Predicting mortality and hospitalization in heart failure using machine learning: A systematic literature review

Mpanya, Dineo; Çelik, Turgay; Klug, Eric; Ntsinjana, Hopewell

doi:10.1016/j.ijcha.2021.100773

Cited by 22 publications

(17 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Even though a more complicated model might have better predictive power, than for example a logistic regression model, if it is a black box model it might not allow for this confidence. While explainable and interpretable AI have been a recent focus within the machine learning field, there is still evidence to show that clinicians and patients distrust in machine learning methods prevents greater uptake (Elish, 2018 ; Mpanya et al, 2021 ; Joshi et al, 2022 ).…”

Section: Discussionmentioning

confidence: 99%

A review of risk concepts and models for predicting the risk of primary stroke

Hunter

Kelleher

2022

Front. Neuroinform.

View full text Add to dashboard Cite

Predicting an individual's risk of primary stroke is an important tool that can help to lower the burden of stroke for both the individual and society. There are a number of risk models and risk scores in existence but no review or classification designed to help the reader better understand how models differ and the reasoning behind these differences. In this paper we review the existing literature on primary stroke risk prediction models. From our literature review we identify key similarities and differences in the existing models. We find that models can differ in a number of ways, including the event type, the type of analysis, the model type and the time horizon. Based on these similarities and differences we have created a set of questions and a system to help answer those questions that modelers and readers alike can use to help classify and better understand the existing models as well as help to make necessary decisions when creating a new model.

show abstract

Section: Discussionmentioning

confidence: 99%

A review of risk concepts and models for predicting the risk of primary stroke

Hunter

Kelleher

2022

Front. Neuroinform.

View full text Add to dashboard Cite

show abstract

“…ML for HF-related tasks has been widely used in many studies, using different kinds of data (e.g., ECG, EHR, etc.) to obtain indications related to the diagnosis or risk analysis [18,19]. A reduced version of the dataset used in this study was utilized in other studies as previously cited in the introductory section.…”

Section: Discussionmentioning

confidence: 99%

“…Of particular interest in this study is the evaluation of mortality and hospitalization due to HF since it is one of the greatest challenges associated with the disease. In a recent systematic review [18], many ML-based studies were reported to have dwelt on the prediction of these outcomes. The measured Area Under the Receiver Operating Characteristic (AUROC or AUC) curve varied from between 0.47 to 0.84 for hospitalizations, and from 0.48 to 0.92 for mortality predictions.…”

Section: Related Workmentioning

confidence: 99%

Deep Learning for Predicting Congestive Heart Failure

et al. 2022

View full text Add to dashboard Cite

Congestive heart failure (CHF) is one of the most debilitating cardiac disorders. It is a costly disease in terms of both lives and financial outlays, given the high rate of hospital re-admissions and mortality. Heart failure (HF) is notoriously difficult to identify on time, and is frequently accompanied by additional comorbidities that further complicate diagnosis. Many decision support systems (DSS) have been developed to facilitate diagnosis and to raise the standard of screening and monitoring operations, even for non-expert staff. This is confirmed in the literature by records of highly performing diagnosis-aid systems, which are unfortunately not very relevant to expert cardiologists. In order to assist cardiologists in predicting the trajectory of HF, we propose a deep learning-based system which predicts severity of disease progression by employing medical patient history. We tested the accuracy of four models on a labeled dataset, composed of 1037 records, to predict CHF severity and progression, achieving results comparable to studies based on much larger datasets, none of which used longitudinal multi-class prediction. The main contribution of this work is that it demonstrates that a fairly complicated approach can achieve good results on a medium size dataset, providing a reasonably accurate means of determining the evolution of CHF well in advance. This potentially constitutes a significant aid for healthcare managers and expert cardiologists in designing different therapies for medication, healthy lifestyle changes and quality of life (QoL) management, while also promoting allocation of resources with an evidence-based approach.

show abstract

“…Numerous prediction tools have been proposed to predict a variety of health outcomes, including diabetes mellitus [ 1 ], hypertension [ 2 ], dyslipidemia [ 3 ], cancer [ 4 ], cardiovascular diseases (CVDs) [ 5 ], and mortality [ 6 ]. Because multiple risk factors usually contribute to the development of these health outcomes, prediction tools that account for the effects of multiple risk factors help stratify individuals according to their risk levels.…”

Section: Introductionmentioning

confidence: 99%

Application of deep neural survival networks to the development of risk prediction models for diabetes mellitus, hypertension, and dyslipidemia

Sasagawa,

Inoue,

Futagami

et al. 2023

Journal of Hypertension

View full text Add to dashboard Cite

Objectives : Although numerous risk prediction models have been proposed, few such models have been developed using neural network-based survival analysis. We developed risk prediction models for three cardiovascular disease risk factors (diabetes mellitus, hypertension, and dyslipidemia) among a working-age population in Japan using DeepSurv, a deep feed-forward neural network. Methods : Data were obtained from the Japan Epidemiology Collaboration on Occupational Health Study. A total of 51 258, 44 197, and 31 452 individuals were included in the development of risk models for diabetes mellitus, hypertension, and dyslipidemia, respectively; two-thirds of whom were used to develop prediction models, and the rest were used to validate the models. We compared the performances of DeepSurv-based models with those of prediction models based on the Cox proportional hazards model. Results : The area under the receiver-operating characteristic curve was 0.878 [95% confidence interval (CI) = 0.864–0.892] for diabetes mellitus, 0.835 (95% CI = 0.826–0.845) for hypertension, and 0.826 (95% CI = 0.817–0.835) for dyslipidemia. Compared with the Cox proportional hazards-based models, the DeepSurv-based models had better reclassification performance [diabetes mellitus: net reclassification improvement (NRI) = 0.474, P ≤ 0.001; hypertension: NRI = 0.194, P ≤ 0.001; dyslipidemia: NRI = 0.397, P ≤ 0.001] and discrimination performance [diabetes mellitus: integrated discrimination improvement (IDI) = 0.013, P ≤ 0.001; hypertension: IDI = 0.007, P ≤ 0.001; and dyslipidemia: IDI = 0.043, P ≤ 0.001]. Conclusion : This study suggests that DeepSurv has the potential to improve the performance of risk prediction models for cardiovascular disease risk factors.

show abstract

Predicting mortality and hospitalization in heart failure using machine learning: A systematic literature review

Abstract: Graphical abstract

Cited by 22 publications

References 46 publications

A review of risk concepts and models for predicting the risk of primary stroke

A review of risk concepts and models for predicting the risk of primary stroke

Deep Learning for Predicting Congestive Heart Failure

Application of deep neural survival networks to the development of risk prediction models for diabetes mellitus, hypertension, and dyslipidemia

Contact Info

Product

Resources

About