Prediction model for patients with acute respiratory distress syndrome: use of a genetic algorithm to develop a neural network model

Zhang, Zhongheng

doi:10.7717/peerj.7719

Cited by 13 publications

(13 citation statements)

References 46 publications

(48 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These two studies on ARDS indicated that ML was superior to the traditional APACHE score; however, the robustness requires further validation owing to the relatively small sample size or lack of multi-source validation. Similarly, our results demonstrated the ability of ML to predict the mortality of patients with ARDS, achieving comparable or even better performance than Ding et al (13) and Zhang (14) in a larger and multisource dataset. Compared with existing scoring systems, the RF-based algorithm has the following advantages.…”

Section: Discussionsupporting

confidence: 74%

“…Ding et al (13) built an RF model to predict ARDS events in a small sample of 296 patients, and achieved an AUROC of 0.82. Zhang (14) developed a neural network model with a genetic algorithm to predict 90-day mortality in a database of 745 patients, and in the testing cohort of 272 patients achieved an AUROC of 0.821. These two studies on ARDS indicated that ML was superior to the traditional APACHE score; however, the robustness requires further validation owing to the relatively small sample size or lack of multi-source validation.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Mortality prediction for patients with acute respiratory distress syndrome based on machine learning: a population-based study

et al. 2021

View full text Add to dashboard Cite

Background: Traditional scoring systems for patients' outcome prediction in intensive care units such as Oxygenation Saturation Index (OSI) and Oxygenation Index (OI) may not reliably predict the clinical prognosis of patients with acute respiratory distress syndrome (ARDS). Thus, none of them have been widely accepted for mortality prediction in ARDS. This study aimed to develop and validate a mortality prediction method for patients with ARDS based on machine learning using the Medical Information Mart for Intensive Care (MIMIC-III) and Telehealth Intensive Care Unit (eICU) Collaborative Research Database (eICU-CRD) databases.Methods: Patients with ARDS were selected based on the Berlin definition in MIMIC-III and eICU-CRD databases. The APPS score (using age, PaO 2 /FiO 2 , and plateau pressure), Simplified Acute Physiology Score II (SAPS-II), Sepsis-related Organ Failure Assessment (SOFA), OSI, and OI were calculated. With MIMIC-III data, a mortality prediction model was built based on the random forest (RF) algorithm, and the performance was compared to those of existing scoring systems based on logistic regression. The performance of the proposed RF method was also validated with the combined MIMIC-III and eICU-CRD data. The performance of mortality prediction was evaluated by using the area under the receiver operating characteristics curve (AUROC) and performing calibration using the Hosmer-Lemeshow test.Results: With the MIMIC-III dataset (308 patients, for comparisons with the existing scoring systems), the RF model predicted the in-hospital mortality, 30-day mortality, and 1-year mortality with an AUROC of 0.891, 0.883, and 0.892, respectively, which were significantly higher than those of the SAPS-II, APPS, OSI, and OI (all P<0.001). In the multi-source validation (the combined dataset of 2,235 patients in MIMIC-III and 331 patients in eICU-CRD), the RF model achieved an AUROC of 0.905 and 0.736 for predicting inhospital mortality for the MIMIC-III and eICU-CRD datasets, respectively. The calibration plots suggested good fits for our RF model and these scoring systems for predicting mortality. The platelet count and lactate level were the strongest predictive variables for predicting in-hospital mortality.Conclusions: Compared to the existing scoring systems, machine learning significantly improved performance for predicting ARDS mortality. Validation with multi-source datasets showed a relatively robust generalisation ability of our prediction model.

show abstract

Section: Discussionsupporting

confidence: 74%

Section: Discussionmentioning

confidence: 99%

Mortality prediction for patients with acute respiratory distress syndrome based on machine learning: a population-based study

et al. 2021

View full text Add to dashboard Cite

show abstract

“…The top 5 features in the model were minimum tidal volume, Glasgow coma score, respiratory rate, P aO 2 , and age. 31 In a secondary analysis of 2 multi-center RCTs from 44 hospitals, Zhang 32 conducted a study in the same area. However, rather than predict development of ARDS, the author sought to predict mortality following diagnosis of ARDS and to provide risk stratification in an effort to help clinicians choose appropriate treatment.…”

Section: Ardsmentioning

confidence: 99%

Monitoring Big Data During Mechanical Ventilation in the ICU

Smallwood

2020

Respir Care

View full text Add to dashboard Cite

The electronic health record allows the assimilation of large amounts of clinical and laboratory data. Big data describes the analysis of large data sets using computational modeling to reveal patterns, trends, and associations. How can big data be used to predict ventilator discontinuation or impending compromise, and how can it be incorporated into the clinical workflow? This article will serve 2 purposes. First, a general overview is provided for the layperson and introduces key concepts, definitions, best practices, and things to watch out for when reading a paper that incorporates machine learning. Second, recent publications at the intersection of big data, machine learning, and mechanical ventilation are presented.

show abstract

“…Algorithms used for the prediction of ARDS include mostly supervised ML algorithms including regression [ 15 ], classification [ 17 ], decision trees [ 17 , 19 ], or neural network [ 20 ]. There have also been attempts to identify ARDS with the help of unstructured data such as text from radiology reports using natural language processing and ML [ 21 ].…”

Section: Reviewmentioning

confidence: 99%

“…Furthermore, it is claimed to be relatively easy to use and does not require mathematical operations. Another approach using a genetic algorithm was developed by Zhan, who identified seven important variables (age, AIDS, leukemia, metastatic tumor, hepatic failure, lowest albumin, and FiO 2 ) for the prediction of mortality in ARDS patients [ 20 ]. These approaches achieved comparable performance to widely used APACHE III scoring in predicting mortality of patients in hospitals admitted with ARDS.…”

Section: Reviewmentioning

confidence: 99%

Can Big Data and Machine Learning Improve Our Understanding of Acute Respiratory Distress Syndrome?

Bhattarai

Gupta

Ali

et al. 2021

Cureus

View full text Add to dashboard Cite

Acute respiratory distress syndrome (ARDS) accounts for 10% of all diagnoses in the Intensive Care Unit, and about 40% of the patients succumb to the disease. Clinical methods alone can result in the under-recognition of this heterogeneous syndrome. The purpose of this study is to evaluate the role that big data and machine learning (ML) have played in understanding the heterogeneity of the disease and the development of various prediction algorithms. Most of the work in the field of ML in ARDS has been in the development of prediction models that have comparable efficacies to that of traditional models. Prediction algorithms have been useful in identifying new variables that may be important to consider in the future, supplementing the unknown information with the help of available noninvasive parameters, as well as predicting mortality. Phenotype identification using an unsupervised ML algorithm has been pivotal in classifying the heterogeneous population into more homogenous classes. Big data generated from ventilators in the form of ventilator waveform analysis and images in the form of radiomics have also been leveraged for the identification of the syndrome and can be incorporated into a clinical decision support system. Although the results are promising, lack of generalizability, “black box” nature of algorithms and concerns about “alarm fatigue” should be addressed for more mainstream adoption of these models.

show abstract

Prediction model for patients with acute respiratory distress syndrome: use of a genetic algorithm to develop a neural network model

Cited by 13 publications

References 46 publications

Mortality prediction for patients with acute respiratory distress syndrome based on machine learning: a population-based study

Mortality prediction for patients with acute respiratory distress syndrome based on machine learning: a population-based study

Monitoring Big Data During Mechanical Ventilation in the ICU

Can Big Data and Machine Learning Improve Our Understanding of Acute Respiratory Distress Syndrome?

Contact Info

Product

Resources

About