Predicting ICU Mortality Among Septic Patients Using Machine Learning Technique DOI Open Access
Abdulla Al‐Ansari,

Fatima A. Bahman Nejad,

Roudha J. Al-Nasr

et al.

Journal of Clinical Medicine, Journal Year: 2025, Volume and Issue: 14(10), P. 3495 - 3495

Published: May 16, 2025

Introduction: Sepsis leads to substantial global health burdens in terms of morbidity and mortality is associated with numerous risk factors. It crucial identify sepsis at an early stage order limit its escalation sequelae the condition. The purpose this research predict ICU evaluate predictive accuracy machine learning algorithms for among septic patients. Methods: study used a retrospective cohort from computerized records accumulated 280 hospitals between 2014 2015. Initially sample size was 23.47K. Several models were trained, validated, tested using five-fold cross-validation, three sampling strategies (Under-Sampling, Over-Sampling, Combination). Results: under-sampled approach combined augmentation Extra Trees model produced best performance Accuracy, Precision, Sensitivity, Specificity, F1-Score, AUC 90.99%, 84.16%, 94.89%, 88.48%, 89.20%, 91.69%, respectively, Top 30 features. For 29 features showed 82.99%, 51.38%, 71.72%, 85.41%, 59.87%, 78.56%, respectively. Down-Sampling, 31 81.78%, 49.08%, 79.76%, 82.21%, 60.76%, 80.98%, Conclusions: Machine can reliably when suitable clinical predictors are utilized. that proposed 90.99% only single-entry data. Incorporating longitudinal data could further enhance performance.

Language: Английский

Enhancing Real-Time Patient Monitoring in Intensive Care Units with Deep Learning and the Internet of Things DOI
Yongqiang Bai, Bingfei Gu, Chao Tang

et al.

Big Data, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 17, 2025

The demand for intensive care units (ICUs) is steadily increasing, yet there a relative shortage of medical staff to meet this need. Intensive work inherently heavy and stressful, highlighting the importance optimizing these units' working conditions processes. Such optimization crucial enhancing efficiency elevating level diagnosis treatment provided in ICUs. intelligent ICU concept represents novel ward management model that has emerged through advancements modern science technology. This includes communication technology, Internet Things (IoT), artificial intelligence (AI), robotics, big data analytics. By leveraging technologies, aims significantly reduce potential risks associated with human error improve patient monitoring outcomes. Deep learning (DL) IoT technologies have huge revolutionize surveillance patients ICUs due critical complex nature their conditions. article provides an overview most recent research applications linical critically ill patients, focus on execution AI. In ICU, seamless continuous critical, as even little delays decision-making can result irreparable repercussions or death. looks at how like DL monitoring, clinical results, Furthermore, it investigates function wearable advanced health sensors coupled networking systems, which enable secure connection analysis various forms predictive remote by professionals. assessing existing outlining roles IoT, analyzing benefits limitations integration, study hopes shed light future identify opportunities further research.

Language: Английский

Citations

2

21st century critical care medicine: An overview DOI Open Access
Smitesh Padte, Vikramaditya Samala Venkata, Priyal Mehta

et al.

World Journal of Critical Care Medicine, Journal Year: 2024, Volume and Issue: 13(1)

Published: March 5, 2024

Critical care medicine in the 21st century has witnessed remarkable advancements that have significantly improved patient outcomes intensive units (ICUs). This abstract provides a concise summary of latest developments critical care, highlighting key areas innovation. Recent include Precision Medicine: Tailoring treatments based on individual characteristics, genomics, and biomarkers to enhance effectiveness therapies. The objective is describe recent Care Medicine. Telemedicine: integration telehealth technologies for remote monitoring consultation, facilitating timely interventions. Artificial intelligence (AI): AI-driven tools early disease detection, predictive analytics, treatment optimization, enhancing clinical decision-making. Organ Support: Advanced life support systems, such as Extracorporeal Membrane Oxygenation Continuous Renal Replacement Therapy provide better organ support. Infection Control: Innovative infection control measures combat emerging pathogens reduce healthcare-associated infections. Ventilation Strategies: ventilation modes lung-protective strategies minimize ventilator-induced lung injury. Sepsis Management: Early recognition aggressive management sepsis with tailored Patient-Centered Care: A shift towards patient-centered focusing psychological emotional well-being addition medical needs. We conducted thorough literature search PubMed, EMBASE, Scopus using our strategy, incorporating keywords telemedicine, management. total 125 articles meeting criteria were included qualitative synthesis. To ensure reliability, we focused only published English language within last two decades, excluding animal studies,

Language: Английский

Citations

8

Artificial Intelligence in the Intensive Care Unit: Present and Future DOI
Jhossmar Cristians Auza-Santiváñez, Ariel Sosa Remón,

Freddy Ednildon Bautista-Vanegas

et al.

Deleted Journal, Journal Year: 2025, Volume and Issue: 4, P. 464 - 464

Published: March 18, 2025

Introduction: Artificial intelligence (AI) is significantly transforming critical medicine and intensive care. Its ability to process large volumes of data generate accurate predictions has improved medical decision-making, optimizing diagnosis, treatment, reducing the workload healthcare personnel. Methodology: A literature review was conducted between November 2024 February 2025, consulting databases such as SciELO, LILACS, Scopus, PubMed-MedLine, Google Scholar, ClinicalKeys. Original articles, case reports, open-access systematic reviews from last 5 years were selected, using descriptors in Health Sciences (DeCS) Boolean operators for search. Development: Current applications AI ICU include: Monitoring early detection adverse events sensors machine learning algorithms; diagnosis prognosis through deep neural networks image interpretation; treatment optimization, including adjustments mechanical ventilation pharmacogenomics; efficient management hospital resources. The future care oriented towards more explanatory transparent systems, personalized precision medicine, integration with emerging technologies automation clinical processes. Conclusions: redefining units, improving diagnostic accuracy, treatments, decision-making thus allowing management. However, advanced it is, will never replace empathy judgment professionals. By integrating responsibly, we not only save lives, but also humanize patient care, always remembering that, at heart there compassion commitment each patient.

Language: Английский

Citations

0

Predicting ICU Mortality Among Septic Patients Using Machine Learning Technique DOI Open Access
Abdulla Al‐Ansari,

Fatima A. Bahman Nejad,

Roudha J. Al-Nasr

et al.

Journal of Clinical Medicine, Journal Year: 2025, Volume and Issue: 14(10), P. 3495 - 3495

Published: May 16, 2025

Introduction: Sepsis leads to substantial global health burdens in terms of morbidity and mortality is associated with numerous risk factors. It crucial identify sepsis at an early stage order limit its escalation sequelae the condition. The purpose this research predict ICU evaluate predictive accuracy machine learning algorithms for among septic patients. Methods: study used a retrospective cohort from computerized records accumulated 280 hospitals between 2014 2015. Initially sample size was 23.47K. Several models were trained, validated, tested using five-fold cross-validation, three sampling strategies (Under-Sampling, Over-Sampling, Combination). Results: under-sampled approach combined augmentation Extra Trees model produced best performance Accuracy, Precision, Sensitivity, Specificity, F1-Score, AUC 90.99%, 84.16%, 94.89%, 88.48%, 89.20%, 91.69%, respectively, Top 30 features. For 29 features showed 82.99%, 51.38%, 71.72%, 85.41%, 59.87%, 78.56%, respectively. Down-Sampling, 31 81.78%, 49.08%, 79.76%, 82.21%, 60.76%, 80.98%, Conclusions: Machine can reliably when suitable clinical predictors are utilized. that proposed 90.99% only single-entry data. Incorporating longitudinal data could further enhance performance.

Language: Английский

Citations

0