Critical Transitions in Intensive Care Units: A Sepsis Case Study

Abstract

The progression of complex human diseases is associated with critical transitions across dynamical regimes. These transitions often spawn early-warning signals and provide insights into the underlying disease-driving mechanisms. In this paper, we propose a computational method based on surprise loss (SL) to discover data-driven indicators of such transitions in a multivariate time series dataset of septic shock and non-sepsis patient cohorts (MIMIC-III database). The core idea of SL is to train a mathematical model on time series in an unsupervised fashion and to quantify the deterioration of the model’s forecast (out-of-sample) performance relative to its past (in-sample) performance. Considering the highest value of the moving average of SL as a critical transition, our retrospective analysis revealed that critical transitions occurred at a median of over 35 hours before the onset of septic shock, which suggests the applicability of our method as an early-warning indicator. Furthermore, we show that clinical variables at critical-transition regions are significantly different between septic shock and non-sepsis cohorts. Therefore, our paper contributes a critical-transition-based data-sampling strategy that can be utilized for further analysis, such as patient classification. Moreover, our method outperformed other indicators of critical transition in complex systems, such as temporal autocorrelation and variance.

Prediction of Mortality in Hemodialysis Patients Using Moving Multivariate Distance

“…In neuroscience, Bashan et al (2012) demonstrated that several integrated physiological systems play part in topological transition in sleep stages. Ghalati et al (2019) used the Surprise Loss (SL) approach to characterize the critical transition that occurs before a septic shock. Dakos et al (2008) illustrated the critical slowing down that precedes tipping points in climate transitions, using a time-series autocorrelation approach.…”

“…While the structure of this network is far from being elucidated and many nodes are still unidentified, it may nonetheless be possible to infer much about the network via sampling of small subsets of nodes (i.e., molecules), due to the emergent properties of the network as a whole that lend coherence to its state and dynamics (Cohen, 2016;Cohen et al, 2021). This approach can be linked to a more general area of complex systems theory that has recently gained substantial attention: critical transitions (Dakos et al, 2008(Dakos et al, , 2012Scheffer et al, 2009;Bashan et al, 2012;Ghalati et al, 2019). In complex systems, system dynamics such as a change in variability may provide early warning signs (EWSs) of impending state changes known as critical transitions (e.g., ecological collapse, financial crises, and shifts in climate regime).…”

“…In complex systems, system dynamics such as a change in variability may provide early warning signs (EWSs) of impending state changes known as critical transitions (e.g., ecological collapse, financial crises, and shifts in climate regime). Only a few studies have explicitly applied this framework to health and disease ( Gijzel et al, 2017 ; Ghalati et al, 2019 ; Nakazato et al, 2020 ), though many studies have reported higher variation in various biomarkers preceding the onset of adverse outcomes in subjects with chronic diseases ( Holzel, 1987 ; Ma et al, 2012 ; Mendez et al, 2013 ; Myers et al, 2018 ; Woo et al, 2018 ), including Chronic Kidney Disease (CKD; Yang et al, 2007 ; Flythe et al, 2013 ; Selvarajah et al, 2014 ; Corte et al, 2015 ; Nakazato et al, 2017 ). Increased variance is indeed one of the main characteristics of resilience loss ( Gijzel et al, 2017 ).…”

Zukunft der Intensivmedizin – Künstliche Intelligenz

“…In einer Studie von Ghalet et al wurde die sog. "Surprise Loss"-Methode angewendet, um klinische Veränderungen an Sepsispatienten zu identifizieren und zu charakterisieren [7]. Hierbei werden minimale Instabilitäten in großen Datensätzen identifiziert, in diesem Fall Veränderungen von Parametern, die auf eine Progression einer Organdysfunktion bei einer Sepsis ("Transition der kritischen Erkrankung") hinweisen.…”

WISDoM: Characterizing Neurological Time Series With the Wishart Distribution

“…As a consequence, clipping or padding techniques are applied, altering the underlying temporal structure. In recent years, studies on such data have seen an increasing popularity in a wide range of fields, from functional magnetic resonance imaging (fMRI) (Van Den Heuvel and Pol, 2010 ; Azarmi et al, 2019 ; Yan et al, 2019 ) to time series exploration for critical transition prediction in clinical scenarios (Cuesta-Frau et al, 2019 ; Ghalati et al, 2019 ). The common goal of this type of research is to develop models and algorithms capable of reaching the highest possible classification and prediction performances, for diagnostic and real-time applications, while unveiling underlying information about a system.…”