BRAVO/TeleTrend: a comprehensive WWW-based neuromonitoring system for the neurosurgery ICU.

This paper describes BRAVO/TeleTrend--a comprehensive client/server-based system for remote access, review and analyses of continuously acquired multiparametric physiological data from Intensive Care unit (ICU) patients. The system is designed as a distributed three tier model and implemented in Java (Sun Microsystems). TeleTrend is a data review package, which interfaces to existing physiological bedside monitors such as the BRAVO suite of products (Nicolet Biomedical, Madison, WI) and the vital signs monitors compatible with the Unity Network (Marquette Electronics, Milwaukee, WI). It does not transfer over the web the entire patient record, which can be hundreds of megabytes. Instead, it provides tools to view a compressed representation of the raw data in a trend display and to zoom into the raw data if needed. Thus, it eliminates the need for a high-bandwidth Internet connection and makes possible the use of a slower modem access to the vast amount of physiological data acquired per patient. In addition, TeleTrend features a rule-based module capable of generating clinical alerts, which is a potentially useful tool for neurointensivists and other critical care personnel. Finally, TeleTrend is intended as a multi-user, semi real-time telemedical application, which features built-in white-board and chat components. These components allow several physicians at different locations around the world to simultaneously view and brainstorm over critical chunks of continuously recorded raw and trend data. By allowing the end-user user to switch on-the-fly from monitoring patients in one ICU to those in another, and by integrating an HL7 interface TeleTrend steps over the boundaries of a single ICU. Thus, it can be provide a medical enterprise-wide solution to the remote access of an important component of the electronic patient medical record. Currently in house validation, verification and alpha testing of the system are underway.

Segmentation analysis in functional MRI: activation sensitivity and gray-matter specificity of RARE and FLASH.

Brain activation is accompanied by local decreases in vascular deoxyhemoglobin. Theoretically, gradient-echo and spin-echo sequences show similar sensitivity to capillary deoxyhemoglobin, but spin-echo sequences should be less sensitive to venous deoxyhemoglobin. This is an important distinction in the context of cortical localization. We report herein a direct experimental comparison of a gradient-echo sequence (fast low-angle shot [FLASH]) with a spin-echo sequence (rapid acquisition with relaxation enhancement [RARE]) for functional MRI (fMRI) in seven subjects undergoing visual stimulation. A Student t test analysis was used to locate areas of significant activation, and then computerized image segmentation was performed to determine the type of activated tissue. Contrary to previous reports, both sequences proved equally sensitive to overall activation. RARE activation, however, was more specific for gray matter, as suggested by prior theoretical models.

Functional MR and PET imaging of rolandic and visual cortices for neurosurgical planning.

Magnetic resonance (MR) imaging has recently been used to demonstrate physiological activation of the human brain. This development is of considerable interest to the neurosurgeon planning procedures near brain regions involving specific functions. In the present study, rolandic and visual cortices were imaged with a conventional 1.5-tesla clinical MR imager using a spoiled gradient-recalled acquisition in the steady state sequence. Two patients, one with a right frontal astrocytoma and the other with a left parietal meningioma, underwent MR imaging of rolandic cortex while performing a repetitive finger apposition task. Two patients with complex partial seizures referable to the temporal and occipital regions underwent MR imaging of visual cortex while exposed to repetitive photic stimulation (8.3 Hz). Significant signal intensity changes up to 15% between the activation and rest conditions were observed near the surgical targets at the expected anatomical location of the rolandic and visual cortices. In two of these cases activation measured by MR was compared and found similar to the activation measured at the same plane by H2(15)O positron emission tomography (PET). These results suggest that functional MR and PET techniques can be used to obtain preoperative brain mapping in individual patients considered for neurosurgical procedures.

Computer applications in the intensive care unit.

Incorporating computers into all aspects of daily ICU operations is a formidable task both technically and logistically. To coordinate a project as complex as this, it is imperative to achieve close cooperation between physicians, nurses, basic scientists, computer specialists, hospital administrators, and equipment manufacturers. In this article, we have provided a blueprint and have discussed the implementation of a COmputational Model of PATient Health Status (COMPATHS) designed to carry out such integration.

The neurosurgical intensive care unit in an era of health care reform.

Health care reform, public concern, and managed care will create an environment that demands highly creative strategies to deliver quality care while reducing costs. Patient satisfaction and outcomes will take on a high priority. To meet this challenge, the neurosurgical ICU of the future will be designed with a patient-focused theme wherein the physical environment embodies healing and humanism. Services will be brought to the patient rather than the patient accommodating the system. Patients and families will be the directors of their own care. Staff and families will have access to a highly sophisticated clinical information system, and learning for staff at all levels will be a part of everyday life in the ICU. Unit management will be within a framework of shared governance wherein the power base is with the direct care givers, and decision and policy making happens at the point closest to the patient. Patient outcomes will be a result of a highly organized collaborative model that includes primary nursing, critical paths, and case management. Partnerships between nurses and unit support staff will create skill-mix changes that allow the nurse to spend less time on nonclinical unit maintenance-type functions and more time with the patient and family. This will have a positive fiscal impact as well as enhance patient satisfaction and outcomes.

Role of inhibition in memory retrieval by hippocampal area CA3.

A network model with some general properties of hippocampal area CA3, and the results of its simulation on a massively parallel processor, are described. This network performs the tasks of recent declarative memory including recovery of complete traces from partial cues and recognition of familiarity. Immediate recurrent inhibition is essential for providing sensitivity to small cues while preventing spurious recall. Tonic inhibition seems to set the retrieval speed/accuracy trade-off. Delayed inhibition resets hippocampal activity. The behavior under excessive or deficient inhibition resembles that of amnesics with lesions in brainstem areas known to modulate hippocampal inhibition. The rate of recall and dynamics of inhibition by the model are similar to those inferred to occur in the human hippocampus from unit and evoked potential recordings. The model suggests a mechanism whereby the hippocampus can control its own plasticity. These simulations demonstrate that the retrieval mechanism in several hippocampal models is feasible and that the theta rhythm and the cognitive evoked potentials may be generated by synaptic events modulating network parameters.

Human brain metabolic responses to familiarity during lexical decision.

Local cerebral glucose metabolic rate was measured using F18-DG/PET in 11 normal righthanded male volunteers performing two versions of a visual lexical decision task differing only with respect to whether the stimuli were all novel and shown only once, or whether they were all repeated many times. Each subject was also scanned in a resting state (eyes open and ears unplugged). Scalp recordings of cognitive evoked potentials in the same subjects and tasks confirmed that the effects of familiarity were present during the entire 30 min FDG uptake period. The N4 (a left temporoparietal negativity 400 ms after stimulus onset) decreased with repetition, and the P3b (a 600 ms latency parietal positivity) increased. The pattern of metabolic activation was also significantly different between the three experimental conditions. Planned comparisons of the tasks with the resting condition revealed significant relative activations of visual and motor areas. Novel words evoked significantly greater activation in the left and right anterior cingulate gyri and right hippocampal formation than did repeated words in the same task. Relative glucose metabolism in the left angular gyrus was significantly greater to novel words than to resting. Thus, two tasks equated in sensory, motor, and decision processes, but differing in the familiarity of the stimuli, evoke significantly different patterns of brain activation. The patterns observed provided evidence that recent episodic memory for words engages a system that includes the bilateral anterior cingulate gyri, the medial temporal lobe, and the left angular gyrus. © 1994 Wiley-Liss, Inc.

Localized brain metabolic response correlated with potentials evoked by words.

Evoked potentials (EPs) measure synaptic current flows that propagate from brain to scalp, Alternatively, positron emission tomography (PET) using fluoro-deoxyglucose (FDG) can measure the increased glucose metabolism supporting this synaptic activation. It is difficult to localize the brain activity-generating EPs from their scalp distribution, because activity originating in different regions tends to produce overlapping scalp topographies. In contrast, FDG-PET provides better spatial resolution for activity throughout the brain, but shows only the total metabolism integrated over a 30-min uptake period. We combined the temporal and psychological resolution of EPs with the spatial resolution of PET to help define when and where in the brain words are encoded for meaning.