Shermeen Nizami

Implementation of Artifact Detection in Critical Care: A Methodological Review

Artifact detection (AD) techniques minimize the impact of artifacts on physiologic data acquired in critical care units (CCU) by assessing quality of data prior to clinical event detection (CED) and parameter derivation (PD). This methodological review introduces unique taxonomies to synthesize over 80 AD algorithms based on these six themes: 1) CCU; 2) physiologic data source; 3) harvested data; 4) data analysis; 5) clinical evaluation; and 6) clinical implementation. Review results show that most published algorithms: a) are designed for one specific type of CCU; b) are validated on data harvested only from one OEM monitor; c) generate signal quality indicators (SQI) that are not yet formalized for useful integration in clinical workflows; d) operate either in standalone mode or coupled with CED or PD applications; e) are rarely evaluated in real-time; and f) are not implemented in clinical practice. In conclusion, it is recommended that AD algorithms conform to generic input and output interfaces with commonly defined data: 1) type; 2) frequency; 3) length; and 4) SQIs. This shall promote: a) reusability of algorithms across different CCU domains; b) evaluation on different OEM monitor data; c) fair comparison through formalized SQIs; d) meaningful integration with other AD, CED and PD algorithms; and e) real-time implementation in clinical workflows.

Heart disease classification through HRV analysis using Parallel Cascade Identification and Fast Orthogonal Search

Heart rate variability (HRV) is an established indicator of cardiac health. Recent developments have shown the potential of nonlinear metrics for pattern classification of various heart conditions. Evidence indicates that the combination of multiple linear and nonlinear features leads to increased classification accuracy. In our paper, we demonstrate HRV classification using two dynamic nonlinear techniques called Parallel Cascade Identification (PCI) and Fast Orthogonal Search (FOS). We investigate the use of these two techniques for feature extraction from publicly available Physionet electrocardiogram (ECG) data to differentiate between normal sinus rhythm of the heart and 3 undesired conditions: arrhythmia, supraventricular arrhythmia, and congestive heart failure. Results compare well with previous studies which have used more features over the same dataset. We hypothesize that combining PCI and FOS features with traditional HRV features will show further improvement in classification accuracy and so can assist in real-time patient monitoring.

Service oriented architecture to support real-time implementation of artifact detection in critical care monitoring

The quality of automated real-time critical care monitoring is impacted by the degree of signal artifact present in clinical data. This is further complicated when different clinical rules applied for disease detection require source data at different frequencies and different signal quality. This paper proposes a novel multidimensional framework based on service oriented architecture to support real-time implementation of clinical artifact detection in critical care settings. The framework is instantiated through a Neonatal Intensive Care case study which assesses signal quality of physiological data streams prior to detection of late-onset neonatal sepsis. In this case study requirements and provisions of artifact and clinical event detection are determined for real-time clinical implementation, which forms the second important contribution of this paper.

Comparing metrological properties of pressure-sensitive mats for continuous patient monitoring

Pressure-sensitive mat (PSM) technology offers several advantages as a sensor modality for patient monitoring since it is non-contact and unobtrusive. However, as we move to deploy PSM for long-term continuous patient monitoring, we must consider and characterize their metrological properties that arise due to their electrical, mechanical or optical construction. We evaluate the dynamic metrological properties of rise time, creep, percent change in creep, drift, and repeatability for three different PSM technologies from three vendors, namely, S4 (Kinotex fiber-optics), Tekscan (resistive ink), and XSensor (capacitive). Both long-term (14.5 hrs) and repeated short-term experiments (1 min) were conducted using two anthropometric models exhibiting contact pressures representative of adult and neonatal patients. Long-term experiments were conducted to characterize rise time, creep, percent change in creep, and drift for each sensor. With both pressure models, the XSensor exhibited the fastest dynamic response in terms of rise and recovery times, while Tekscan exhibited the slowest responses. S4 and Tekscan present with an expected decrease in drift with application of the adult model, but XSensor shows the opposite trend. Short-term experiments were conducted to measure repeatability with four application-removal repetitions for 1 min each. The coefficient of variation (CoV) was computed for each sensor as a measure of repeatability. For both pressure models, the smaller CoV of XSensor implies greater repeatability and hence, greater reliability.

Integrative learning through the design of an electrocardiogram acquisition system

This paper presents an electrocardiogram acquisition system course design project for an upper year bioinstrumentation course. The objective of this design project is to provide students an opportunity for an integrative learning, enhancing their educational experience. Unlike similar electrocardiogram instrumentation projects, this project is based on a commercially available instrumentation amplifier enabling better systems-level thinking. The project is described along with observations made from our pilot implementation of the project. The initial offering of the project is considered a success with positive feedback from students. Recommendations for improvements are also discussed.

Comparing time and frequency domain estimation of neonatal respiratory rate using pressure-sensitive mats

Pressure-sensitive mats (PSM) have proved to be useful in the estimation of respiratory rates (RR) in adult patients. However, PSM technology has not been extensively applied to derive physiologic parameters in infant and neonatal patients. This research evaluates the applicability of the capacitive XSensor PSM technology to estimate a range of RR in neonatal patient simulator trials conducted under several experimental conditions. PSM data are analyzed in both the time and frequency domain and comparative results are presented. For the frequency-domain approach, in addition to estimating RR, a measure of confidence is also derived from the relative height of peaks in the periodogram. The study demonstrates that frequency domain analysis of mean-shifted PSM data achieves the best possible RR estimation, with zero percent error, as compared to the lowest achievable RMS error of 1.57 percent in the time domain. The frequency domain approach outperforms the time domain analysis whether examining raw data or those preprocessed by normalizing, detrending and median filtering.

An artifact detection framework for clinical decision support systems

This research develops a standardized framework to integrate artifact detection (AD) in computerized Clinical Decision Support Systems (CDSS). Review of the state of the art has revealed a number of limitations currently preventing the widespread implementation of AD algorithms within CDSS. To address those limitations, this paper develops a novel component-based AD framework for integration in CDSS. The novelty of this research is the development of a Common Reference Model (CRM) with standard definitions for component interfaces. These definitions include common physiologic data attributes of: (1) type; (2) frequency; (3) length; and (4) Signal Quality Indicator.

A high-capacity scheduling algorithm for systems employing embedded modulation

The paper proposes a simple scheduling algorithm based on adaptive embedded modulation (EM) for the downlink in urban wireless networks. Embedded modulation allows multiple users to share a single downlink channel simultaneously. The proposed EM scheduling algorithm embeds the user which has the largest carrier-to-interference ratio (C/I) with each of the lower C/I users in turn. Simulation results show that the EM scheduler, employing adaptive coded modulation, can improve the average throughput of a given downlink by up to 33% over a round robin scheduler, while still maintaining acceptable fairness amongst the system users.