- Email: [email protected]
Changes in temporal structure of heart rate variability during clinical stress testing
e34
Meeting Abstracts for the Society for Complexity in Acute Illness (SCAI)
the use of the neutral networks is restricted to assigning IVA strains to reservoirs. All of the populations are treated as well mixed, so there is no network structure within any reservoir. That is, any agent in a given reservoir is equally likely to contact any other agent. The reservoirs are connected by weighted edges, where the weight of an edge determines how strongly the IVA strains can be passed between the different reservoirs. Each population is assigned characteristics such as birth and death rates, transmission rates, and seasonal effects. Within each population, we model every individual as an agent, where on the order of 105 agents are currently being simulated. Each agent retains a history of all previous IVA infections as well as the strains with which it is currently infected. Agents are added and removed from the populations. The trans-reservoir transmissions are done between specific agents. Two main types of evolutionary processes are incorporated in our model. The first is point mutations of the IVA strains' genotypes. Here we conceptualize each of the 8 genes to have a certain number of loci. This allows the strains to evolve over a particular reservoir's neutral network. If a strain evolves onto a network that can infect multiple reservoirs, we can have trans-reservoir infections. The second evolutionary process is the reassortment of 2 IVA strains within an agent. The genome of IVA is composed of 8 individual strands of RNA. If an agent is currently infected by more than a single strain, the RNA strands can be reassorted, allowing for the rise of novel strains. Results: We hope to demonstrate that using a multi-reservoir model can reproduce the effects of using neutral networks. In this way, the model will reproduce the phylodynamics of the IVA. We have also included the important features of multiple different species and strain reassortment for IVA evolution. Conclusions: We have examined an agent-based model of IVA evolution and epidemiology. The model includes multiple interacting reservoirs (eg, avian, swine and human) of agents. Each reservoir is assigned IVA strains by which it can be infected. One of the reservoirs acts as a bridge between other reservoirs. In this, transreservoir transmission is allowed. In addition, IVA can evolve within each reservoir through the processes of mutation and reassortment. doi:10.1016/j.jcrc.2009.06.040
References [1] Koelle K, Cobey S, Grenfell B, Pascual M. Epochal evolution shapes the phylodynamics of interpandemic influenza A (H3N2) in humans. Science 2006;314:1898-903. [2] Shih ACC, Hsiao TC, Ho MS, Li WH. Simultaneous amino acid substitutions at antigenic sites drive influenza A hemagglutinin evolution. PNAS 2007;104:6283-8. [3] Suzuki Y. Positive selection operates continuously on hemagglutinin during evolution of H3N2 human influenza A virus. Gene 2008;427: 111-6. Changes in temporal structure of heart rate variability during clinical stress testing☆ Timothy G. Buchman a,b, Robert E. Karsch a a Barnes-Jewish Hospital, Saint Louis, MO b Emory Center for Critical Care, Atlanta, GA
☆
Supported by DARPA and the James S. McDonnell Foundation.
Objectives: Biology is modular. The modules are structural (eg, genes, cells, tissues, organs, organisms) and evolutionary adaptation typically occurs when environmental pressure leads to reuse of a duplicate or near duplicate of an existing module. We hypothesized (1) that within-lifetime adaptation also occurs by module reuse and (2) that the complex variability observed in biological time series reflects reuse of temporal modules across multiple scales. Methods: We performed a retrospective analysis of 100 deidentified 12-lead electrocardiographic signals sampled at 500 Hz during standardized clinical stress testing in patients with normal sinus rhythms. The stresses studied included both treadmill sessions using a modified Bruce protocol as well as infusion of dobutamine. Preprocessing of the digitized files was performed to mitigate wandering baselines and to filter out ambient noise. Fiducial points were identified automatically using a filter-bank strategy. We generated 2 data sets for each record. The first data set consisted of R-R intervals and the second data set consisted of the (preprocessed) voltages between the R waves. We studied moving windows of 400 beats with 200 beat overlaps between windows. We calculated multiscale entropy over 6 scale factors from the R-R interval data and beat-by-beat covariance from the voltage data using a top 10% best-fit strategy. Results: The most general multiscale entropy pattern shows loss of multiscale entropy with increasing stress, which reaccumulates during the cool-down period. Some patients displayed intermediate adaptation with transient improvements in multiscale entropy before a final loss of that entropy as they reached their exercise capacity. The most general covariance pattern showed loss of the ability to access temporally remote modules as stress increased, with restoration of that access during the cool-down period. Conclusions: Variability analysis of clinical exercise tests shows changes in complex variability (temporal modularity). At the highest reversible stress levels, access to temporal modules across multiple scales appears to be limited and possibly limiting. These findings suggest that concurrent complex variability analysis might be used to infer that a patient is not only stressed but also approaching physiologic limits. Such monitoring could be used to predict impending physiologic collapse. doi:10.1016/j.jcrc.2009.06.041 Assessing the prediction potential of an in silico computer model of intracranial pressure dynamics Wayne Wakeland PhDa, Rachel Agbeko MDb, Kevin Vinecore BS c, Mark Peters MRCP, PhDd, Brahm Goldstein MD, MCRe,f a Systems Science Graduate Program, Portland State University, Portland, OR, USA b Critical Care Group, Portex Unit, Institute of Child Health, London, UK c Clinical Neurophysiology, Oregon Health and Science University, Portland, OR, USA d Paediatric Intensive Care Unit, Great Ormond Street Hospital for Children NHS Trust, London, UK e Department of Pediatrics, University of Medicine and Dentistry of New Jersey, New Brunswick, NJ, USA f Ikaria, Inc., Clinton, NJ, USA
Objective: Traumatic brain injury (TBI) frequently results in poor outcome, suggesting that new approaches are needed. We hypothesized that a patient-specific in silico computer model of
Meeting Abstracts for the Society for Complexity in Acute Illness (SCAI)
the use of the neutral networks is restricted to assigning IVA strains to reservoirs. All of the populations are treated as well mixed, so there is no network structure within any reservoir. That is, any agent in a given reservoir is equally likely to contact any other agent. The reservoirs are connected by weighted edges, where the weight of an edge determines how strongly the IVA strains can be passed between the different reservoirs. Each population is assigned characteristics such as birth and death rates, transmission rates, and seasonal effects. Within each population, we model every individual as an agent, where on the order of 105 agents are currently being simulated. Each agent retains a history of all previous IVA infections as well as the strains with which it is currently infected. Agents are added and removed from the populations. The trans-reservoir transmissions are done between specific agents. Two main types of evolutionary processes are incorporated in our model. The first is point mutations of the IVA strains' genotypes. Here we conceptualize each of the 8 genes to have a certain number of loci. This allows the strains to evolve over a particular reservoir's neutral network. If a strain evolves onto a network that can infect multiple reservoirs, we can have trans-reservoir infections. The second evolutionary process is the reassortment of 2 IVA strains within an agent. The genome of IVA is composed of 8 individual strands of RNA. If an agent is currently infected by more than a single strain, the RNA strands can be reassorted, allowing for the rise of novel strains. Results: We hope to demonstrate that using a multi-reservoir model can reproduce the effects of using neutral networks. In this way, the model will reproduce the phylodynamics of the IVA. We have also included the important features of multiple different species and strain reassortment for IVA evolution. Conclusions: We have examined an agent-based model of IVA evolution and epidemiology. The model includes multiple interacting reservoirs (eg, avian, swine and human) of agents. Each reservoir is assigned IVA strains by which it can be infected. One of the reservoirs acts as a bridge between other reservoirs. In this, transreservoir transmission is allowed. In addition, IVA can evolve within each reservoir through the processes of mutation and reassortment. doi:10.1016/j.jcrc.2009.06.040
References [1] Koelle K, Cobey S, Grenfell B, Pascual M. Epochal evolution shapes the phylodynamics of interpandemic influenza A (H3N2) in humans. Science 2006;314:1898-903. [2] Shih ACC, Hsiao TC, Ho MS, Li WH. Simultaneous amino acid substitutions at antigenic sites drive influenza A hemagglutinin evolution. PNAS 2007;104:6283-8. [3] Suzuki Y. Positive selection operates continuously on hemagglutinin during evolution of H3N2 human influenza A virus. Gene 2008;427: 111-6. Changes in temporal structure of heart rate variability during clinical stress testing☆ Timothy G. Buchman a,b, Robert E. Karsch a a Barnes-Jewish Hospital, Saint Louis, MO b Emory Center for Critical Care, Atlanta, GA
☆
Supported by DARPA and the James S. McDonnell Foundation.
Objectives: Biology is modular. The modules are structural (eg, genes, cells, tissues, organs, organisms) and evolutionary adaptation typically occurs when environmental pressure leads to reuse of a duplicate or near duplicate of an existing module. We hypothesized (1) that within-lifetime adaptation also occurs by module reuse and (2) that the complex variability observed in biological time series reflects reuse of temporal modules across multiple scales. Methods: We performed a retrospective analysis of 100 deidentified 12-lead electrocardiographic signals sampled at 500 Hz during standardized clinical stress testing in patients with normal sinus rhythms. The stresses studied included both treadmill sessions using a modified Bruce protocol as well as infusion of dobutamine. Preprocessing of the digitized files was performed to mitigate wandering baselines and to filter out ambient noise. Fiducial points were identified automatically using a filter-bank strategy. We generated 2 data sets for each record. The first data set consisted of R-R intervals and the second data set consisted of the (preprocessed) voltages between the R waves. We studied moving windows of 400 beats with 200 beat overlaps between windows. We calculated multiscale entropy over 6 scale factors from the R-R interval data and beat-by-beat covariance from the voltage data using a top 10% best-fit strategy. Results: The most general multiscale entropy pattern shows loss of multiscale entropy with increasing stress, which reaccumulates during the cool-down period. Some patients displayed intermediate adaptation with transient improvements in multiscale entropy before a final loss of that entropy as they reached their exercise capacity. The most general covariance pattern showed loss of the ability to access temporally remote modules as stress increased, with restoration of that access during the cool-down period. Conclusions: Variability analysis of clinical exercise tests shows changes in complex variability (temporal modularity). At the highest reversible stress levels, access to temporal modules across multiple scales appears to be limited and possibly limiting. These findings suggest that concurrent complex variability analysis might be used to infer that a patient is not only stressed but also approaching physiologic limits. Such monitoring could be used to predict impending physiologic collapse. doi:10.1016/j.jcrc.2009.06.041 Assessing the prediction potential of an in silico computer model of intracranial pressure dynamics Wayne Wakeland PhDa, Rachel Agbeko MDb, Kevin Vinecore BS c, Mark Peters MRCP, PhDd, Brahm Goldstein MD, MCRe,f a Systems Science Graduate Program, Portland State University, Portland, OR, USA b Critical Care Group, Portex Unit, Institute of Child Health, London, UK c Clinical Neurophysiology, Oregon Health and Science University, Portland, OR, USA d Paediatric Intensive Care Unit, Great Ormond Street Hospital for Children NHS Trust, London, UK e Department of Pediatrics, University of Medicine and Dentistry of New Jersey, New Brunswick, NJ, USA f Ikaria, Inc., Clinton, NJ, USA
Objective: Traumatic brain injury (TBI) frequently results in poor outcome, suggesting that new approaches are needed. We hypothesized that a patient-specific in silico computer model of