Skip to content

Medical Device Integration: Growth, Trends & Challenges: An Interview with John Zaleski, PhD, CPHIMS

01-Jul-15


 

By signing up, you agree to our Terms of Service and Privacy Policy.

HIMSS Blog: Discussing Medical Device Integration Post Publication of John R. Zaleski’s Book “Connected Medical Devices…”

I was interviewed for the HIMSS Blog by the Editors of HIMSS Books and the interview in total is available for viewing at the following link:

Interview with John R. Zaleski: Medical Device Integration: Growth, Trends & Challenges

Why Medical Device Integration Now?

From my recent HIMSS Blog post:

More than half of U.S. hospitals and health systems are planning to purchase and implement a medical device integration (MDI) solution. This is quite a difference from, say, 5 years ago. There are a number of reasons motivating this. Partially, the maturing deployment of electronic health record systems; partially, the maturing of the complexity of integration that requires higher-frequency, higher accuracy, higher fidelity data, such as clinical decision support methods within electronic health record systems; partially, the motivation of Meaningful Use and needs for improvement in patient safety; partially the PP-ACA. Other specific motivations, such as the recognition that improved patient care management can be achieved through better, more accurate data. Furthermore, MDI is an essential element for achieving better patient safety.

Need for Higher Fidelity Data Drove Medical Device Integration Exposure

As a researcher with more than 20 years working with medical devices and as a product developer and inventor, the following trends and major milestones are, in my opinion, the recognition of the value of MDI, which occurred not too long after electronic health record systems became widespread, perhaps not quite 10 years ago; then, the motivating Federal guidelines surrounding Meaningful Use Stages 1 & 2; the PP-ACA also provided some motivation. But, beyond that, my opinion is that receipt of data into the electronic health record systems motivated new ideas about what to do with those data. This, in my opinion, is leading to “higher-level” use cases other than charting. For example, use of the data to assist in improved patient care management and clinical decision making. When I started in this field, I was a graduate student and needed to collect data on live patients whom I was studying to develop methods for weaning from post-operative mechanical ventilation. I was running a study on patients recovering from coronary artery bypass grafting surgery and was following “my” patients from surgery through to extubation from mechanical ventilation in surgical intensive care and general surgery. When I was conducting this study, it was years before the commercial electronic health record system was widely publicized. Furthermore, none of the medical devices with which I was working had any automated data collection capability that was exploited within the hospital system I was working. Hence, I had to write my own code and perform data collection on my own, right at the patient bedside. My purpose was in using the data to develop models of patient state; to better predict time-based changes in physiologic and respiratory parameters, and guided by patient demographics, intakes and outputs, and other information. Yet, what was really lacking was a way to collect this information using an automated, standardized approach. So, I got into the MDI field as it was a necessary utility to meet my ultimate needs: complete data for better clinical decision making.

Tracking an Object in Free-Fall Using the Discrete Kalman Filter

29-Jun-15


 

By signing up, you agree to our Terms of Service and Privacy Policy.

Modeling Ballistic Free-Fall

I have written previously in several posts on applications of Kalman filtering to estimating and tracking physiologic data:

Tracking Medical Device Data: Example of an Excel Macro-based Kalman Filter

Kalman Filtering of Medical Device Data

Signal data artifact smoothing using the Kalman filter on arterial blood pressure measurements

The white paper, which can be downloaded below, describes the use of the Kalman Filter in tracking the vertical position (altitude) of a spherical object. The purpose of this white paper is to illustrate the system dynamics, control, and tracking algorithm application to a system of known true dynamics. Gauss-Markov noise is applied to the measurements to simulate uncertainty in observing the true position of the object. The mathematics and theory can then be applied to other dynamic systems.

Kalman Filtering of Ballistic Free Fall Object

Kalman Filtering System Dynamics

It is necessary first to establish a model of the system dynamics. This is a mathematical representation of the applied external forces that define the motion of the sphere as it descends. Knowing the system dynamics, the accuracy of the prediction of the object’s pathway is increased immensely, particularly if the object is accelerating at a known rate. If the object system dynamics are not known, it is still possible to track the object, with the system dynamics being derived from the observed motion. Yet, the predicted motion of the object and the accuracy with which its position and speed are known from measurement to measurement will not be as accurate and can deviate particularly if external forces cause unpredictable accelerations on the object as it descends. Consider the diagram of Figure 1, in which a sphere of mass, m, and radius, r, is falling under the influence of the two principal forces weight and drag .

Figure 1: Forces on a sphere in ballistic free-fall.

Figure 1: Forces on a sphere in ballistic free-fall.

The attached white paper is developed into a Java program to compute the track of the object as it falls. The purpose of this posting is to provide the capability to those interested in reproducing for academic reasons and for general education. The white paper will be used as the starting point for other posts I intend to write on the tracking of physiological processes, such as heart and lung dynamics.

Evolving from data to knowledge: Using physiologic data to facilitate clinical decisions

22-Jun-15

Better data means better decision making, implying better care

The trend in data collection is that of evolution from episodic, ad-hoc, to more continuous collection of higher frequencies. This higher frequency data collection increases the likelihood that events will be “caught” — hence, better surveillance.

More continuously available information undergirds better care

More continuously available information undergirds better care

“Expert systems are the most common type of clinical decision support system technology in routine clinical use.” [1]

The types of data and uses for expert systems include:

  • Alerts & Reminders
  • Therapy Planning
  • Prescription
  • Data & Information Retrieval
  • Image Recognition & Interpretation

Common Clinical Decision Support System Myths [2]

  • Diagnosis is dominant decision-making issue in medicine: Not “what does patient have,” but, rather, “What should I do?”
  • Clinicians will use expert systems if they can be shown to function at the level of experts: “What do we know?”
  • Clinicians will use standalone tools, so long as they are integrated into clinical workflow.

What is the motivation for better Decision Support?

All across the specialty landscape the prediction for increased demand within the next 20 years is considerable. The motivation for more clinicians is clear. Yet, in addition to increasing the available pool of care providers, improving the capabilities of those in the field through providing better knowledge and decision making capability through access to the right information is key to improved care.

The growing demand and decreasing supply motivates the need for better decision support tools using the available data.

The growing demand and decreasing supply motivates the need for better decision support tools using the available information.

Display is Key to the Right Decisions.

The famous graphic by Charles Minard showing Napolean’s Russian Campaign, is a testament to the genius of his graphic design around 6 data elements: (1)army size, (2,3) position (x,y), (4) direction, (5) temperature, (6) time. Knowing how to display and integrate data into the workflow of the clinician is key to the usability as well as providing a valuable tool for informing on clinical decisions.

Data in 6 variables from Minard (1861) showing Napolean's Russian Campaign. Illustrated are: army size, (x,y) position, direction, temperature, time

Data in 6 variables from Minard (1861) showing Napolean’s Russian Campaign. Illustrated are: army size, (x,y) position, direction, temperature, time

The original presentation by John R. Zaleski is available for download here.

References

[1] Enrico Coiera, Guide to Health Informatics 2nd Edition, Chapter 25, October 2003.

[2] Edward Shortliffe, “Medical Thinking: What Should We Do?” Conference on Medical Thinking, University College of London, June 23rd, 2006.

Modeling Spontaneous Minute Volume in Patients Weaning from Postoperative Mechanical Ventilation

22-Jun-15

A Method for Predicting Spontaneous Minute Volume (SMV) Trajectory in Postoperative Weaning

Weaning is the process by which patients are gradually removed from mechanical ventilation. This post references a legacy article A Method for Predicting SMV on modeling postoperative weaning following the completion of my dissertation. The following figure illustrates the systemic process of surgery followed by transfer of the patient from the operating room (OR) to the surgical intensive care unit (SICU).

Patients re-awaken and resume breathing following coronary bypass grafting procedures in which they experience cessation of breathing as a function of the procedure and require assistance in the form of postoperative mechanical ventilation.

Patients re-awaken and resume breathing following coronary bypass grafting procedures in which they experience cessation of breathing as a function of the procedure and require assistance in the form of postoperative mechanical ventilation.

Study of Mechanically Ventilated CABG Patients

The process of weaning patients from postoperative mechanical ventilation is complex and dependent upon a number of factors that affect the process and viability of a patient to wean. A group of 28 patients who had undergone coronary artery bypass grafting (CABG) in the age range of 28 through 78 years of age were monitored from the time they returned from surgery until extubated from postoperative mechanical ventilation. SMV, the product of respiratory rate and tidal volume, was found to evolve from zero upon patient arrival to between 6 & 10 liters per minute as the effects of the anesthetic abated. A rough model of SMV is shown in the following figure.

Spontaneous Minute Volume qualitative trend versus time.

Typical qualitative trend of the SMV versus time.

The sampled SMV from each patient was binned and all patients who were successfully weaned were found to have measurements in the 6-10 lpm range, as illustrated in the following figure.

Sample patient spontaneous minute volume frequency distribution showing ranges of values just prior to extubation.

Sample patient SMV frequency distribution showing ranges of values just prior to extubation.

and integrated into a distribution function and compared with the distribution function of a normal (Gaussian) distribution. This is illustrated in the following figure.

Sample spontaneous minute volume distribution integrated across all patients and compared with distribution function of a Gaussian.

Sample SMV distribution integrated across all patients and compared with distribution function of a Gaussian.

As can be seen in the illustration, all patients determined to be viable for weaning possessed SMV in the range of 6-10 lpm. In all normal CABG patients, the state variable that was found to evolve over time. In all patients, SMV was immeasurable upon arrival. The method outlined in the article enables critical care staff to predict the SMV trajectory up to 30 minutes into the future, thereby facilitating weaning management and patient care.

Systems Engineering in the Intensive Care Unit: A Model CABG Patient Post-Operative Patient Re-Awakening Time — A Legacy Paper | fentanyl

15-Jun-15

Study of Post-Operative Respiratory Recovery

Paper originally published at INCOSE 1998. Request for reproduction caused me to include the text here. Systems Engineering in the Intensive Care Unit.

In a study of the post-operative recovery of Coronary Artery Bypass Graft (CABG) patients in surgical intensive care units, it was observed that these patients tended to awaken from the effects of anesthesia at varying rates. Because the time at which patients awoke was critical to latter phases of the study, determining the factors which affected the rate of re-awakening were of primary importance. To determine the critical parameters which influenced the re-awakening time of CABG patients, a closed-form model of the patients’ cardiovascular and respiratory systems was constructed to isolate and identify the number of parameters which were thought to affect reawakening time. As a result of this analysis, a mathematical relationship was determined which described the re-awakening time of a patient as a function of several physiological parameters. These included patient body temperature, body mass, body size, and the quantity of anesthetic (fentanyl) administered during surgery.  The work discussed in this paper was performed at the University of Pennsylvania in conjunction with my Ph.D. dissertation requirements.

Key Modeling Parameters

The key input to the process was the amount of anesthesia administered during surgery, and the
key output was the time to re-awaken. Patient physiological parameters affected the reawakening
time by affecting the rate at which the metabolism processes the anesthetic. Therefore, a  mathematical representation of the patient’s rate of metabolism was required to model the
effects of the physiological parameters on reawakening time. The anesthesia, the patient
(system), and the output are represented by the block diagram of Figure 1.

Model of re-awakening CABG patient as a system represented by the parameters mass, height, body temperature and with input the fentanyl dosage and output the re-awakening time.

Figure 1. Model of re-awakening CABG patient as a system represented by the parameters mass, height, body temperature and with input the fentanyl dosage and output the re-awakening time as determined by the time to begin spontaneous breathing.

Knowing when a patient will re-awaken is important to patient care management because
patients must be weaned from post-operative mechanical ventilation. However, before weaning
can occur patients must show that they can sustain spontaneous respiratory load. This cannot happen until the patient’s body temperature has returned to normal and the effects of the anesthesia have worn off sufficiently to enable the respiratory muscles to begin operating autonomously. To facilitate patient care management, it was hypothesized that a model of a patient’s respiratory and cardiovascular systems could be constructed to provide an
estimate of when patients would begin to exhibit spontaneous respiratory response. In brief, it was hypothesized that the re-awakening time, tRA, was a function of the following parameters:

System Equation for Re-Awakening

In deriving the functional form of this model it was necessary to isolate those parameters which were thought to play some role in determining the re-awakening time. These parameters are discussed below.

Patient Age, Mass, Height

Patients varied in ages, mass, and size. Most patients were elderly, as shown in the patient age distribution of Fig. 2. However, the correlation between increased patient age and the need for the surgery is normal as CABG surgery is frequently required in individuals who have experienced many years of arterial plaque buildup.

Patient Age-Frequency Histogram

Figure 2. Patient Age-Frequency Histogram

The functional form of equation (1) indicates that patient mass can be a factor in the reawakening time. To understand the impact of this parameter on re-awakening time it is necessary to understand how changes in patient mass can result in variations in metabolization of the anesthesia. It was hypothesized that a correlation existed between the size of a patient and the patient’s mass. In average people (not abnormally obese or thin), a loose correlation appears to exist, as illustrated in Figure 3. This figure is a scatter plot of patient height versus patient mass, wherein mass varied from approximately 50 kg to 115 kg. Patient height varied between 150 cm and 210 cm. The mass and height of a patient can be related to patient body surface area. Body surface area is used as a normalizing parameter in medicine. Body surface area is incorporated as a normalizing parameter in the relationship between anesthetic dosage and re-awakening time.

Figure 3. Patient height versus mass.

Figure 3. Patient height versus mass.

Patient Body Temperature & Achieving Normothermia

Patients return from surgery cold. The reason for this reduced body temperature is a result of the surgical procedure. Because the body must be at normal temperature so that the heart and the rest of the body functions can operate optimally, reaching normal body temperature normothermia) is key to the re-awakening process. It is only after normothermia is reached that the heart pumps oxygenated blood most efficiently throughout the body. Therefore, patient temperature is a key indicator to achieving efficient cardiovascular performance.
All patients achieved normothermia before the re-awakening process commenced. The average
time to reach normal body temperature, tNORM, was approximately 2.5 hours. Figure 4 is a scatter plot of patient arrival temperature versus time to achieve normothermia. The average temperature profile is shown overlaid on this scatter plot. As can be seen in the figure, the average re-warming time of most patients was found to be in excess of 2 hours.

Figure 4. Patient temperature warming profile.

Figure 4. Patient temperature warming profile.

Deriving the Re-Awakening Model

Patients arrive from surgery dependent on a mechanical ventilator for their breathing. This results from the effect of the anesthesia: it behaves as a muscular inhibitor to ensure that the patient does not move involuntarily during surgery. As the effects of the anesthesia wear off, the patient’s respiratory performance returns with a gradual increase in the amount of air breathed by the patient as a function of time. One measure of this respiratory function is the volume of air breathed by the patient in a given minute. A qualitative relationship between the volume of air breathed in a minute (spontaneous minute volume, Vesp ) and time after surgery is shown in Figure 5.

Figure 5. Evolution of spontaneous minute volume over time.

Figure 5. Evolution of spontaneous minute volume over time.

The re-awakening time was defined by the author as the ability to breathe a steady volume of
air, Vesp, of not less than 1 L/min with 80% confidence. The confidence was determined by monitoring Vesp over a 10 minute interval. If it was found that 80% of all Vesp measurements
exceeded 1 L/min, then the patient was said to be re-awakening. This criterion was selected by the author as an empirical threshold to affirm the reawakening of a patient. Qualitatively, this is illustrated in Figure 6.

Figure 6. Illustration of re-awakening time.

Figure 6. Illustration of re-awakening time.

It was found that once patients began breathing at approximately 1 L/min the volume of air they
continued to breathe spontaneously increased monotonically until they were able to breathe
completely on their own, defined in Figure 5 as breathing between 6 and 10 L/min (depending on body mass). The confidence value of 80% was selected as a threshold for measurements to ensure that the re-awakening of patients was indeed genuine and not spurious or an artifact of a
procedure that may have stimulated the patient to breathe spontaneously. The setting of the threshold on this parameter is of itself a subject of some study. The generalized mathematical relationship between fentanyl dosage and re-awakening time was determined to be most highly correlated to normalized fentanyl dosage, ƒn. This parameter is calculated by computing the ratio of the fentanyl dosage to average patient body surface area. Rollings [Rollings 1984] provides an expression for body surface area, BSA, as a function of patient mass and height:

Equation 2, Body Surface Area

where:

BSA is the patient body surface area measured in square-meters;

m is the mass of the patient, measured in kilograms; and,

h is the patient height, measured in centimeters.

Analysis revealed that a more definitive relationship between anesthetic dosage and time to
reach 1 L/min was achieved by relating this parameter to both the patient body mass and body
surface area, as opposed to body mass alone. Body surface area was selected as an additional
normalizing parameter because the equation defining body mass incorporates an expression
relating to patient height. Patient height affects the size of the body torso, which ultimately affects the size of the lungs and their volume, as well as the cardiovascular system. The relationship between lung volume and body surface area motivates the argument that larger, heavier patients need more anesthetic than smaller, lighter patients. Hence, the patient body mass and surface area (BSA) are used to normalize the dosage. The normalizing parameter, ƒn, is defined as follows:

fentanyl dosing

The regression curve defining the re-awakening time is as follows [1]:

Re-awakening time

A second-order regression was chosen as this represented the lowest order expression with an adequately high correlation coefficient. A plot of the raw data, the mean time to reach 1 L/min., and the curve of equation (4) (solid line) is provided in Figure 7. The upper and lower curves that bound the mean of Figure 7 also define the 1-s standard deviation about the mean value, or 64 minutes. Coincidentally, these bounding curves also define the 80% confidence region for measurements. The plot of Figure 7 provides and indication of when the patient will re-awaken after surgery. Having an estimate of re-awakening time can provide critical care staff with a marker for the time from arrival until the time at which patients will begin breathing on their own. This knowledge provides critical care staff with both a guide for
managing the patients within their care, and as a measuring device with which to determine whether a patient’s progress is normal.

Figure 7. Model to estimate time for patient to begin breathing 1 L/min of air, in comparison with measurements.

Figure 7. Model to estimate time for patient to begin breathing 1 L/min of air, in comparison with measurements.

Discussion

The model of re-awakening time derived here provides a simple and useful tool for estimating patient re-awakening time. This knowledge is vital for the critical care physician as it facilitates the patient care management process, enabling ICU staff to anticipate when and how patients should be evolving during the normal recovery process. It was generally observed that the larger the fentanyl dosage, the longer the time interval until a patient exhibited spontaneous respiratory function. Why? One simple reason is that larger fentanyl doses result in patients who tend to re-awaken more slowly. Patients given heavier doses tended not to re-awaken for perhaps 3-4 hours, and then their respiratory systems responded more slowly. As a result, these patients required ventilatory support for extended periods of time. However, wide variations existed which made the coupling between fentanyl dosage and re-awakening time less significant. What might cause these wide variations? There are a number of reasons. The most obvious is that patients metabolic rates differ, resulting in some patients being able to metabolize
the drug more quickly. Ultimately, by treating the patient as a system of processes, it can be shown that even as complex as the human body is, simple mathematical relationships can be developed to model the outcome of the effects of anesthetic on human respiratory and cardiovascular behavior, and that these simple models can provide a useful tool for physicians to estimate and monitor the expected respiratory responses of patients recuperating from the effects of surgery.

References

[Zaleski 1996] Zaleski, John R. Modeling Post-Operative Respiratory State in Coronary
Artery Bypass Graft Patients: A Methodology for Weaning Patients from Post-Operative Mechanical Ventilation. Ph.D. Dissertation, The University of Pennsylvania, 1996.
[Rollings 1984] Rollings, Robert C. Facts and Formulas. Nashville, TN: By the author, 1984.

Heart Rate Variability (HRV) Analysis Using the Lomb-Scargle Periodogram—Simulated ECG Analysis — Part 2

01-May-15


 

By signing up, you agree to our Terms of Service and Privacy Policy.

Medical Device Integration, Informatics & Heart Rate Variability (HRV) Analysis

The purpose of the attached white paper and the analysis is to study the use of signal processing analysis on heart rate variability data collected via medical device integration to assess the value of this method for advanced informatics & analytics. The selection of the LSP was made because the ability of this algorithm to operate on data that contains gaps is quite important when considering physiologic and other real-world data.

The use of the Lomb-Scargle Periodogram (LSP) for the analysis of biological signal rhythms has been well-documented. [1,2]

“The analysis of time-series of biological data often require special statistical procedures to test for the presence or absence of rhythmic components in noisy data, and to determine the period length of rhythms.” [3]

“In the natural sciences, it is common to have incomplete or unevenly sampled time series for a given variable. Determining cycles in such series is not directly possible with methods such as Fast Fourier Transform (FFT) and may require some degree of interpolation to fill in gaps. An alternative is the Lomb-Scargle method (or least-squares spectral analysis, LSSA), which estimates a frequency spectrum based on a least squares fit of sinusoid.” [4]

[1] T. Ruf, “The Lomb-Scargle Periodogram in Biological Rhythm Research: Analysis of Incomplete and Unequally Spaced Time-Series.” Biological Rhythm Research, 1999, Vol. 30, No. 2, pp. 178-201.

[2] Jozef Púčik, “Heart Rate Variability Spectrum: Physiologic Aliasing and Nonstationarity Considerations.” Trends in Biomedical Engineering. Bratislava, September 16-18, 2009.

[3] T. Ruf, “The Lomb-Scargle Periodogram in Biological Rhythm Research: Analysis of Incomplete and Unequally Spaced Time-Series”. Biological Rhythm Research, 1999, Vol. 30, No. 2, pp. 178-201.

[4] Marc in the box, “Lomb-Scargle periodogram for unevenly sampled time series.” Link: http://www.r-bloggers.com/lomb-scargle-periodogram-for-unevenly-sampled-time-series/. Published January 10th, 2013. Accessed 20-April-2015.

This Paper’s Contribution

This paper focuses on the use of the Lomb-Scargle Periodogram to survey Heart Rate Variabililty (HRV). In a preceding analysis, our focus was on the use of signal processing methods, such as the Lomb-Scargle Periodogram, detect power spectral density versus frequency in time-domain signals, such as heart rate variability. The purpose of that analysis was to illustrate the identification of power spectral density associated with time domain signals using a signal processing method known as the Lomb-Scargle Periodogram (LSP). The LSP is deemed a better method for evaluating power spectral density in time-varying signals where there may be missing or data gaps, or irregular measurements. For this reason, it is deemed superior to the discrete Fourier transform for power spectral analysis related to signals processing involving unevenly sample data, which is frequently the case in biology and medicine.

Download the White Paper Here

A copy of the white paper on HRV analysis using Lomb-Scargle Periodogram can be downloaded here. The following figure depicts the power spectrum produced by the Lomb-Scargle Periodogram when applied to a time-varying signal containing several frequencies. The mathematics of the Lomb-Scarlge Periodogram are such that the method can be applied to time signals with gaps or missing data, thereby improving its utility in real-world settings where such gaps may be expected to occur.

Sample Lomb-Scargle Periodogram associated with a signal

Sample Lomb-Scargle Periodogram associated with a signal

John Zaleski on Medical Device Integration: HIMSS Media Interview on Connected Medical Devices

29-Apr-15


 

By signing up, you agree to our Terms of Service and Privacy Policy.

HIMSS Media Announces New Book on Medical Device Integration: Connected Medical Devices

John Zaleski is interviewed by HIMSS Media (update: #1 best seller at conference 2015) on his new book on the topic of medical device integration. In this book, titled Connected Medical Devices, he describes best practices for medical device integration. This book is intended for the healthcare enterprise that is beginning the process of integrating medical device data into their electronic health record systems. A link to Connected Medical Devices interview is included here:

John R. Zaleski, Ph.D., CPHIMS–HIMSS15 Interview on Connected Medical Devices

Connected Medical Devices: Integrating Patient Care Data in Healthcare Systems

Update: Book of the month for June 2015!

Within a healthcare enterprise, patient vital signs and other automated measurements are communicated from connected medical devices to end-point systems, such as electronic health records, data warehouses and standalone clinical information systems. Connected Medical Devices: Integrating Patient Care Data in Healthcare Systems explores how medical device integration (MDI) supports quality patient care and better clinical outcomes by reducing clinical documentation transcription errors, improving data accuracy and density within clinical records and ensuring the complete capture of medical device information on patients.  The book begins with a comprehensive overview of the types of medical devices in use today and the ways in which those devices interact, before examining factors such as interoperability standards, patient identification, clinical alerts and regulatory and security considerations. Offering lessons learned from his own experiences managing MDI rollouts in both operating room and intensive care unit settings, the author provides practical guidance for healthcare stakeholders charged with leading an MDI rollout. Topics include working with MDI solution providers, assembling an implementation team and transitioning to go-live. Special features in the book include a glossary of acronyms used throughout the book and sample medical device planning and testing tools.

About the Author

John Zaleski, PhD, CPHIMS, brings more than 25 years of experience in researching and ushering to market devices and products to improve healthcare. Dr. Zaleski received his PhD from the University of Pennsylvania, with a dissertation that describes a novel approach for modeling and prediction of post-operative respiratory behavior in post-surgical cardiac patients. He has a particular expertise in designing, developing, and implementing clinical and non-clinical point-of-care applications for hospital enterprises. Dr. Zaleski is the named inventor or co-inventor on seven issued patents related to medical device interoperability. He is the author of numerous peer-reviewed articles on clinical use of medical device data, information technology and medical devices and wrote two seminal books on medical device integration into electronic health records and the use of medical device data for clinical decision making.

Using Radiofrequency Identification (RFID) to promote improved patient identification in telemonitoring

27-Apr-15


 

By signing up, you agree to our Terms of Service and Privacy Policy.

What does RFID and Positive Patient Identification have to do with Telemonitoring, anway?

Chronic ailments, such as congestive heart failure (CHF), diabetes, chronic obstructive pulmonary order, stroke, arthritis and others usually afflict us as we get older. More often than not, we experience more than one of these ailments simultaneously. To facilitate the monitoring and management of these ailments, various types of telemonitoring are being employed to collect vital signs information such as blood pressure, glucose and weight. But, in addition, telemonitoring is being used as a remote method to communicate visually and audibly with nurses and other practitioners, particularly for those patients who are home-bound and infirm. When monitoring vital signs remotely, methods need to be brought to bear to ensure that data are collected from the correct patient, and this is where positive patient identification methods such as barcode or radiofrequency identification (RFID) can come into play.

The original paper, which I wrote a number of years ago, appeared in Practical Patient Care. Subsequent to that time, I opted to create this white paper to highlight key aspects of the subject matter, and to put forth a thought on the use in the home.

RFID for use in medication administration and positive patient identification

The attached paper is an abridgment of an article that originally appeared in Practical Patient Care a number of years ago on the use of RFID technologies in the application of medication administration. The application of barcode and RFID technologies in the areas of medication administration reporting and safety checking suggest there is an expanding role in the use of these technologies at the point of care, regardless of where that point of care might be. In particular, the use of RFID in validating and verifying patient, drug, and order identity as part of the “5 Rights” for medication administration checking and infusion pump drug validation via the use of universal product code (UPC) symbol or RFID-embedded patient bracelets is a key step in nursing workflow within acute care wards. Verifying patient identity is often accompanied by clinician requests to orally verify patient identity via independent checks for patient name, dates of birth, and social security numbers. Patients at home could also employ a similar approach to placing their identification on any vital signs information transmitted from there to a remote telemonitoring facility. This would improve the association and linkage between raw numeric vital signs and related observations to a particular identity.

White paper on telemonitoring and positive patient identification

The attached white paper captures my thoughts on the topic and on other technologies, such as telemonitoring.

Autonomic Heart Controller Device Concept

27-Apr-15


 

By signing up, you agree to our Terms of Service and Privacy Policy.

What is an Autonomic Heart Rate Controller?

The idea for extending the performance of left-ventricular assist devices (LVADs) occurred to me more than 15 years ago. The idea led me to write a white paper at that time which has been maintained and archived on this web site.

Discussions regarding heart rate variability (HRV) caused me to research a conceptual controller for autonomic, chemoreceptor-based sinoatrial heart control. As HRV is affected by sympathetic and parasympathetic control, this fact reminded me of a paper (unpublished) which I had written 14 years ago. While I included a web-adaptation of this paper very early in the history of the web log, I never included the actual paper itself. Much has transpired over the years in the regard to medical device integration, control and research into physiologic monitoring. Yet, I have not seen any writing in particular associated or closely related to the topic at hand.

Heart Rate Autonomic Control White Paper

Originally written in August 2002, the attached white paper presents a concept for a mechanism to automatically controller for heart rate pacing and contractile force (stroke volume) of either an artificial left ventricular assist device (LVAD) or a patient’s own heart who has experienced degenerative performance of the Sinoatrial node.

Autonomic control is based on the hormonal action of concentrations of catecholamines within the blood stream. These, in turn, influence the sinoatrial node through uptake. The concept laid out in the attached paper “operates” by analyzing the chemical nature of the epinephrine, norepinephrine, and dopamine content of the return blood flow through the superior vena cava and then using this information via cyclic voltammetry, neural network control, and feedback to the pacing device to control the heart rate of the assist device.

Why Autonomic Control is Potentially Interesting?

The basic premise of left ventricular assist devices (LVAD) and in heart pacing in general is to provide enough contractile force to move blood throughout the system. Most systems these days rely upon vasodilation, which is related to hormones as well as work. Yet, heart contractility is also related to emotion and release of hormones which are unrelated to direct movement or action of the human body. Hence, the idea was to accommodate a more “realistic” concept that took into account not only the contractile or vasodilation / vasoconstriction aspects of the arteries and veins, but also the hormonal changes measured through catecholamine changes in the blood stream. It is recognized that the attached paper is very raw and, clinically, it is somewhat naive. Yet, the objective was to present the idea as a potential starting point for research and, eventually, a product that could extend the performance of existing LVADs to support more human-like, natural behavior of artificial hearts.

Update on measurement of blood enzymes to support autonomic control:

Recently, a team at Ecole Polytechnique Federale de Lausanne (EPFL) in Switzerland developed an implant to monitor chemicals in the blood. This implant, according to this recent article is reportedly the world’s smallest at 14 mm and measures a maximum of 5 indicators to include troponin, lactase, glucose, ATP, to show whether a heart attack has occurred or to track (in the case of diabetic patients) blood enzymes and protein levels. The information can subsequently be transmitted via Bluetooth to a smartphone for online tracking.

This work is encouraging as it lays the foundation for real-time sensing, data collection and analysis at a level that is necessary for more accurate modeling and monitoring of cardiovascular systems. Such work can lead to earlier detection and more accurate prostheses — particularly, artificial hearts utilizing autonomic control to augment heart rate pacing– as well as more accurate clinical decision support methods that can determine whether patients have experienced critical events. I can see applications to earlier stroke detection where time-is-brain. Such real-time autonomic methods, when integrated with mobile technology and health information technology, can lead to great advances in patient care management through early warning resulting in early intervention.

Medical Device Integration can aid Rapid Response and Early Warning

27-Apr-15

What is rapid response and early warning?

The concept of the Rapid Response Team (RRT) is being introduced in many hospitals throughout the United States and Europe as a means of intervening in patient decline and decompensation prior to the onset of a mortal adverse event. Key measures for the early onset warnings that trigger the deployment of such teams are changes in vital signs that may herald the onset of potentially irreversible cardiac, neurologic, respiratory, deterioration in patients, resulting in the potential for “failure to rescue” in the case of adverse events. Availability of accurate, dense and reliably collected data is key to accomplishing this. Hence, medical device integration, or the ability to extract patient care data measured through medical devices at the point of care, is an enabler and enhancement for RRT early warning score calculation.

Early Warning Scores Calculation

The linked white paper captures some my interests and research into this area together with the calculation of early warning scores from certain vital signs, such as heart rate, blood pressure and temperature. A key aspect of early warning score development in support of rapid response is having the data necessary to process through early warning algorithms in order to create scores or measures of trending decompensation. The early warning score (EWS) and modified early warning score (MEWS) employ measurement of blood pressure and heart rate. These two vital parameters are available through physiologic monitors at the point of care. Data from these physiologic monitors can be extracted either through the network or through serial ports associated with these physiologic monitors.

The two monitors shown below measure various vital signs. The Capnostream 20 monitor measures end tidal carbon dioxide, pulse, pulse oximetry and respiratory rate. The Mindray monitor below that measures many other vital signs, including blood pressure–necessary for the early warning score calculation.

Data from these physiologic monitors may be received through serial ports whereupon the data can be translated into more standardized formats (e.g.: HL7) via hardware and software and then sent to rules engines whereby early warning scores may be calculated. These rules can then be used to notify clinical staff, such as members of a clinical rapid response team (RRT) who can then decide whether this information when taken together with other observations and measurements (such as visual observation of the patient and laboratory data) merit intervention.

Capnostream 20 etCO2 monitor

Capnostream 20 etCO2 monitor

Mindray V12 Physiologic Monitor

Mindray V12 Physiologic Monitor

Predictive Analytics enabled by Medical Device Integration

27-Apr-15


 

By signing up, you agree to our Terms of Service and Privacy Policy.

How can medical device integration enhance prediction?

Data are the heart of decision making. The old adage “what cannot be measured cannot be controlled” is apt here. Historically, clinically quantifiable benefits of connected medical devices within the healthcare enterprise have been measured in terms of time-in-motion studies and workflow relating to time saving associated with accomplish a specific task or end goal. These are valid measures. Yet, the question remains as to whether there is something more tangible clinically that can be used as a measure of effectiveness related to medical device integration. The analysis of data made available from these sources is temporal in nature (i.e., time-varying and collected over time), is multidimensional (i.e., is a vector and represents the patient cardiovascular and respiratory system evolution over time), and is objective if collected automatically from the bedside.

Are Data Access is Key to Improved Prediction and Predictive Analytics?

The attached white paper captures some of the references and measures for improvement relative to medical device integration. Prediction of patient clinical outcome has been the subject of much research and many papers over the years. The difference between large scale data mining and predictive analytics in this context is that data from medical devices are multidimensional time series. Hence, temporal trends in behavior as measured by patient state changes over time provide the ability to track how a patient is “evolving” with time.

The figure below is from a presentation (“FDA Regulatory Submission Prototype Use Case”) I gave at the 2nd Annual Medical Connectivity Conference (San Diego). The figure depicts multiple sources of medical device data, from physiologic monitors to mechanical ventilators to infusion pumps and laboratory systems. The data from each of these medical devices are brought forward to an integrator, whereupon the data can be combined, processed, analyzed and then the output of which can be individual indices, measures, or time predictions. These outputs are, in ensemble, the objective measures of the patient, time-based, and comparative. The scope of the analysis is limited only by the needs and imagination of the researcher and the clinical end user.

What types of algorithms can be fed using data from these sources?

  • Weaning algorithms (i.e., weaning from post-operative mechanical ventilation)
  • Sepsis algorithms (i.e., modified early warning scores combining vital signs, laboratory and visual observations)
  • Respiratory sufficiency assessments and ventilator acquired events (e.g.: ARDS, COPD, PCA management, Extubation Criteria, VAP, etc.)

Many other methods and analysis can be performed, as well, to provide predictive assessments of the patient while in-situ in the critical care, medical surgical, or operating room.

Medical Device Integration -- multiple sources

Medical Device Integration

Legacy INCOSE Healthcare Interest Group: Are Systems Engineering Processes the Key to Treating the Chronically Ill?

27-Apr-15


 

By signing up, you agree to our Terms of Service and Privacy Policy.

Legacy INCOSE Healthcare Interest Group & Systems Engineering

Respiratory and Cardiovascular Systems

Respiratory and Cardiovascular Systems

Almost 14 years ago I participated and ran the Healthcare Interest Group of the International Council on Systems Engineering (INCOSE). As a result of the new book on connected medical devices, several individuals have approached me asking whether there might be another book in the works. In the figure above, I include an image from a presentation I gave in 1999 at a Systems Engineering at the University of Pennsylvania. The cardiovascular and respiratory systems are two excellent examples of human or animal subsystems that can be modeled at macroscopic (systems level) of microscopic levels.

In reviewing past material I had written over the years, I came upon the following article originally published in INCOSE INSIGHT in April 2002. While some of the material is dated, the message is still accurate:

“In pondering the relationship between systems engineering and medicine, it seems that seems that an analogy for medical treatment exists that mimics effective management of software systems development projects. This analogy derives from the concept of proactive development methodologies in comparison with proactive or preventive, medicine. The essence of the analogy is … taking steps to plan for variance…as opposed to treating the cause of the problem…once symptoms appear.”

“System knowledge relies on many different tools and models for predictive purposes, including identification of normal or abnormal behaviors, or indicators for such behaviors…”

Why Systems Engineering Processes in Medicine?

In the case of systems engineering, we adhere to a set of processes to identify the key requirements of any project or product, and identify the opportunities to mitigate risk, improve quality and patient care. Adherence to processes such as identification of risk, improving quality and providing feedback to clinical and support staff to improve patient care through continuous feedback of what works (and what does not work) are extremely important since one can control only what one can measure.

Furthermore, the application of mathematical rigor, statistical and process control techniques and statistics are key in determining whether measurable benefits are realized through the application of specific workflows or techniques. Hence, the application of systems engineering processes can lead to better or increased predictability, particularly through the application of modeling and simulation and by viewing the patient as a complex system of systems, involving inputs and outputs.

A PDF of the original INCOSE Insight article is included as a link here.

 

Medical Device Integration, the Lomb-Scargle Periodogram, and Heart Rate Variability (HRV) — Part 1

21-Apr-15


 

By signing up, you agree to our Terms of Service and Privacy Policy.

Medical Device Integration, HRV and Lomb-Scargle Periodogram — What’s the Connection?

Predicting the future accurately is a capability essential to the basic functioning of our lives. Many fields identify the benefits of forecasting behavior. These include, but are not limited to, weather, financial, sales, and defense. In these cases, the objective is to estimate with as high a degree of confidence as possible the expected outcome such that the estimated result will match the actual result, once the actual result occurs.

The accuracy of the predictions is, of course, dependent upon many factors. Some of these factors include the accuracy of the models used to predict the future events, the amount and fidelity of information these models require to ensure accuracy, and the length of time into the future over which the prediction is estimated to be valid.

Medical device integration provides access to the raw data collected for heart rate variability assessment (that is, raw ECG signals). The variability of these signals is well known in terms of diagnostic inference and, hence, the data provide the source for data analysis and predictive assessment. The HRV analysis of the raw signals in the case of the LSP focuses on determining the periodicity of the heart rate, how it changes over time, and given other observations, can be used in concert to assess weather there is an impending issue.

Signal processing of time-varying signals can produce information and knowledge that are useful in diagnosis and analysis of underlying ailments. Hence, one benefit of medical device integration is providing these time-varying signals at relatively high frequency. One technique for determining the frequency of events in measurements — periodic signal behavior — is the Lomb-Scargle Periodogram.

The LSP is a technique that is in a class of predictive analytic algorithms for detecting signal periodicity and identifying frequency occurrence of events in raw data

Lomb-Scargle Periodogram derived from Data Sourced through Medical Device Integration for HRV Assessment

The use of the Lomb-Scargle Periodogram (LSP) for the analysis of biological signal rhythms has been well-documented in the literature. I include a White Paper as the start of my analysis into Heart Rate Variability (HRV) and its calculation for the purpose of alert notification on change. Heart Rate Variability (HRV) has been used as an assessment of the autonomic nervous system, based on sympathetic and parasympathetic tone (SNS versus PSNS). High HRV is indicative of parasympathetic tone. Low HRV is indicative of sympathetic tone. Low HRV has been associated with coronary heart disease and those who have had heart attacks.

Book on Medical Device Integration: “Connected Medical Devices”… Available Now Through HIMSS Media

08-Apr-15


 

By signing up, you agree to our Terms of Service and Privacy Policy.

About the Author

John Zaleski, PhD, CPHIMS, brings more than 25 years of experience in researching and ushering to market devices and products to improve healthcare. Dr. Zaleski received his PhD from the University of Pennsylvania, with a dissertation that describes a novel approach for modeling and prediction of post-operative respiratory behavior in post-surgical cardiac patients. He has a particular expertise in designing, developing, and implementing clinical and non-clinical point-of-care applications for hospital enterprises. Dr. Zaleski is the named inventor or co-inventor on seven issued patents related to medical device integration and interoperability. He is the author of numerous peer-reviewed articles on clinical use of medical device data, information technology and medical devices and wrote two seminal books on integrating medical device data into electronic health records and the use of medical device data for clinical decision making.

More about the author available through the author’s LinkedIn site. Click on the image to be taken there.

John R. Zaleski, Ph.D., CPHIMS, Author of Connected Medical Devices -- his 3rd book on medical device integration

John R. Zaleski, Ph.D., CPHIMS // LinkedIn Photo Author of Connected Medical Devices — his 3rd book on medical device integration

Author’s 3rd Book on Medical Device Integration — a Practicum

Within a healthcare enterprise, patient vital signs and other automated measurements are communicated from connected medical devices to end-point systems, such as electronic health records, data warehouses and standalone clinical information systems. Connected Medical Devices: Integrating Patient Care Data in Healthcare Systems explores how medical device integration (MDI) supports quality patient care and better clinical outcomes by reducing clinical documentation transcription errors, improving data accuracy and density within clinical records and ensuring the complete capture of medical device information on patients.  The book begins with a comprehensive overview of the types of medical devices in use today and the ways in which those devices interact, before examining factors such as interoperability standards, patient identification, clinical alerts and regulatory and security considerations. Offering lessons learned from his own experiences managing MDI rollouts in both operating room and intensive care unit settings, the author provides practical guidance for healthcare stakeholders charged with leading an MDI rollout. Topics include working with MDI solution providers, assembling an implementation team and transitioning to go-live. Special features in the book include a glossary of acronyms used throughout the book and sample medical device planning and testing tools.

Connected Medical Devices: Integrating Patient Care Data in Healthcare Systems

This text on medical device integration (MDI) focuses on the practical aspects of implementing MDI in the hospital. Book contents are as follows:

Introduction

  • The Mechanics of MDI;
  • Medical Device Driver Software;
  • The MDI Intermediary between the Medical Device and the Health IT system;
  • Major MDI Solution Providers;
  • Vendor Agnostic Representation of MDI Solutions;
  • Some Tips on Selecting an MDI Solution; and,
  • Chapter Summary.

Chapter 1: Medical Device Types and Classes Used in Hospital Departments and How They Communicate

  • Healthcare Enterprise Departments most often in need of MDI;
  • Medical Device Topologies;
  • Surgical Services Environments (Operating Room, OR; Post-Anesthesia Care Unit, PACU);
  • Essential OR Data Elements;
  • Parameter Transmission Intervals – OR;
  • Redundant Parameter Transmission;
  • Intensive Care Unit (ICUs) / Critical Care Units (CCUs);
  • Physiologic Monitors;
  • Mechanical Ventilators;
  • Infusion Systems and Tourniquet Pumps;
  • Specialty Medical Devices;
  • Emergency Departments (EDs);
  • Medical Surgical / Step-Down Units;
  • Chapter Summary.

Chapter 2: MDI Solution Acquisition and Implementation

  • Starting the MDI acquisition process: build or buy
  • Building an MDI solution;
  • Acquiring an MDI solution;
  • The Request for Information (RFI) / Request for Proposal (RFP) Process;
  • Communicating Enterprise Requirements to MDI Solution Providers;
  • Medical Device Driver Development & Timelines;
  • Communicating with the Health IT system;
  • Hospital Facilities and Enterprise Networking Requirements;
  • Building the MDI Implementation Team;
  • Project Management;
  • Staging the MDI Solution Implementation;
  • Assembling the MDI Implementation Team;
  • Estimating Timelines for MDI Implementation Completion;
  • Installation;
  • Testing;
  • Transition to go-live;
  • Chapter Summary.

Chapter 3: Semantic Data Alignment and Time Synchronization of Medical Devices

  • Interoperability Continuum;
  • Semantic Harmonization of Medical Device Data;
  • Temporal Alignment of Medical Device Data;
  • Validating Medical Device Data in the Health IT System Patient Chart;
  • Preparation for Go-Live Checklist; and,
  • Chapter Summary.

Chapter 4: Standards Surrounding Medical Device Integration to Health IT Systems

  • Medical Device Standards Specific to Medical Device Integration;
  • Health Level Seven (HL7) Standards Developing Organization;
  • IEEE 11073 Medical / Personal Health Device;
  • Health Level Seven (HL7) Observation Reporting;
  • Conditioning and Translating Connected Medical Device Data for IT System Consumption;
  • Patient Administration;
  • A Few Words About HL7 Fast Health Interoperable Resources;
  • Integrating the Healthcare Enterprise® (IHE);
  • Other Medical Device Integration-Related Standards; and,
  • Chapter Summary.

Chapter 5: Notification, Alerts & Clinical Uses of Medical Device Data

  • Interface Health and Status Notification and Technical Alerts;
  • Clinical Alerts and Notifications;
  • Aperiodic versus Periodic Data Collection;
  • Clinical Uses of Medical Device Data; and,
  • Chapter Summary.

Chapter 6: Patient Identification and Medical Device Association

  • Methods for Patient Identification;
  • Barcode and RFID;
  • Medical Device Association Workflows;
  • Chapter Summary.

Chapter 7: Regulatory and Security Considerations of MDI

  • Medical Device Data Systems (MDDS);
  • Regulatory Classification and Identification of Risk;
  • Medical Device Security;
  • IEC 80001;
  • Software Development Methodologies and Testing;
  • Chapter Summary.

Appendix A.1: Medical Device Quantity Planning Table

Appendix A.2: Testing Tools

Appendix A.3: HL7 Testing Simulator

Ordering Connected Medical Devices Through HIMSS Media

Click on the image to be taken to the HIMSS Media site to order Connected Medical Devices:

John R. Zaleski's 3rd book on medical device integration. Available through HIMSS Media

Connected Medical Devices: Integrating Patient Care Data in Healthcare Systems

 

Tracking Medical Device Data: Example of an Excel Macro-based Kalman Filter

24-Mar-15
Medical Device Data Tracking: Kalman Filter in Orange; Medical Device Data in Blue.

Medical Device Data Tracking: Kalman Filter in Orange; Medical Device Data in Blue.

Excel Spreadsheet Showing Medical Device Data.

Excel Spreadsheet Showing Medical Device Data.

Detail of spin buttons associated with modifying process and measurement noise. Noise is added to simulate inaccuracies in measurements.

Detail of spin buttons associated with modifying process and measurement noise. Noise is added to simulate inaccuracies in measurements.

Tracking Device Data

Microsoft Excel is an extremely powerful computing tool. I have had a number of people contact me privately asking me to include examples of processing algorithms developed over the years. A key algorithm which I use and of which am fond is the Kalman Filter. In the figures above I provide a sample plot of raw data collected from a glucometer obtained through medical device integration with the tracked values (estimated by the Kalman Filter) overlaid. Below this figure are included a screen snapshot of the columns of data and calculations associated with the filter. Finally, the screen snapshot below this includes the macro code associated with the spinner buttons used to adjust the process and measurement noise of the filter.

MS Excel Kalman Filter Model

The code is saved in a macro-enabled workbook (.xlsm), included below as a download below:

Sample_Glucose Filter

Note that the algorithm included in this analysis is for demonstration purposes only to illustrate the calculation process and to show the flexibility and capabilities of the spreadsheet software. Microsoft Excel is an extremely powerful tool and the Author oftentimes uses it for development prior to writing in a more formalized language (such as Java). The ability to develop Macros in addition to the in-built capabilities of the Excel worksheets make it very capable for fairly complex calculations and algorithms.

A nice tutorial on the Filter can be found here on Bilgin Esme’s Blog.

Modeling Gauss-Markov Noise to Simulate Artifact from Medical Device Measurements

16-Mar-15

Why modeling of measurement artifact is a useful tool

Oftentimes, in analyzing medical device driver performance as part of medical device driver development, it is necessary to recreate physiologic signals that emulate actual measurements. Time-varying signals reflect more accurately on the actual measurements from a patient as they are obtained from medical devices.

Why Gaussian Modeling?

Every so often different types of mathematical models require the modeling of imperfections or noise. The use of the normal distribution is commonly employed to represent the population of events and measurements about some sample mean with some sample standard deviation. In reality, nothing is truly normally-distributed — this is only an approximation. Furthermore, the use of pure Gaussian noise (“white noise”) is rarely the situation when it comes to physical events. More often than not, the use of a form of “colored” noise (i.e., normally-distributed random variates with a Markov “coloration”) is closer to the operational behavior of physical system.

Noise Coding Example

This simple random noise generator, written in an MS Excel macro, that creates random variates using a Gauss-Markov process. In this model, the random number generator can be set to create Gaussian white noise or can be set to incorporate colored noise on top of the Gaussian white noise.

Gauss-Markov Modeling of Medical Device Signal Obtained through Medical Device Integration; beta = 100

Gauss-Markov Modeling of Medical Device Signal Obtained through Medical Device Integration; beta = 100

The figure above shows the Excel Macro running with no Markov noise (i.e., beta set to 100 — effectively infinity). Such is the case in which the current random variate is not dependent on the previous variate. In the following figure, the Markov variable (beta) is set to zero, implying that the current random variate is dependent only on the time interval between the previous measurement and the current measurement. This time interval is set to 1 (unity). Hence, the current random variate is influenced by the previous measurement by the amount exp (-1):

Gauss-Markov Modeling of Medical Device Signal Obtained through Medical Device Integration; beta = 0

Gauss-Markov Modeling of Medical Device Signal Obtained through Medical Device Integration; beta = 0

Mathematically, the new measurement is based on the previous with a noise term given by the exponential, representative of time “memory”:

X[i+1] = X[i] + X[i-1]*exp(-b*dt)

Where:

X[i] is the current Gaussian random variate;

X[i+1] is the new random variate;

X[i-1] is the previous random variate;

b (beta) is the Markov factor; and,

dt is the time interval between the previous and current measurement.

 

JEFFREY HAYZLETT ANNOUNCES JOHN ZALESKI AS AN AUTHOR IN NEW C-SUITE BOOK CLUB

12-Mar-15


 

By signing up, you agree to our Terms of Service and Privacy Policy.

C-SUITE BOOK CLUB

New York, NY, March 12th, 2015: Jeffrey Hayzlett, former Fortune 100 CMO, bestselling author and Bloomberg TV host, announced John R. Zaleski has been inducted into C-Suite Book Club, the premier source for the world’s leading business books for c-suite leaders, business executives, and celebrities.

Showcase C-Suite Bo#127EF2F

Featuring premium content from top thought leaders, designed to increase knowledge, deepen understanding, and build skills to enhance readers’ personal and professional lives, John Zaleski, Ph.D., CPHIMS, the author of CONNECTED MEDICAL DEVICES: INTEGRATING PATIENT CARE DATA IN HEALTHCARE SYSTEMS, has been selected as recommended reading in C-Suite Book Club.

In his debut, John Zaleski gives readers the opportunity to learn about the process of integrating medical devices within enterprise healthcare systems and using their data to support clinical charting and decision support functions. Please click on image or click on the following link to be taken to C-Suite Book Club book reference.

For more information, visit http://www.c-suitebookclub.com.

###

About Jeffrey Hayzlett

Jeffrey Hayzlett is the Contributing Editor and Host of C-Suite with Jeffrey Hayzlett on Bloomberg Television. He is also host of the digital television show Mind Your Own Business on C-Suite TV.  Hayzlett is a global business celebrity, speaker, bestselling author, and Chairman of C-Suite Network, home of the world’s most powerful network of C-Suite leaders.  Connect with Hayzlett on Twitter, FacebookLinkedIn, Google+ or email.

About C-Suite Book Club:

C-Suite Book Club is the premier source of the world’s leading business books for C-Suite leaders and business executives, featuring bestselling authors covering a range of topics, including sales, marketing, leadership, social media, finance, and management. C-Suite Book Club features premium content from top thought leaders, designed to increase knowledge, deepen understanding, and build skills to enhance readers’ personal and professional lives. Visit C-Suite Book Club online and follow them on Facebook and Twitter. C-Suite Book Club is part of C-Suite Network, the world’s most powerful network of C-Suite leaders.

John R. Zaleski, Ph.D., CPHIMS

John R. Zaleski, PhD, CPHIMS, brings more than 25 years of experience in researching and ushering to market devices and products to improve healthcare. Dr. Zaleski received his PhD from the University of Pennsylvania, with a dissertation that describes a novel approach for modeling and prediction of post-operative respiratory behavior in post-surgical cardiac patients. He has a particular expertise in designing, developing, and implementing clinical and non-clinical point-of-care applications for hospital enterprises. Dr. Zaleski is the named inventor or co-inventor on seven issued patents related to medical device interoperability. He is the author of numerous peer-reviewed articles on clinical use of medical device data, information technology and medical devices and wrote two seminal books on integrating medical device data into electronic health records and the use of medical device data for clinical decision making.

Weblog: http://www.medicinfotech.com

Email: john@medicinfotech.com

CONTACT INFORMATION

Media Contact:

John Lee

Account Executive, TallGrass PR

john.lee@tallgrasspr.com

917.653.3444

Kalman Filtering of Medical Device Data

09-Feb-15


 

By signing up, you agree to our Terms of Service and Privacy Policy.

Why Filter Noisy Measurements?

Data obtained from medical devices can be noisy…it can contain artifact or aberrant measurements obtained from the patient. Filtering these measurements to reveal the basic trends in the time-varying signals can help reduce false alarms, if the filtering is chosen carefully. In this post, a simple Kalman filter is presented for the filtering of noisy signals data — such as derived from medical devices or other sensor measurements.

What’s The Motivation? Data!

This post is dedicated in response to those who have contacted me expressing interest in various mathematical models for Kalman filtering of medical device data I have developed over the years. The one-dimensional model presented below makes use of standard methodologies but has been simplified to the point that the results can be expressed as mathematical renderings in the Excel Worksheet columns. Assumptions on noise are that both process and measurement error can be varied using a spinner button. Process noise is kept small (value of 0.01) to ensure proper filter operation. But, both of these noise sources can be varied as desired to show the result in graphical format. The noise that is added is transformed uniformly distributed random variates into Gaussian white noise, which is sampled and overlaid on the “true” measurements of blood glucose to simulate artifact.

The following link provides an example of the Kalman filtering application. Hence, I will not go into the theory here: Signal artifact smoothing using the EKF…

An MS Excel model has been created for tracking blood glucose. An image of the spreadsheet is provided below:

Excel spreadsheet view of data with columns showing the resultant estimates from the one-dimensional Kalman filter.

Excel spreadsheet view of data with columns showing the resultant estimates from the one-dimensional Kalman filter.

Kalman Filter Specifics

Spinner buttons are used to provide for the adjustment of process noise (qk) and measurement noise (vk). The following screen shots show various levels of measurement noise on the tracked overlay on the original signal.

The following plot shows vk = 0; the one following, vk = 0.05.

Any interest in obtaining the actual spreadsheet, please send me an email at john@medicinfotech.com.

Kalman filter of glucose measurements with no measurement noise (process noise = 0.01)

Kalman filter of glucose measurements with no measurement noise (process noise = 0.01)

Kalman filter of glucose measurements with measurement noise = 0.05  (process noise = 0.01)

Kalman filter of glucose measurements with measurement noise = 0.05 (process noise = 0.01)

Haar Wavelet Transform (HWT) Modeling in Microsoft Excel

09-Feb-15


 

By signing up, you agree to our Terms of Service and Privacy Policy.

Can you really create a wavelet model using Excel?

Recently, I received a comment from a reader on an older posting regarding HWTs. This reader requested that I post an example of the MS Excel file that actually creates a  basis function and performs the transformation. One of the benefits of this type of HWT is its relative simplicity: the HWT basis function is very easy to code and process and the creation of the transformation is straightforward.

In the example, I include a set of 8 (x,y) pairs, which are then transformed using the wavelet to their basis coefficients. The plot below illustrates the random data points in X,Y coordinate point form. These data are one-dimensional, representing time along the horizontal, or abscissa coordinate and the value of the signal along the vertical, or ordinate.

Model of Raw (x,y) data points for Haar Wavelet Transformation

Model of Raw (x,y) data points for Haar Wavelet Transformation

Lossy signal reconstruction: all basis function values < 2 removed.

Lossy signal reconstruction: all basis function values < 2 removed.

Calculations are carried out directly in the spreadsheet itself. The transformations are direct and straightforward. The MS Excel spreadsheet, which is shown below, is used to create the basis coefficients and the inverse HWT matrix. A threshold spinner button was added to allow for removing basis coefficients experimentally to see the effect on the recreated signal when certain coefficients are removed from the calculation. This has the effect of “lossy compression”, illustrating quite nicely the effect on recreating an imperfect reproduction of the original signal. Thus, to remove various of the coefficients from the calculation, adjusting the spinner threshold will define the minimum limit below which the coefficients will be excluded. Those coefficients failing to meet the criterion specified by the value indicated by the spinner button will be zeroed out. The recreated signal will then be visible using only those coefficients that meet the specified criterion.

Note: if you would like a copy of the spreadsheet, please contact me through email at john@medicinfotech.com.

HWT Spreadsheet

Haar-Excel

INFORMS 2014 Joint Session on Analytics / Early Warning Notifications

16-Nov-14


 

By signing up, you agree to our Terms of Service and Privacy Policy.

John R. Zaleski Presentation on Alarms, Early Warning Notifications At INFORMS 2014

This past week, I hosted a session on alarms and early warning notifications at the 2014 INFORMS Annual Conference in San Francisco. Also present on the panel were Jim Fackler, MD, then with Johns Hopkins University Medical Center, and Adam Seiver, MD, Ph.D., CMIO at Philips Healthcare Respironics.

The title of the session was Clinical Analytics, Informatics and Clinical Decision Making. The purpose was to propose methods to help mitigate alarm fatigue. During this session I gave a presentation titled Early Warning Notifications Using Medical Device Data. Two slides drawn from the presentation make the point about the seriousness of alarm fatigue in the high acuity hospital settings.

Medical Device Alarms and Alarm Fatigue

Medical Device Alarms and Alarm Fatigue – Average Number of Alarms to which Staff Exposed in Critical Care

Medical Device Alarms and Alarm Fatigue -- Attenuating Alarms

Medical Device Alarms and Alarm Fatigue — Attenuating Alarms

Unique Session on Spurs Recollection of AAMI Alarm Summit

The purpose of this presentation was to introduce types of early warning notifications to hypothetically address alarm fatigue, and to introduce analytics that can be employed using medical device data retrieved from the bedside medical devices for use in clinical decision making. A copy of the presentation is available via the hyperlink above.

I was reminded of presentations given several years ago, at the AAMI Alarms Summit in Virginia. For medical devices that produce alarms, the regulatory pathways are Class II, 510(k) pre-market notification and substantial equivalence to other systems. Algorithm testing is required to quantify false negative (Type II) and false positive (Type I) rates. At this same conference (AAMI Alarms Summit), Maria Cvach of JHMI reported that, per Johns-Hopkins experience with alarms:

  • Crisis delays < 10 seconds
  • Total delays < 15 seconds
  • Crisis counts ~ 585 per day
FDA regulations of monitors producing alarms

FDA regulations of monitors producing alarms

 

Wordpress SEO Plugin by SEOPressor