PredictPTB: an interpretable preterm birth prediction model using attention-based recurrent neural networks

<h3>Background</h3><p dir="ltr">Early identification of pregnant women at risk for preterm birth (PTB), a major cause of infant mortality and morbidity, has a significant potential to improve prenatal care. However, we lack effective predictive models which can accurately...

وصف كامل

محفوظ في:
التفاصيل البيبلوغرافية
المؤلف الرئيسي: Rawan AlSaad (14159019) (author)
مؤلفون آخرون: Qutaibah Malluhi (3158757) (author), Sabri Boughorbel (846228) (author)
منشور في: 2022
الموضوعات:
الوسوم: إضافة وسم
لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!
_version_ 1864513519057108992
author Rawan AlSaad (14159019)
author2 Qutaibah Malluhi (3158757)
Sabri Boughorbel (846228)
author2_role author
author
author_facet Rawan AlSaad (14159019)
Qutaibah Malluhi (3158757)
Sabri Boughorbel (846228)
author_role author
dc.creator.none.fl_str_mv Rawan AlSaad (14159019)
Qutaibah Malluhi (3158757)
Sabri Boughorbel (846228)
dc.date.none.fl_str_mv 2022-02-14T03:00:00Z
dc.identifier.none.fl_str_mv 10.1186/s13040-022-00289-8
dc.relation.none.fl_str_mv https://figshare.com/articles/journal_contribution/PredictPTB_an_interpretable_preterm_birth_prediction_model_using_attention-based_recurrent_neural_networks/25659111
dc.rights.none.fl_str_mv CC BY 4.0
info:eu-repo/semantics/openAccess
dc.subject.none.fl_str_mv Biological sciences
Biochemistry and cell biology
Genetics
Mathematical sciences
Numerical and computational mathematics
Deep learning
Predictive models
Attention mechanism
Electronic health record
Preterm birth
Pregnancy
dc.title.none.fl_str_mv PredictPTB: an interpretable preterm birth prediction model using attention-based recurrent neural networks
dc.type.none.fl_str_mv Text
Journal contribution
info:eu-repo/semantics/publishedVersion
text
contribution to journal
description <h3>Background</h3><p dir="ltr">Early identification of pregnant women at risk for preterm birth (PTB), a major cause of infant mortality and morbidity, has a significant potential to improve prenatal care. However, we lack effective predictive models which can accurately forecast PTB and complement these predictions with appropriate interpretations for clinicians. In this work, we introduce a clinical prediction model (PredictPTB) which combines variables (medical codes) readily accessible through electronic health record (EHR) to accurately predict the risk of preterm birth at 1, 3, 6, and 9 months prior to delivery.</p><h3>Methods</h3><p dir="ltr">The architecture of PredictPTB employs recurrent neural networks (RNNs) to model the longitudinal patient’s EHR visits and exploits a single code-level attention mechanism to improve the predictive performance, while providing temporal code-level and visit-level explanations for the prediction results. We compare the performance of different combinations of prediction time-points, data modalities, and data windows. We also present a case-study of our model’s interpretability illustrating how clinicians can gain some transparency into the predictions.</p><h3>Results</h3><p dir="ltr">Leveraging a large cohort of 222,436 deliveries, comprising a total of 27,100 unique clinical concepts, our model was able to predict preterm birth with an ROC-AUC of 0.82, 0.79, 0.78, and PR-AUC of 0.40, 0.31, 0.24, at 1, 3, and 6 months prior to delivery, respectively. Results also confirm that observational data modalities (such as diagnoses) are more predictive for preterm birth than interventional data modalities (e.g., medications and procedures).</p><h3>Conclusions</h3><p dir="ltr">Our results demonstrate that PredictPTB can be utilized to achieve accurate and scalable predictions for preterm birth, complemented by explanations that directly highlight evidence in the patient’s EHR timeline.</p><h2>Other Information</h2><p dir="ltr">Published in: BioData Mining<br>License: <a href="https://creativecommons.org/licenses/by/4.0" target="_blank">https://creativecommons.org/licenses/by/4.0</a><br>See article on publisher's website: <a href="https://dx.doi.org/10.1186/s13040-022-00289-8" target="_blank">https://dx.doi.org/10.1186/s13040-022-00289-8</a></p>
eu_rights_str_mv openAccess
id Manara2_0e5a295d6d1c4c9f5c3707f7caadc979
identifier_str_mv 10.1186/s13040-022-00289-8
network_acronym_str Manara2
network_name_str Manara2
oai_identifier_str oai:figshare.com:article/25659111
publishDate 2022
repository.mail.fl_str_mv
repository.name.fl_str_mv
repository_id_str
rights_invalid_str_mv CC BY 4.0
spelling PredictPTB: an interpretable preterm birth prediction model using attention-based recurrent neural networksRawan AlSaad (14159019)Qutaibah Malluhi (3158757)Sabri Boughorbel (846228)Biological sciencesBiochemistry and cell biologyGeneticsMathematical sciencesNumerical and computational mathematicsDeep learningPredictive modelsAttention mechanismElectronic health recordPreterm birthPregnancy<h3>Background</h3><p dir="ltr">Early identification of pregnant women at risk for preterm birth (PTB), a major cause of infant mortality and morbidity, has a significant potential to improve prenatal care. However, we lack effective predictive models which can accurately forecast PTB and complement these predictions with appropriate interpretations for clinicians. In this work, we introduce a clinical prediction model (PredictPTB) which combines variables (medical codes) readily accessible through electronic health record (EHR) to accurately predict the risk of preterm birth at 1, 3, 6, and 9 months prior to delivery.</p><h3>Methods</h3><p dir="ltr">The architecture of PredictPTB employs recurrent neural networks (RNNs) to model the longitudinal patient’s EHR visits and exploits a single code-level attention mechanism to improve the predictive performance, while providing temporal code-level and visit-level explanations for the prediction results. We compare the performance of different combinations of prediction time-points, data modalities, and data windows. We also present a case-study of our model’s interpretability illustrating how clinicians can gain some transparency into the predictions.</p><h3>Results</h3><p dir="ltr">Leveraging a large cohort of 222,436 deliveries, comprising a total of 27,100 unique clinical concepts, our model was able to predict preterm birth with an ROC-AUC of 0.82, 0.79, 0.78, and PR-AUC of 0.40, 0.31, 0.24, at 1, 3, and 6 months prior to delivery, respectively. Results also confirm that observational data modalities (such as diagnoses) are more predictive for preterm birth than interventional data modalities (e.g., medications and procedures).</p><h3>Conclusions</h3><p dir="ltr">Our results demonstrate that PredictPTB can be utilized to achieve accurate and scalable predictions for preterm birth, complemented by explanations that directly highlight evidence in the patient’s EHR timeline.</p><h2>Other Information</h2><p dir="ltr">Published in: BioData Mining<br>License: <a href="https://creativecommons.org/licenses/by/4.0" target="_blank">https://creativecommons.org/licenses/by/4.0</a><br>See article on publisher's website: <a href="https://dx.doi.org/10.1186/s13040-022-00289-8" target="_blank">https://dx.doi.org/10.1186/s13040-022-00289-8</a></p>2022-02-14T03:00:00ZTextJournal contributioninfo:eu-repo/semantics/publishedVersiontextcontribution to journal10.1186/s13040-022-00289-8https://figshare.com/articles/journal_contribution/PredictPTB_an_interpretable_preterm_birth_prediction_model_using_attention-based_recurrent_neural_networks/25659111CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/256591112022-02-14T03:00:00Z
spellingShingle PredictPTB: an interpretable preterm birth prediction model using attention-based recurrent neural networks
Rawan AlSaad (14159019)
Biological sciences
Biochemistry and cell biology
Genetics
Mathematical sciences
Numerical and computational mathematics
Deep learning
Predictive models
Attention mechanism
Electronic health record
Preterm birth
Pregnancy
status_str publishedVersion
title PredictPTB: an interpretable preterm birth prediction model using attention-based recurrent neural networks
title_full PredictPTB: an interpretable preterm birth prediction model using attention-based recurrent neural networks
title_fullStr PredictPTB: an interpretable preterm birth prediction model using attention-based recurrent neural networks
title_full_unstemmed PredictPTB: an interpretable preterm birth prediction model using attention-based recurrent neural networks
title_short PredictPTB: an interpretable preterm birth prediction model using attention-based recurrent neural networks
title_sort PredictPTB: an interpretable preterm birth prediction model using attention-based recurrent neural networks
topic Biological sciences
Biochemistry and cell biology
Genetics
Mathematical sciences
Numerical and computational mathematics
Deep learning
Predictive models
Attention mechanism
Electronic health record
Preterm birth
Pregnancy