Use of a Neural Network as a Predictive Instrument for Length of Stay in the Intensive Care Unit Following Cardiac Surgery

A patient's intensive care unit (ICU) length of stay following cardiac surgery is an important issue in Canada, where cardiovascular intensive care resources are limited and waiting lists for cardiac surgery exist. We trained a neural network with a database of 713 patients and 15 input variables to predict patients who would have a prolonged ICU length of slay, defined as a stay greater than 2 days. In an independent test set of 696 patients, the network was able to stratify patients into three risk groups for prolonged stay (low, intermediate, and high), corresponding to frequencies or prolonged stay of 16.3, 35.3, and 60.86%, respectively. The trained network could potentially be used as a predictive instrument for optimizing the scheduling of cardiac surgery patients in times of limited ICU resources. Neural networks are a new method for developing predictive instruments that offer both advantages and disadvantages when compared to other more widely used statistical techniques.     

Technical description of the IBIS data library. Improved Monitoring for Brain Dysfunction in Intensive Care and Surgery

The IBIS Data Library (DL) is an annotated data library that contains practically all the monitored data and other clinical information from critically ill patients during surgery and in intensive care. The data have been collected at three sites: the intensive care unit of the Kuopio University Hospital, Finland; Royal Brompton Hospital, London, UK; and St. Bartholomew's Hospital, London, UK. The purpose of the DL is to form the basis for development of biosignal interpretation methods in the Improved Monitoring for Brain Dysfunction in Intensive Care and Surgery project in the European Union (EU) BIOMED2 programme (BMH4-97-2570). The DL contains continuous electroencephalography signals, multimodal evoked potential recordings and diagnostic electrocardiography recorded during intensive care and surgery. In addition, signal types similar to those recorded during an earlier project, the EU-BIOMED1 project IMPROVE, are stored in the DL. In addition, trend data from patient monitors, laboratory data, annotations, nursing actions, and medications recorded and stored by a Patient Data Management System (PDMS) during routine care are included. The data obtained routinely are complemented by special annotations made by a physician who observes the patient during the data collection session. Annotations include, for example, assessment of the awareness of the patient and specific events during surgery not recorded routinely by the PDMS. Inclusion of information about the care plan and the aims of the care make the contents of the DL complete. The present paper describes the technical set-up used for recording of the DL and the contents of the DL. The paper also includes an appendix defining a new data format, the extended evoked potentials format, used for storage of sweep data in the DL.     

Exploring the IBIS data library contents: tools for data visualisation, (pre-) processing and screening. Improved Monitoring for Brain Dysfunction in Intensive Care and Surgery

During the IBIS project a high-quality data library of continuous and intermittent physiological signals and variables from patients during intensive care and surgery has been collected. To facilitate exploration of the full content of this data library a data browser was developed, which offers a flexible graphical display of the collection of multivariate data. To supplement the functionality of the display of the 'raw' data, a set of screening and pre-processing tools has been developed. A separate trend analysis tool offers a convenient overview of an entire recording focusing on the slow changes in the general state of the patient and the interaction between different physiological subsystems seen from a long-term perspective. A frequency analysis tool for processing the electroencephalography (EEG) signals has been integrated in the data browser to facilitate a quick screening of the cerebral function. The data library is the foundation of the development and validation of biosignal interpretation methods. This process can potentially be more productive using the described tool for algorithm prototyping based on a graphical network specifying the interaction between data processing primitives.     

Clinical Decision Support and Closed‐Loop Control for Intensive Care Unit Sedation

Critical care patients undergoing surgery require drug administration to regulate physiological variables such as blood pressure, cardiac output, heart rate, and degree of consciousness. The rate of infusion of each administered drug is critical, requiring constant monitoring and frequent adjustments. Patients in the intensive care unit who require mechanical ventilation due to acute respiratory failure also frequently require the administration of sedative agents. Open‐loop control (manual control) by clinical personnel can be tedious, imprecise, time‐consuming, and sometimes of poor quality, depending on the skills and judgment of the clinician. Dynamical system pharmacokinetic and pharmacodynamic modeling and closed‐loop control system design methodologies can significantly advance our understanding of the wide effects of pharmacological agents and anesthetics, as well as advance the state‐of‐the‐art in active control of drug delivery systems for clinical pharmacology. In this paper, we discuss the challenges and opportunities of clinical decision support and closed‐loop control for intensive care unit sedation.     

The IBIS project: data collection in London. Improved Monitoring for Brain Dysfunction during Intensive Care and Surgery

The primary aim of the Improved Monitoring for Brain Dysfunction during Intensive Care and Surgery (IBIS) project was to create a unique and comprehensively annotated data library (DL) of multiple physiological, including neurophysiological, signals. Data collection was undertaken in Kuopio, Finland and London, UK, and comparable protocols were used at all the sites. In London, 43 patients were recruited at the Royal Brompton Hospital, followed by nine at St. Bartholomew's Hospital, all of whom underwent cardiac or combined cardiac and carotid artery surgery. Thirty-seven patients underwent a single operation, while 15 underwent two procedures. The protocols and equipment used, problems specific to the electrically hostile environment and preliminary results are described, including those of clinical interest. The DL is being used for the development of clinically applicable neurophysiological monitoring tools.     

Detection of artifacts in monitored trends in intensive care

In intensive care, decision-making is often based on trend analysis of physiological parameters. Artifact detection is a pre-requisite for interpretation of trends both for clinical and research purposes. In this study, we developed and tested three methods of artifact detection in physiological data (systolic, mean and diastolic artery and pulmonary artery pressures, central venous pressure, and peripheral temperature) using pre-filtered physiological signals (2-min median filtering) from 41 patients after cardiac surgery. These methods were: (1) the Rosner statistic; (2) slope detection with rules; and (3) comparison with a running median (median detection). After tuning the methods using data from 20 randomly chosen patients, the methods were tested using the data from the remaining patients. The results were compared with those obtained by manual identification of artifacts by three senior intensive care unit physicians. Out of an average of 22,480 data points for each variable, the three observers labelled 0.98% (220 data points) as artifacts. The inter-observer agreement was good. The average (range) sensitivity for artifact detection in all variables in the test database was 66% (33-92%) for the Rosner statistic, 64% (24-98%) for slope detection and 72% (41-98%) for median detection. All methods had a high specificity (> or = 94%). Slope detection had the highest mean positive prediction rate (53%; 21-85%). When the performance was measured by the cost function, slope detection and running median performed equally well and were superior to Rosner statistics for systemic arterial and central venous pressure and peripheral temperature. None of the methods produced acceptable results for pulmonary artery pressures. We conclude that median filtering of physiological variables is effective in removing artifacts. In post-operative cardiac surgery patients, the remaining artifacts are difficult to detect among physiological and pathophysiological changes. This makes large databases for tuning artifact algorithms mandatory. Despite these limitations, the performance of running median and slope detection were good in selected physiological variables.     

Parameter identification and sedative sensitivity analysis of an agitation-sedation model

Sedation administration and agitation management are fundamental activities in any intensive care unit. A lack of objective measures of agitation and sedation, as well as poor understanding of the underlying dynamics, contribute to inefficient outcomes and expensive healthcare. Recent models of agitation-sedation dynamics have enhanced understanding of the underlying dynamics and enable development of advanced protocols for semi-automated sedation administration. In this research, the agitation-sedation model parameters are identified using an integral-based fitting method developed in this work. Parameter variance is then analysed over 37 intensive care unit patients. The parameter identification method is shown to be effective and computationally inexpensive, making it suited to real-time clinical control applications. Sedative sensitivity, an important model parameter, is found to be both patient-specific and time-varying. However, while the variation between patients is observed to be as large as a factor 10, the observed variation in time is smaller, and varies slowly over a period of days rather than hours. The high fitted model performance across all patients show that the agitation-sedation model presented captures the fundamental dynamics of the agitation-sedation system. Overall, these results provide additional insight into the system and clinical dynamics of sedation management.     

A portable device for intensive care brain function monitoring with event-related potentials

Monitoring level of consciousness or depth of sedation is essential in modern intensive care units and emergency rooms. Current methods are based on spontaneous EEG providing only indirect information on the reactivity of the brain. Measurement of auditory event-related potentials (ERPs) has been shown to have additional potential for evaluation of the level of consciousness. Unfortunately, compact and easy-to-use devices are not commercially available. In this study a portable battery-powered device for clinical auditory ERP measurements was designed, constructed and validated. The device consists of a five-channel data logger and a 16-bit stereo audio stimulator. The signals are digitized with a 22-bit sigma-delta analogue-to-digital converter and stored to a PC Card hard disk. Prior to the in vivo application, the device was validated with extensive technical tests. Importantly, the RMS noise amplitude of the EEG channels was found to be less than 1 mivroV and the delivered auditory stimulus intensity corresponded accurately the settings (mean difference 0.2+/-0.5 dB). In addition to technical tests the device was successfully validated in vivo. To summarize, a novel portable instrument for measurement of auditory event-related potentials in intensive care environment is introduced.     

Cluster analysis of clinical data measured in the surgical intensive care unit

A set of 13 extensively used hemodynamic, ventilatory and gas analysis variables are measured (on-line or off-line) on 200 patients in an intensive care unit (ICU) during the 6 h immediately following cardiac surgery. In order to identify both low- and high-risk patterns, a clustering method is applied to these data at three equidistant observation times. Application of the divergence criterion allows a quantitative evaluation of the diversity between the clusters identified, showing that the two patterns are really distinct in the 13-D space. The same criterion is then used to find possible subsets of variables capable of maintaining, in time, an effective separation power. The latter always include the cardiac index (CI), representative of cardiac performance, and two indices related to respiratory efficiency and metabolic rate, i.e., the carbon dioxide production index (VCO2I) and the arterio-venous oxygen difference (avO2D).     

Estimation of systemic vascular bed parameters for artificial heartcontrol

An extended Kalman filter estimator for the identification of systemic circulation model parameters during cardiac ejection and cardiac filling is described. The estimator has been developed for use in the control of a cardiac ventricular assist device. A lumped element circuit with a time-varying capacitor was used to represent the systemic circulation and the left ventricle. Since the haemodynamic variables that are measurable in patients with impaired cardiac function vary dramatically as the patients move through different levels of care, the estimator was designed so that it can be used with different sets of blood pressure and flow measurements. Preliminary evaluation of the performance of the estimator using data from a computer simulation and from a patient during open-heart surgery is presented. The robustness of the estimator to variations in parameter initialization is also described     

PACS in Töölö hospital

A 3-year PACS project was started in 1997 and completed in 1999 with filmless radiology and surgery. An efficient network for transferring images provides the infrastructure for integration of different distributed imaging systems and enables efficient handling of all patient-related information on one display station. Because of the need for high-speed communications and the massive amount of image data transferred in radiology, ATM (25, 155 Mbit/s) was chosen to be the main technology used. Both hardware and software redundancy of the system have been carefully planned. The size of the Dicom image library utilizing MO discs is currently 1.2 TB with 300 GB RAID capacity. For the increasing amount of teleradiologic consultations, a special Dicom gateway is planned. It allows a centralized and resilient handling and routing of received images around the hospital. Hospital-wide PACS has already improved the speed and quality of patient care by providing instant access to diagnostic information at multiple locations simultaneously. The benefits of PACS are considered from the viewpoint of the entire hospital: PACS offers a method for efficiently transporting patient-related images and reports to the referring physicians.     

Adaptive coordination in cardiac anaesthesia: a study of situational changes in coordination patterns using a new observation system

Patient care in hospital settings requires coordinated team performance. Studies in other industries show that successful teams adapt their coordination processes to the situational task requirements. This prospective field study aimed to test a new observation system and investigate patterns of adaptive coordination within operating room teams. A trained observer recorded coordination activities during 24 cardiac surgery procedures. The study tested whether different patterns occur during different phases of and between different types of surgical procedures (two-way multivariate ANOVA with repeated measure). A statistically significant increase was found in clinical and coordination activities in phases of the operation with high task interdependence. The highest level of 'coordination via the work environment' (i.e. an implicit coordination mechanism) was recorded during the actual procedure on the beating heart. These findings prove the sensitivity of the observation system developed and evaluated in this study and provide insight into patterns of adaptive coordination in cardiac anaesthesia. This study furthers our understanding of adaptive coordination as a cornerstone of effective team performance in complex work environments. Using a new observation system, it describes patterns employed by health care professionals in response to changing task demands in an acute patient care setting.     

Proteomics in paediatric cardiac surgery: Is a personalised approach feasible?

The incidence of congenital cardiac abnormalities remains high. Paediatric patients with congenital cardiac defects often require surgery at a young age. The surgeries are often long and complex, rendering this population particularly vulnerable to the deleterious effects of cardiopulmonary bypass and cardiac surgery. The search for cardioprotective strategies is ongoing in an attempt to reduce the morbidity in this population. In the post-genomic era, it is apparent that simply determining the genomic sequences holds little diagnostic potential and means to determine progression of disease and response to treatment. The field of proteomics is expanding and application of proteomic techniques in the clinical setting holds great potential to advance our understanding of the proteomic changes involved in specific disease stages. This review will assess the application of proteomic techniques in the setting of paediatric cardiac surgery and highlight the need to obtain a clear understanding of the role of various proteins in children with cardiac conditions. The success and challenges of the available proteomic technology will be discussed as well as the future potential of proteomic methods for advancing our understanding of protein changes in children requiring cardiac surgery.     

Autonomic information flow improves prognostic impact of task force HRV monitoring

Heart rate variability (HRV) represents the cardiovascular control mediated by the autonomic nervous system and other mechanisms. In the established task force HRV monitoring different cardiovascular control mechanisms can approximately be identified at typical frequencies of heart rate oscillations by power spectral analysis. HRV measures assessing complex and fractal behavior partly improved clinical risk stratification. However, their relationship to (patho-)physiology is not sufficiently explored. Objective of the present work is the introduction of complexity measures of different physiologically relevant time scales. This is achieved by a new concept of the autonomic information flow (AIF) analysis which was designed according to task force HRV. First applications show that different time scales of AIF improve the risk stratification of patients with multiple organ dysfunction syndrome and cardiac arrest patients in comparison to standard HRV. Each group's significant time scales correspond to their respective pathomechanisms.     

基于多元统计分析的变电建筑工程造价因素研究

从实际工程数据出发,以设计方案因子为切入点,按主成分和多元回归的方式,识别变电建筑的主要因素,初步建立科学合理的全局快速估价模型.     

Single sweep analysis of event related auditory potentials for the monitoring of sedation in cardiac surgery patients

Event-related potentials (ERPs) from the auditory system were investigated in 28 post-operative cardiac patients in order to assess their relevance in the monitoring of patient sedation level. Midazolam (17 patients) and propofol (11 patients) were the sedative agents used. The auditory ERP components of N100 (HAB100) and mismatch negativity (MMN) were considered. A single sweep method based on the AutoRegressive with eXogenous input (ARX) model, which is able to enhance the evoked responses to each single stimulus, was used to process each sweep and to compute traditional parameters on a sweep-by-sweep basis. Differences in the measured parameters were related to variations in the patient sedation levels classified through Ramsay score. Significant differences (P<0.05) in both MMN and HAB100 parameters were found between light sedation (LS) and deep sedation (DS) levels.     

Neural network cost estimates for heart bypass surgery

This paper reports on the results of using artificial neural network (ANN) technology to estimate treatment costs of heart bypass patients based on their diagnostic condition and clinical criteria. Our applications include: (1) predicting total episode cost using clinical data; (2) a method for providing rapid feedback to assess change in total costs within a turbulent environment; and (3) a procedure for identifying activity-based cost driver candidates that would normally not surface from an analysis of accounting data. Clinical data were collected on 250 heart bypass patients at the University of Ottawa Heart Institute. The data analysed support the following conclusions: (1) clinical and diagnostic indicators obtained before surgery for individual patients can be used to estimate the total cost of their heart bypass surgery; (2) the average cost estimate error decreases as we add clinical information available during and after the surgical event; (3) the procedure used to estimate individual patient cost does not require access to accounting records; (4) the forecasting system we describe may improve exception reporting for individual patients by tracking costs and clinical criteria on a real-time basis throughout the treatment episode.     

基于动态库的社会医疗保险数据实时上传接口设计与探讨

基于动态链接库方式,实现医院信息系统实时与医保信息系统进行数据交换,保证医保病人费用不用二次录入,医院库存数据、病人费用数据实时上传,在异地医保端记费扣库存,从而使两个相对独立的系统数据保持一致,建立更加完善的HIS和医保系统之间的接口。     

Making Library Screencast Tutorials: Factors and Processes

Many library tutorials are currently being created by using screencasts, a form of web-based streaming media. This case study describes a tutorial project undertaken by a group of three librarians who had not had previous experience in this area. The article explains the planning, documentation, and production processes involved with making the tutorials as well as the multiple roles that emerged. It outlines the review process that involved project supervisors and the input of library staff from various departments. Upon project completion, the team revisited matters that needed to be a part of the initial process: marketing, faculty buy-in, and assessment.     

The transformation of surgery patient care with a clinical research information system

Implementing a computerized clinical research information system (CRIS) can make clinical research easier and more efficient while improving patient care by providing surgeons with performance feedback. To transform the original manual patient information management system so it delivers patient care, this study developed a CRIS for cardiovascular disease to facilitate surgery treatment tracking. The CRIS tracks hundreds of pieces of data through surgical stages and converts these data into computerized registries, provides surveillance mechanisms, and generates clinical interpretive reports in a timely manner. Surgeons can use the CRIS to identify surgical-related data and interventional cardiovascular procedure risks based on specific patient characteristics, and it has increased the quality and efficiency of patient care. An intelligent data analysis (IDA) tool based on the Weka library that seamlessly integrates with the CRIS has helped provide models for clinical research.