Predicting the Next-Day Perceived and Physiological Stress of Pregnant Women by Using Machine Learning and Explainability: Algorithm Development and Validation

Abstract

Background Cognitive behavioral therapy–based interventions are effective in reducing prenatal stress, which can have severe adverse health effects on mothers and newborns if unaddressed. Predicting next-day physiological or perceived stress can help to inform and enable pre-emptive interventions for a likely physiologically and perceptibly stressful day. Machine learning models are useful tools that can be developed to predict next-day physiological and perceived stress by using data collected from the previous day. Such models can improve our understanding of the specific factors that predict physiological and perceived stress and allow researchers to develop systems that collect selected features for assessment in clinical trials to minimize the burden of data collection. Objective The aim of this study was to build and evaluate a machine-learned model that predicts next-day physiological and perceived stress by using sensor-based, ecological momentary assessment (EMA)–based, and intervention-based features and to explain the prediction results. Methods We enrolled pregnant women into a prospective proof-of-concept study and collected electrocardiography, EMA, and cognitive behavioral therapy intervention data over 12 weeks. We used the data to train and evaluate 6 machine learning models to predict next-day physiological and perceived stress. After selecting the best performing model, Shapley Additive Explanations were used to identify the feature importance and explainability of each feature. Results A total of 16 pregnant women enrolled in the study. Overall, 4157.18 hours of data were collected, and participants answered 2838 EMAs. After applying feature selection, 8 and 10 features were found to positively predict next-day physiological and perceived stress, respectively. A random forest classifier performed the best in predicting next-day physiological stress (F1 score of 0.84) and next-day perceived stress (F1 score of 0.74) by using all features. Although any subset of sensor-based, EMA-based, or intervention-based features could reliably predict next-day physiological stress, EMA-based features were necessary to predict next-day perceived stress. The analysis of explainability metrics showed that the prolonged duration of physiological stress was highly predictive of next-day physiological stress and that physiological stress and perceived stress were temporally divergent. Conclusions In this study, we were able to build interpretable machine learning models to predict next-day physiological and perceived stress, and we identified unique features that were highly predictive of next-day stress that can help to reduce the burden of data collection.

How Wearable Sensors Can Support the Research on Foetal and Pregnancy Outcomes: A Scoping Review

“…Twenty-one studies described the application of wearable sensors to collect information about maternal parameters and activities during pregnancy [ 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 , 34 , 35 , 36 , 37 , 38 , 39 , 40 , 41 ]. These studies are described in Table 2 , together with a summary of the design and population, wearable device used, and main findings.…”

“…In fact, wearable sensing could be a very important part of studies evaluating interventions to improve wellbeing and physiological stress during and after pregnancy. Ng and colleagues conducted a prospective study to predict physiological and perceived stress of pregnant women using a body-conforming flexible ECG sensor, ecological momentary assessment (EMA) surveys, and machine learning models [ 34 ]. Their results showed that it was possible to predict next-day physiological and perceived stress; however, sensor-based data alone had poor predictive performance and needed to be integrated with EMA-based data [ 34 ].…”

“…Ng and colleagues conducted a prospective study to predict physiological and perceived stress of pregnant women using a body-conforming flexible ECG sensor, ecological momentary assessment (EMA) surveys, and machine learning models [ 34 ]. Their results showed that it was possible to predict next-day physiological and perceived stress; however, sensor-based data alone had poor predictive performance and needed to be integrated with EMA-based data [ 34 ]. Another aspect to consider in these studies is the elevated burden for participants, which might affect the response to interventions.…”

“…For the second domain of application, selected studies reported on wearable devices used to collect maternal parameters and activities during pregnancy, and to investigate their association with maternal and neonatal health [ 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 , 34 , 35 , 36 , 37 , 38 , 39 , 40 , 41 ]. Devices used were varied, including wristband and finger-based trackers, body-conforming flexible sensors, and more complex systems made of different sensors.…”

“…Although promising in terms of improving the monitoring of maternal parameters, these devices have not yet produced many concrete results. Some studies reported high correlations between sensor-based measurements and those obtained using current monitoring technologies, which are too often expensive and complex [ 34 , 36 ]. However, these studies largely focused on uncomplicated pregnancies [ 21 , 26 , 34 , 36 ], while future research will have to be done on women with pregnancy complications.…”