and read by a team of obstetricians. Readings were recorded to evaluate 1456 fetal cardiotocography (qualified fetal cardiotocography, left side of the flowchart), while the other fetal cardiotocography results without the reading were used to evaluate the model in clinical situations (clinical validation dataset, right side of the flowchart). For the qualified fetal cardiotocography dataset, the classification model was trained to find abnormal data. Clinical validation datasets were created to represent three clinical situations according to the time window selected from the fetal cardiotocography results. NST, non stress test; CTU-UHB, Czech Technical University-University Hospital in Brno; SQI, signal quality index; IUGR, intrauterine growth restriction.
Fig. 2
Fig. 2. Classification model structure. As an input of the classification model, 2-channel waveform data were created based on the fetal heart rate (FHR) wave and uterine contraction wave (UC). Randomly initialised convolutional kernels were applied to the input and transformed to two features that were used to classify the fetus status by lightGBM classification machine learning model.
Fig. 3
Fig. 3. Data pre-processing flow for transforming images to CSV format. Data pre-processing consisted of three image process techniques. Waveform pixels were extracted by using RGB channel differences. Then, the edges of the graph were identified through the Hough transform algorithm results. The value of the waveform pixels was calculated using the relative position of the pixel compared to the edge. Finally, data were saved to csv format. FHR, fetal heart rate; RGB, Red Green Blue; CSV, comma-separated values.
Fig. 4
Fig. 4. Description of the clinical validation process. The clinical validation dataset consisted of fetal cardiotocography without reads combined with maternal demographic information (age, IUP) and postpartum neonatal status (IUGR, Apgar, Fetal distress). Based on delivery, the closest 20 minutes, random 20 minutes, and all non-overlapping time windows were selected to calculate the abnormal probability score. The calculated abnormal probability score, maternal age, and IUP were used as independent variables. Regression models were fitted on IUGR, Apgar, and fetal distress as dependent variables. NST, non stress test; IUP, intrauterine pregnancy; IUGR, intrauterine growth restriction.
Fig. 5
Fig. 5. The receiver operating characteristic (ROC; left) and precision-recall (PR; right) curves of the internal (A and B) and external validation dataset (C and D). The results were calculated through soft-voting models (10-fold each). The value of the area under the curve is shown at the right-bottom side of the graph. AUROC, area under the receiver operating characteristic curve; AUPRC, area under the precision-recall curve.
See this image and copyright information in PMC

References

    1. Ayres-de-Campos D. Electronic fetal monitoring or cardiotocography, 50 years later: what’s in a name? Am J Obstet Gynecol. 2018;218:545–546. - PubMed
    1. Kawakita T, Reddy UM, Landy HJ, Iqbal SN, Huang CC, Grantz KL. Neonatal complications associated with use of fetal scalp electrode: a retrospective study. BJOG. 2016;123:1797–1803. - PMC - PubMed
    1. O’Neill E, Thorp J. Antepartum evaluation of the fetus and fetal well being. Clin Obstet Gynecol. 2012;55:722–730. - PMC - PubMed
    1. Hoyer D, Åŧebrowski J, Cysarz D, Gonçalves H, Pytlik A, Amorim-Costa C, et al. Monitoring fetal maturation-objectives, techniques and indices of autonomic function. Physiol Meas. 2017;38:R61–R88. - PMC - PubMed
    1. Gierałtowski J, Hoyer D, Schneider U, Åŧebrowski JJ. Formation of functional associations across time scales in the fetal autonomic control system--a multifractal analysis. Auton Neurosci. 2015;190:33–39. - PubMed