Trends and methods in intensive care unit (ICU) research using machine learning: latent dirichlet allocation (LDA)-based thematic literature review

Abstract

Introduction: The use of machine learning (ML) in intensive care units (ICUs) has led to a large yet fragmented body of literature. It is imperative to conduct a systematic analysis and synthesis of this research to identify methodological trends, clinical applications, and knowledge deficits.

Methods: A systematic literature review was conducted in accordance with the PRISMA guidelines, encompassing 2,507 ICU-focused ML publications from 2019 to 2024. Latent Dirichlet Allocation (LDA), an unsupervised topic modeling approach, was used with n-gram and no-n-gram tokenization strategies. Bayesian optimization approaches were used to increase model coherence and diversity.

Results: The analysis demonstrated a substantial degree of methodological variability, emphasizing the predominance of studies on infection surveillance and complication prediction. N-gram tokenization efficiently identified clinically specific topics, but no-n-gram techniques produced larger interpretative groups. Underexplored fields include emerging research areas like drug response prediction, pediatric-specific modeling, and surgical risk classification.

Conclusion: In conclusion, the study highlights the significance of methodological transparency and tokenization strategies while offering a thorough topic overview and identifying methodological trends in the literature on ICU - ML. Future research should prioritize neglected areas such as pediatric care modeling and therapy response, utilizing advanced ML techniques and multimodal data integration to enhance the outcomes of ICU patients.

Keywords: Clinical applications; Intensive care unit; Latent dirichlet allocation; Literature review; Machine learning; Topic modeling; Unsupervised modeling.

Fig. 1

PRISMA flow diagram for study selection

Fig. 2

Yearly distribution of ICU + ML publications (2019–2024)

Fig. 3

Dataset usage across publications (n = 2507)

Fig. 4

Frequency distribution of ML algorithms utilized in the studies

Fig. 5

Temporal trends in the use of the top 10 learning-based ML algorithms in ICU-related studies (2019–2024)

Fig. 6

Distribution of clinical focus areas in ICU-related ML studies (excluding mortality) Abbreviations: AKI (acute kidney injury), ARDS (acute respiratory distress syndrome), NICU (neonatal intensive care unit), PICU (pediatric intensive care unit)

Fig. 7

No N-Gram Article—Topic distribution

Fig. 8

N-Gram Article—Topic distribution

Fig. 9

Topic similarity-based dendrogram

Fig. 10

Word cloud for thematic clusters C1, C2, C3, C4, C5, C6 and C7