Machine Learning Methods for Mortality Prediction of Polytraumatized Patients in Intensive Care Units – Dealing with Imbalanced and High-Dimensional Data

García, María N. Moreno; Robledo, J. González; Martín-González, F.; Sánchez, F.; Barba, Mercedes Sánchez

doi:10.1007/978-3-319-10840-7_38

Cited by 7 publications

(5 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We also list high-performing machine learning models that have only been strictly validated in ICU settings; those include SANMF [41], SICULA (a.k.a. the super learner) [50] and [44]. It is worth noting that none of the models predicts ICU readmission.…”

Section: ) Comparison With Existing Outcome Prediction Modelsmentioning

confidence: 99%

“…We, therefore, resort to comparing with KD-OP's average performance in predicting mortality when applied to pneumonia and CKD using the MIMIC-III ICU dataset. Also, apart from [44], which reports sensitivity, the models strictly rely on ROC-AUC to report their performance. We will, therefore, resort to comparing with KD-OP's performance using ROC-AUC.…”

Section: ) Comparison With Existing Outcome Prediction Modelsmentioning

confidence: 99%

See 1 more Smart Citation

A Knowledge Distillation Ensemble Framework for Predicting Short- and Long-Term Hospitalization Outcomes From Electronic Health Records Data

Ibrahim

Bean

Searle

et al. 2022

IEEE J. Biomed. Health Inform.

View full text Add to dashboard Cite

The ability to perform accurate prognosis is crucial for proactive clinical decision making, informed resource management and personalised care. Existing outcome prediction models suffer from a low recall of infrequent positive outcomes. We present a highly-scalable and robust machine learning framework to automatically predict adversity represented by mortality and ICU admission and readmission from time-series of vital signs and laboratory results obtained within the first 24 hours of hospital admission. The stacked ensemble platform comprises two components: a) an unsupervised LSTM Autoencoder that learns an optimal representation of the time-series, using it to differentiate the less frequent patterns which conclude with an adverse event from the majority patterns that do not, and b) a gradient boosting model, which relies on the constructed representation to refine prediction by incorporating static features. The model is used to assess a patient's risk of adversity and provides visual justifications of its prediction. Results of three case studies show that the model outperforms existing platforms in ICU and general ward settings, achieving average Precision-Recall Areas Under the Curve (PR-AUCs) of 0.891 (95% CI: 0.878 -0.939) for mortality and 0.908 (95% CI: 0.870-0.935) in predicting ICU admission and readmission.

show abstract

Section: ) Comparison With Existing Outcome Prediction Modelsmentioning

confidence: 99%

Section: ) Comparison With Existing Outcome Prediction Modelsmentioning

confidence: 99%

A Knowledge Distillation Ensemble Framework for Predicting Short- and Long-Term Hospitalization Outcomes From Electronic Health Records Data

Ibrahim

Bean

Searle

et al. 2022

IEEE J. Biomed. Health Inform.

View full text Add to dashboard Cite

show abstract

“…However, the only other available metric for LightGBM is specificity, which is at a low 0.641 and entails a high rate Deterioration (14 days) -0.78 (0.74 -0.82) Guo [18] Deterioration (14 days) -0.67 (0.61 -0.73) Liu [24] Mortality & Deterioration -0.74 (0.69 -0.79) Galloway [14] Deterioration -0.72 (0.68 -0.77) Gong [16] Deterioration -0.853 (0.790-0.916) We also list high-performing machine learning models that have only been strictly validated in ICU settings; those include SANMF [35], SICULA (a.k.a. the super learner) [42] and [38]. It is worth noting that none of the models predicts ICU readmission.…”

Section: ) Comparison With Existing Outcome Prediction Modelsmentioning

confidence: 99%

“…We, therefore, resort to comparing with KD-OP's average performance in predicting mortality when applied to pneumonia and CKD using the MIMIC-III ICU dataset. Also, apart from [38], which reports sensitivity, the models strictly rely on ROC-AUC in reporting their performance. We will, therefore resort to comparing with KD-OP's performance using ROC-AUC.…”

Section: ) Comparison With Existing Outcome Prediction Modelsmentioning

confidence: 99%

A Knowledge Distillation Ensemble Framework for Predicting Short and Long-term Hospitalisation Outcomes from Electronic Health Records Data

Ibrahim¹,

Bean²,

Searle³

et al. 2020

Preprint

View full text Add to dashboard Cite

The ability to perform accurate prognosis of patients is crucial for proactive clinical decision making, informed resource management and personalised care. Existing outcome prediction models suffer from a low recall of infrequent positive outcomes. We present a highly-scalable and robust machine learning framework to automatically predict adversity represented by mortality and ICU admission from time-series vital signs and laboratory results obtained within the first 24 hours of hospital admission. The stacked platform comprises two components: a) an unsupervised LSTM Autoencoder that learns an optimal representation of the time-series, using it to differentiate the less frequent patterns which conclude with an adverse event from the majority patterns that do not, and b) a gradient

show abstract

“…The problem of imbalanced data is very common in the medical field and has been addressed in some works, such as studies of mortality [12,31,32], treatment outcomes [33], drug toxicity assessment [34] and medical diagnosis [35][36][37]. Preliminary studies about the behavior of ensemble classifiers, as opposed to single classifiers in imbalanced data contexts, are conducted to predict the mortality of polytraumatized patients [12] and the success of non-invasive mechanical ventilation [33].…”

Section: Related Workmentioning

confidence: 99%

Predictive Modeling of ICU Healthcare-Associated Infections from Imbalanced Data. Using Ensembles and a Clustering-Based Undersampling Approach

et al. 2019

View full text Add to dashboard Cite

Early detection of patients vulnerable to infections acquired in the hospital environment is a challenge in current health systems given the impact that such infections have on patient mortality and healthcare costs. This work is focused on both the identification of risk factors and the prediction of healthcare-associated infections in intensive-care units by means of machine-learning methods. The aim is to support decision making addressed at reducing the incidence rate of infections. In this field, it is necessary to deal with the problem of building reliable classifiers from imbalanced datasets. We propose a clustering-based undersampling strategy to be used in combination with ensemble classifiers. A comparative study with data from 4616 patients was conducted in order to validate our proposal. We applied several single and ensemble classifiers both to the original dataset and to data preprocessed by means of different resampling methods. The results were analyzed by means of classic and recent metrics specifically designed for imbalanced data classification. They revealed that the proposal is more efficient in comparison with other approaches.

show abstract

Machine Learning Methods for Mortality Prediction of Polytraumatized Patients in Intensive Care Units – Dealing with Imbalanced and High-Dimensional Data

Cited by 7 publications

References 8 publications

A Knowledge Distillation Ensemble Framework for Predicting Short- and Long-Term Hospitalization Outcomes From Electronic Health Records Data

A Knowledge Distillation Ensemble Framework for Predicting Short- and Long-Term Hospitalization Outcomes From Electronic Health Records Data

A Knowledge Distillation Ensemble Framework for Predicting Short and Long-term Hospitalisation Outcomes from Electronic Health Records Data

Predictive Modeling of ICU Healthcare-Associated Infections from Imbalanced Data. Using Ensembles and a Clustering-Based Undersampling Approach

Contact Info

Product

Resources

About