Dataset Information

Comparing regression and neural network techniques for personalized predictive analytics to promote lung protective ventilation in Intensive Care Units.

ABSTRACT: Mechanical ventilation is a lifesaving tool and provides organ support for patients with respiratory failure. However, injurious ventilation due to inappropriate delivery of high tidal volume can initiate or potentiate lung injury. This could lead to acute respiratory distress syndrome, longer duration of mechanical ventilation, ventilator associated conditions and finally increased mortality. In this study, we explore the viability and compare machine learning methods to generate personalized predictive alerts indicating violation of the safe tidal volume per ideal body weight (IBW) threshold that is accepted as the upper limit for lung protective ventilation (LPV), prior to application to patients. We process streams of patient respiratory data recorded per minute from ventilators in an intensive care unit and apply several state-of-the-art time series prediction methods to forecast the behavior of the tidal volume metric per patient, 1 hour ahead. Our results show that boosted regression delivers better predictive accuracy than other methods that we investigated and requires relatively short execution times. Long short-term memory neural networks can deliver similar levels of accuracy but only after much longer periods of data acquisition, further extended by several hours computing time to train the algorithm. Utilizing Artificial Intelligence, we have developed a personalized clinical decision support tool that can predict tidal volume behavior within 10% accuracy and compare alerts recorded from a real world system to highlight that our models would have predicted violations 1 hour ahead and can therefore conclude that the algorithms can provide clinical decision support.

SUBMITTER: Hagan R

PROVIDER: S-EPMC7543875 | biostudies-literature | 2020 Oct

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Comparing regression and neural network techniques for personalized predictive analytics to promote lung protective ventilation in Intensive Care Units.

Hagan Rachael R Gillan Charles J CJ Spence Ivor I McAuley Danny D Shyamsundar Murali M

Computers in biology and medicine 20201008

Mechanical ventilation is a lifesaving tool and provides organ support for patients with respiratory failure. However, injurious ventilation due to inappropriate delivery of high tidal volume can initiate or potentiate lung injury. This could lead to acute respiratory distress syndrome, longer duration of mechanical ventilation, ventilator associated conditions and finally increased mortality. In this study, we explore the viability and compare machine learning methods to generate personalized p ...[more]

PMID: 33068808

Similar Datasets

Project description:BackgroundIn the intensive care unit (ICU), critical illness delirium occurs in the context of multiple comorbidities, multi-organ failure, and invasive management techniques, such as mechanical ventilation, sedation, and lack of sleep. Delirium is characterized by an acute confusional state defined by fluctuating mental status, inattention, and either disorganized thinking or an altered level of consciousness. The long-term cognitive and psychosocial function of patients that experience delirium in the ICU is of crucial interest because preliminary data suggest a strong association between ICU-related delirium and long-term cognitive impairment.ObjectiveThe aim of this study is to explore the relationship between delirium in the ICU and adverse outcomes by following mechanically ventilated patients for one year following their discharge from the ICU and collecting data on their long-term cognition and psychosocial function.MethodsThis study will be conducted by enrolling patients in two tertiary ICUs in Australia. We aim to recruit 200 patients who have been mechanically ventilated for more than 24 hours. Data will be collected at the following three time points: (1) at discharge where they will be administered the Mini-Mental State Examination (MMSE); (2) at 6 months after discharge from the ICU discharge where the Impact of Events Scale Revised (IES-R) and the Telephone Inventory for Cognitive Status (TICS) tests will be administered; and (3) at 12 months after discharge from the ICU where the patients will be administered the TICS and IES-R tests, as well as the Informant Questionnaire for Cognitive Decline in the Elderly (IQCODE). The IQCODE will be administered to their "person responsible" or the significant other of the patient.ResultsLong-term cognition and psychosocial function will be the primary outcome of this study. Mortality will also be investigated as a secondary outcome. Active enrollment will take place until the end of September 2016 and data collection will conclude at the end of September 2017. The analysis and results are expected to be available by March 2018.ConclusionDelirium during mechanical ventilation has been linked to longer ICU and hospital stays, higher financial burdens, increased risks of long-term cognitive impairment (ie, dementia), poor functional outcomes and quality of life, and decreased survival. However, delirium during mechanical ventilation in the ICU is not well understood. This study will advance our knowledge of the comprehensive, long-term effects of delirium on cognitive and psychosocial function.Trial registrationAustralian New Zealand Clinical Trials Registry (ANZCTR): ACTRN12616001116415; https://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=371216 (Archived by WebCite at http://www.webcitation.org/ 6nfDkGTcW).

Dataset Information

Comparing regression and neural network techniques for personalized predictive analytics to promote lung protective ventilation in Intensive Care Units.

Publications

Comparing regression and neural network techniques for personalized predictive analytics to promote lung protective ventilation in Intensive Care Units.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure