Donald Krieger

Physiologic data acquisition system and database for the study of disease dynamics in the intensive care unit.

ObjectiveTo describe a real-time, continuous physiologic data acquisition system for the study of disease dynamics in the intensive care unit. DesignDescriptive report. SettingA 16-bed pediatric intensive care unit in a tertiary care children’s hospital. PatientsA total of 170 critically ill or injured pediatric patients. InterventionsNone. Main Outcome MeasuresNone. ResultsWe describe a computerized data acquisition and analysis system for the study of critical illness and injury from the perspective of complex dynamic systems. Both parametric (1 Hz) and waveform (125–500 Hz) signals are recorded and analyzed. Waveform data include electrocardiogram, respiration, systemic arterial pressure (invasive and noninvasive), central venous pressure, pulmonary arterial pressure, left and right atrial pressures, intracranial pressure, body temperature, and oxygen saturation. Details of the system components are explained and examples are given from the resultant physiologic database of signal processing algorithms and signal analyses using linear and nonlinear metrics. ConclusionsWe have successfully developed a real-time, continuous physiologic data acquisition system that can capture, store, and archive data from pediatric intensive care unit patients for subsequent time series analysis of dynamic changes in physiologic state. The physiologic signal database generated from this system is available for analysis of dynamic changes caused by critical illness and injury.

Computer analyses of EEG-sleep in the neonate: methodological considerations.

Neonatal EEG interpretation can aid in the estimation of central nervous system maturation, as well as provide diagnostic and prognostic information of the high-risk infant. However, one cannot easily visualize the complex interrelationships coupling EEG and polysomnographic components of the EEG-sleep rhythm. This is particularly relevant for the preterm neonate, in whom a rudimentary sleep cycle has not yet been clearly delineated. Computer analysis can augment the information derived from the visual interpretation of scalp-generated EEG activity. Automated techniques for EEG-sleep analysis have only recently been applied to a neonatal population. Such studies have been limited to full-term rather than preterm infants and rely on conventional methods that assume stationarity of neurophysiologic signals. We describe a computer system that simultaneously compares behavioral and electrographic components of EEG-sleep in a manner that preserves the integrity of the signals over time, while investigating the time- and frequency-dependent relationships among signals. Strategies for on-line and off-line editing, data storage, and offline signal processing are described. Computational algorithms regarding analyses of EEG power, motility, and cardiorespiratory data are being used to study the ontogeny of EEG-sleep in asymptomatic preterm and full-term neonates. Computer strategies are based on both principles of stationarity and nonstationarity of physiologic signals and are applied depending on the temporal resolution required for specific signal processing needs.

A systems theoretic approach to the study of CNS function.

This paper presents a paradigm for using a general systems theoretic approach to study central nervous system function. Neuronal systems are conceptualized as consisting of different populations of neurons, each of which may be treated as a unit with its own dynamic properties. These dynamic properties of each population are determined by fundamental characteristics of its constituent neurons. Different populations interact with each other through the network structure in which they are embedded. We are modeling the linear and nonlinear properties of these systems using a theoretical framework based on two complementary approaches. First, a functional power series is used to characterize the input/output properties of the system. Second, a state-variable approach is used to characterize the internal structure and function of the system and to identify specific relationships among physiological variables.We have extensively investigated the functional power series approach to study the cat somatosensory system and the rabbit hippocampal formation. Both of these systems exhibit nonlinear properties in their response to electrical stimulation, and these nonlinear properties are characterized by the high-order kernels of the functional power series. State-variable models are being formulated to map these input/output properties onto internal models of the systems.

Multimedia MedNet: a medical collaboration and consultation system

MedNet and its predecessor, NeuroNet, have been in use since 1985. The current system provides real-time monitoring and multiparty consultation during 1,600 brain surgeries per year. >

NeuroNet: collaborative intraoperative guidance and control

Neurophysiological monitoring assesses CNS structure function relationships during surgery. NeuroNet supports remote performance of this task through real time multimodal data processing and multimedia network communication. The system is fully integrated, transparently combining the collection, processing, and presentation of real time data sources, including all physiological monitoring functions, with non real time functions and extensive online database information. Workstations are mounted in instrumentation racks and configured with appropriate electronics to support various data acquisition tasks including electroencephalograms (EEGs), electromyograms (EMGs), and multimodality evoked potentials. Multiple racks can be used in parallel on the same case if the number of variables to be monitored exceeds the capacity of a single tack. The data acquired on these systems is transparently accessible, in real time, across the network for both review and analysis.

Magnetoencephalographic virtual recording: a novel diagnostic tool for concussion

OBJECT Heightened recognition of the prevalence and significance of head injury in sports and in combat veterans has brought increased attention to the physiological and behavioral consequences of concussion. Current clinical practice is in part dependent on patient self-report as the basis for medical decisions and treatment. Magnetoencephalography (MEG) shows promise in the assessment of the pathophysiological derangements in concussion. The authors have developed a novel MEG-based neuroimaging strategy to provide objective, noninvasive, diagnostic information in neurological disorders. In the current study the authors demonstrate a novel task protocol and then assess MEG virtual recordings obtained during task performance as a diagnostic tool for concussion. METHODS Ten individuals (5 control volunteers and 5 patients with a history of concussion) were enrolled in this pilot study. All participants underwent an MEG evaluation during performance of a language/spatial task. Each individual produced 960 responses to 320 sentence stimuli; 0.3 sec of MEG data from each word presentation and each response were analyzed: the data from each participant were classified using a rule constructed from the data obtained from the other 9 participants. RESULTS Analysis of response times showed significant differences (p < 10(-4)) between concussed and normal groups, demonstrating the sensitivity of the task. The MEG measures enabled the correct classification of 8 of 10 individuals as concussed versus nonconcussed (p = 0.055). Analysis of single-trial data classified 70% of trials correctly (p < 10(-10)). Concussed patients showed increased activation in the occipitoparietal and temporal regions during evaluation. CONCLUSIONS These pilot findings are the first evidence of the utility of MEG virtual recording in diagnosing concussion. With further refinements, MEG virtual recordings may represent a noninvasive test to diagnose concussion and monitor its resolution.

Magnetoencephalography-based identification of functional connectivity network disruption following mild traumatic brain injury

Mild traumatic brain injury (mTBI) leads to long-term cognitive sequelae in a significant portion of patients. Disruption of normal neural communication across functional brain networks may explain the deficits in memory and attention observed after mTBI. In this study, we used magnetoencephalography (MEG) to examine functional connectivity during a resting state in a group of mTBI subjects (n = 9) compared with age-matched control subjects (n = 15). We adopted a data-driven, exploratory analysis in source space using phase locking value across different frequency bands. We observed a significant reduction in functional connectivity in band-specific networks in mTBI compared with control subjects. These networks spanned multiple cortical regions involved in the default mode network (DMN). The DMN is thought to subserve memory and attention during periods when an individual is not engaged in a specific task, and its disruption may lead to cognitive deficits after mTBI. We further applied graph theoretical analysis on the functional connectivity matrices. Our data suggest reduced local efficiency in different brain regions in mTBI patients. In conclusion, MEG can be a potential tool to investigate and detect network alterations in patients with mTBI. The value of MEG to reveal potential neurophysiological biomarkers for mTBI patients warrants further exploration.

Real Time Signal Processing in the Clinical Setting

We routinely use a variety of real time signal acquisition, enhancement, and display techniques in the operating room to provide the surgeon with functional information. This enables reduction of surgical morbidity in cases which present a significant risk to the nervous system. Here we present regression based signal processing algorithms which produce considerable signal-to-noise-ratio enhancement with corresponding reduction in the time required to obtain an interpretable neurophysiological signal. We also present the approach we have applied to fault tolerance and distributed data display for our workstation cluster environment.

Theoretical decomposition of neuronal networks

An approach to the calculation of the unobservable elements of the hippocampal formation is presented. The approach is based on characterizing the transformational properties of this neuronal structure using the kernels of a functional power series expansion. In this approach the network properties are characterized as the composite of input/output functions measured for each subsystem in the network. The multidimensional Laplace transforms are then computed and manipulated to provide an estimate of the transformation properties of the unobservable elements. >

An interactive toolset for characterizing complex neural systems

Abstract A set of two computer programs is described which enable the following functions: 1. (1) High-speed data acquisition from up to eight recording channels with presentation of uniform or quasi-random stimulus trains on one or two stimulus channels. 2. (2) On-line computation and display of averaged responses on all recording channels evoked by both stimulus channels. 3. (3) Computation and display of averaged evoked potentials and functional power series (FPS) through third order for characterization of system nonlinearities. 4. (4) Computation and display of stimulated responses to arbitrarily selected stimulus trains using a previously computed FPS. Both programs are implemented in FORTRAN, include on-line help, interactive and batch processing facilities and produce extensive graphics output according to the Tektronix 4010 protocol. Their initial implementation was on an LSI-II microcomputer using TSX + extensions to RT11. Additional capabilities and speed have been gained by porting the analysis routine to the Apollo workstation environment. To date these programs have been used to study the following: 1. (1) The somatosensory system in the human using dual random trains presented to the median and ulnar nerves. 2. (2) Cognitive processing in the human (normals and first-break schizophrenics) using both uniform (P300 paradigm) and dual random trains of tone bursts with the subject either counting or pressing a lever in response to tones with a specific pitch. 3. (3) The network properties of the rabbit hippocampus (both in vivo and in vitro preparations) using random train simulation of the intrinsic pathways.