Thomas M. McKenna

A method for automatic identification of reliable heart rates calculated from ECG and PPG waveforms.

OBJECTIVE The development and application of data-driven decision-support systems for medical triage, diagnostics, and prognostics pose special requirements on physiologic data. In particular, that data are reliable in order to produce meaningful results. The authors describe a method that automatically estimates the reliability of reference heart rates (HRr) derived from electrocardiogram (ECG) waveforms and photoplethysmogram (PPG) waveforms recorded by vital-signs monitors. The reliability is quantitatively expressed through a quality index (QI) for each HRr. DESIGN The proposed method estimates the reliability of heart rates from vital-signs monitors by (1) assessing the quality of the ECG and PPG waveforms, (2) separately computing heart rates from these waveforms, and (3) concisely combining this information into a QI that considers the physical redundancy of the signal sources and independence of heart rate calculations. The assessment of the waveforms is performed by a Support Vector Machine classifier and the independent computation of heart rate from the waveforms is performed by an adaptive peak identification technique, termed ADAPIT, which is designed to filter out motion-induced noise. RESULTS The authors evaluated the method against 158 randomly selected data samples of trauma patients collected during helicopter transport, each sample consisting of 7-second ECG and PPG waveform segments and their associated HRr. They compared the results of the algorithm against manual analysis performed by human experts and found that in 92% of the cases, the algorithm either matches or is more conservative than the humans QI qualification. In the remaining 8% of the cases, the algorithm infers a less conservative QI, though in most cases this was because of algorithm/human disagreement over ambiguous waveform quality. If these ambiguous waveforms were relabeled, the misclassification rate would drop from 8% to 3%. CONCLUSION This method provides a robust approach for automatically assessing the reliability of large quantities of heart rate data and the waveforms from which they are derived.

Decision tool for the early diagnosis of trauma patient hypovolemia

We present a classifier for use as a decision assist tool to identify a hypovolemic state in trauma patients during helicopter transport to a hospital, when reliable acquisition of vital-sign data may be difficult. The decision tool uses basic vital-sign variables as input into linear classifiers, which are then combined into an ensemble classifier. The classifier identifies hypovolemic patients with an area under a receiver operating characteristic curve (AUC) of 0.76 (standard deviation 0.05, for 100 randomly-reselected patient subsets). The ensemble classifier is robust; classification performance degrades only slowly as variables are dropped, and the ensemble structure does not require identification of a set of variables for use as best-feature inputs into the classifier. The ensemble classifier consistently outperforms best-features-based linear classifiers (the classification AUC is greater, and the standard deviation is smaller, p<0.05). The simple computational requirements of ensemble classifiers will permit them to function in small fieldable devices for continuous monitoring of trauma patients.

ROLE OF NITRIC OXIDE IN SEPSIS-INDUCED HYPOREACTIVITY INISOLATED RAT LUNGS

The aim of the present study was to test the hypothesis that pulmonary microvascular reactivity is depressed in sepsis and that inducible nitric oxide synthase (iNOS) contributes to the vascular hyporeactivity. Rats were made septic by cecal ligation and puncture. After 16 h, pulmonary vascular reactivity was evaluated by measurement of perfusion pressures while the vasculature was challenged with angiotensin II and KCI. The results showed that vascular reactivity was significantly depressed in lungs from septic rats in comparison to sham-operated controls. Pretreatment with the nitric oxide synthase inhibitor NG-nitro-L-arginine methyl ester (L-NAME, 100 μM) restored the depressed vasoreactivity while the nitric oxide (NO) synthase substrate L-arginine (1 mM) reversed the contraction-restoring effect of L-NAME. NO production in lungs from septic rats increased about 4-fold in comparison to sham-operated controls. iNOS protein was expressed in lung tissues, mainly the resistance vessels, from septic rats but not from sham-operated controls. Reverse transcription and polymerase chain reaction also showed a strong induction of iNOS mRNA in lung tissues from septic rats. These results suggest that increased iNOS expression and NO production may contribute to depressed pulmonary vascular reactivity in sepsis.

Overexpression of protein kinase Cα enhances lipopolysaccharide-induced nitric oxide formation in vascular smooth muscle cells

Our previous studies showed that lipopolysaccharide (LPS)‐induced nitric oxide (NO) synthesis in cardiovascular tissues is attenuated by protein kinase C (PKC) inhibitors. In the current study, we identify a specific PKC isotype involved in the LPS signal transduction pathway that leads to NO formation in rat vascular smooth muscle cells (VSMC). VSMC were transfected with a mammalian expression vector containing a full length PKCα cDNA insert, and a stable transfectant overexpressing PKCα was obtained as evidenced by increased expression of PKCα mRNA and protein. In response to 100 ng/ml LPS stimulation, the PKCα transfectants showed a 1.8‐fold increase in PKC activity at 30 min and a twofold increase in NO production over 24 hr compared with cells transfected with control plasmids. The LPS‐stimulated increase in NO synthesis in PKCα transfectants was blocked by the specific PKCα inhibitor Gö 6976. After 6 hr LPS treatment, PKCα‐transfected and control cells showed equivalent increases in mRNA and protein for the inducible NO synthase. NO synthase activity of the cell extracts assayed in the presence of excess substrate and cofactors was not significantly different between PKCα‐transfected and control cells after LPS stimulation. However, mRNA levels for GTP cyclohydrolase I, a key enzyme in (6R)‐tetrahydro‐L‐biopterin synthesis, and cationic amino acid transporter‐2, involved in L‐arginine transport, was enhanced in cells overexpressing PKCα compared with control cells. These results suggest that PKCα plays an important role in LPS‐induced NO formation and that a significant portion of this effect may be by means of enhanced substrate availability to the inducible nitric oxide synthase enzyme. J. Cell. Physiol. 176:402–411, 1998.

The physiology analysis system: An integrated approach for warehousing, management and analysis of time-series physiology data

The physiology analysis system (PAS) was developed as a resource to support the efficient warehousing, management, and analysis of physiology data, particularly, continuous time-series data that may be extensive, of variable quality, and distributed across many files. The PAS incorporates time-series data collected by many types of data-acquisition devices, and it is designed to free users from data management burdens. This Web-based system allows both discrete (attribute) and time-series (ordered) data to be manipulated, visualized, and analyzed via a clients Web browser. All processes occur on a server, so that the client does not have to download data or any application programs, and the PAS is independent of the clients computer operating system. The PAS contains a library of functions, written in different computer languages that the client can add to and use to perform specific data operations. Functions from the library are sequentially inserted into a function chain-based logical structure to construct sophisticated data operators from simple function building blocks, affording ad hoc query and analysis of time-series data. These features support advanced mining of physiology data.

Protein Kinase C is a Mediator of Lipopolysaccharide-Induced Vascular Suppression in the Rat Aorta

Treatment of vascular tissue with lipopolysaccharide (LPS) in vitro induces hyporesponsiveness to contractile agonists. We investigated whether protein kinase C (PKC) transduces the LPS signal into contractile dysfunction. Rat aortic tissue was incubated .5–18 h with LPS (10 or 30 ng/mL) or α- and β-phorbol 12,13-dibutyrate (PDB, .1 or 1 μ), either alone or combined with cycloheximide (50 μ) or the kinase inhibitors sphingosine (20 μ), H7 (1-(5-isoquinolinylsulfonyl)-2-methyl piperazine, 25 μ), and HA1004 (N-(2-guanidinoethyl)-5-isoquinolinesulfonamide, 25 μ). LPS and β-PDB induced a sustained translocation of PKC activity from the cytosol to the membrane, an increased protein synthesis-dependent expression of nitric oxide synthase (NOS) activity, and an impaired contractility that could be partially reversed by treatment with the NOS inhibitor Nω-nitro-L-arginine methyl ester. Incubation with α-PDB, an inactive isomer of β-PDB, did not alter any of the tissue functions. Sphingosine blocked LPS- and β-PDB-induced NOS activity and LPS-induced impairments in tissue contractility and PKC translocation. Incubation with H7 also protected against LPS-induced vasoplegia, while HA1004, used as a negative control for H7, provided little protection against LPS. These data indicate that PKC plays a role as an intracellular mediator of LPS-induced NOS activity and vascular suppression.

A fuzzy logic algorithm to assign confidence levels to heart and respiratory rate time series.

We have developed a fuzzy logic-based algorithm to qualify the reliability of heart rate (HR) and respiratory rate (RR) vital-sign time-series data by assigning a confidence level to the data points while they are measured as a continuous data stream. The algorithms membership functions are derived from physiology-based performance limits and mass-assignment-based data-driven characteristics of the signals. The assigned confidence levels are based on the reliability of each HR and RR measurement as well as the relationship between them. The algorithm was tested on HR and RR data collected from subjects undertaking a range of physical activities, and it showed acceptable performance in detecting four types of faults that result in low-confidence data points (receiver operating characteristic areas under the curve ranged from 0.67 (SD 0.04) to 0.83 (SD 0.03), mean and standard deviation (SD) over all faults). The algorithm is sensitive to noise in the raw HR and RR data and will flag many data points as low confidence if the data are noisy; prior processing of the data to reduce noise allows identification of only the most substantial faults. Depending on how HR and RR data are processed, the algorithm can be applied as a tool to evaluate sensor performance or to qualify HR and RR time-series data in terms of their reliability before use in automated decision-assist systems.

Diagnosis of Hemorrhage in a Prehospital Trauma Population Using Linear and Nonlinear Multiparameter Analysis of Vital Signs

In this study, we analyzed a dataset of time-series vital-signs data collected by standard Propaq travel monitor during helicopter transport of 898 civilian trauma casualties from the scene of injury to a receiving trauma center. The goals of the analysis are two fold. First, to determine which combination of the automatically-collected and -qualified vital signs provides the best discrimination between casualties with and without major hemorrhage. Second, to determine whether nonlinear classifiers provide improved discrimination over simpler, linear classifiers. Major hemorrhage is defined by the presence of injuries consistent with hemorrhage in casualties who received one or more units of blood. We randomly selected a subset of the casualties to train and test the classifiers with multiple combinations of the vital-signs variables, and used the area under the receiver operating characteristic curve (ROC AUC) as a decision metric. Based on the results of 100 simulations, we observe that: (i) the best two features obtained are systolic blood pressure and heart rate (mean AUC = 0.75 from a linear classifier), and (ii) the use of nonlinear classifiers does not improve discrimination. These results support earlier findings that the interaction of systolic blood pressure and heart rate is useful for the identification of trauma hemorrhage and that linear classifiers are adequate for many real-world applications.

Lipopolysaccharide-responsive protein kinase C isotypes in the adult rat aorta.

Previous studies have shown that protein kinase C (PKC) activity increases in cardiovascular tissue exposed to lipopolysaccharide (LPS). The objective of these experiments was to identify the PKC isotypes that respond to LPS treatment in the adult rat aorta. We found that PKCα, -δ, -ϵ, and -ζ isotypes are present in endothelium-intact aortas. The PKCα and -ϵ isotypes show two-to threefold increases in abundance after 3 h treatment with 100 ng/mL LPS, while PKCδ and -ζ levels do not increase. In contrast, mRNA for all of the PKC isotypes increased 3.5 to 12-fold during LPS treatment. Both PKC isotype and mRNA levels gradually diminished during 20 h of continuous LPS exposure. Concurrent treatment of the vessels with LPS plus 50 μM cycloheximide caused PKCα, -ϵ, and -ζ, but not -δ, isotypes to rapidly decrease in abundance while blunting the increase in PKC isotype mRNA. The major source for all of the PKC isotypes in the vessel is the vascular smooth muscle cells. These results indicate that LPS treatment induces time-dependent increases in PKC isotype mRNA expression and isotype-specific PKC activation and synthesis in vascular tissue.

LAM: A landscape matching algorithm for respiratory data alignment

Respiratory waveforms and their derived respiratory rate time-series data can become misaligned from each other when they are collected by vital signs monitors under sub-optimal field conditions. The monitor-provided waveforms and rates can be re-aligned by independently calculating respiratory rates from the waveforms and then aligning them with the monitor-provided rates. However, substantially different rates may be generated from the same waveform due to the presence of ambiguous breaths at noisy positions in the waveform. This paper reports a landscape matching (LAM) algorithm to align respiratory rate time-series data with the waveform that they are derived from by using rates that are calculated by different means. The algorithm exploits the intermittent matches between two respiratory rate time series to generate a matching score for an alignment. The best alignment exhibits the highest matching score. The alignment performance of the LAM algorithm is compared to that of a correlation matching (CM) algorithm using field-collected respiratory data. Alignment performance is evaluated by: (1) comparing the ability of the two algorithms to return a shifted waveform to its original, known position; and (2) comparing the percent of points that match between the monitor-provided and calculated respiratory rate time-series data after re-alignment. The LAM alignment algorithm outperforms the CM algorithm in both comparisons at a statistically significant level (p<0.05). Out of 67 samples with shifted time-series data, on average, the LAM aligns respiratory rates within 44 seconds of the original position, which is significantly better the CM-calculated alignment (136 seconds). Out of 465 samples, the LAM performs better, worse, and equal to the CM algorithm in percentage of points matching in 73%, 11%, and 16% of the cases, respectively. This robust alignment algorithm supports the use of reliable post-hoc monitor-provided respiratory rates for data mining purposes.