Call Girls Haridwar Just Call 9907093804 Top Class Call Girl Service Available
Translating into Clinical Practice: NASA 12-lead ECG Inventions for Patient-Centered Medicine & mHealth
1. Translating into Clinical Practice:
NASA 12-lead ECG Inventions
for
Patient-Centered Medicine & mHealth
Todd T. Schlegel, MD
Short Bio:
Born/raised in Minneapolis, Minnesota
Medically trained at Mayo Clinic, Rochester, Minnesota
Joined NASA’s Johnson Space Center in 1992: Medical Officer Senior Scientist
Moved from Houston to Geneva, Switzerland in mid-2013 after 20+ years at NASA
Remain a consultant to NASA, but for management of Ethics Committees, not R&D
*Disclosures:
Present: Nicollier-Schlegel SARL, Switzerland
*New start-up company focused on commercializing the NASA-owned & other IP
2. Summary of presentation
1. New dry electrodes, harnesses and “gloves” for 12lead ECG data acquisition (hardware);
2. New digital-to-analog converter system for enabling
multiple automated interpretations of 12-lead ECGs from
several manufacturers at one time (hardware + software);
3. New advanced ECG (“A-ECG”) software for
improved ECG-based diagnosis of heart diseases (software)
All can be used either locally, or remotely for telemedicine
3. Recent “12-lead ECG Harness” co-developed at NASA
Internal view
Figure:
A. Orbital’s Research Inc.’s FDA-cleared dry electrode as an individual sensor (Orbital also owns harness IP)
B. NanoSonics’ dry electrode as an individual sensor (not yet FDA cleared)
C. Orbital’s electrodes in our 12-lead ECG harness, with wireless ECG transmission (dotted arrow) via Bluetooth
D. Receiving Android smart phone
E. Printout of 12-lead ECG obtained from harness as received by phone.
Abbreviations: RA, RL, LA, LL: right arm, right leg, left arm and left leg electrodes, respectively
4. Why would NASA want a dryelectrode harness for 12-lead ECG?
1) Reduce lead wire “spaghetti” (like that
shown at lower left)
2) Eliminate ECG electrode disposables
3) Eliminate adhesives (irritating and
uncomfortable)
4) Reduce don/doff time and overall
time for 12-lead ECGs
5) Prevent the common clinical problem
of reversed lead wire placements
6) Make 12-lead ECGs more “mobile”
7) Allow non-physician crewmembers to
easily “self-collect” 12-lead ECGs
5. A disposable alternative for 12-lead ECG self-administration:
The ECG Glove from NYC-based “INeedMD, Inc” (already CE & FDA cleared)
- Uses standard adhesives (not dry electrodes) like any other single-use disposable. Mason-Likar configuration.
- Expense a potential issue, but advantages are similar to the harness, plus also higher portability and sterility
- If properly sized, it works very well for resting 12-lead ECG
6. What is meant by “properly sized”?
V5
V6
Clearly one size does not “fit all”…
(And the same thing applies to all harnesses, shirts, etc)
7. Alternative commercialized dry-electrode embodiments for
resting 12-lead ECG (all from Israel, not NASA).
SHL’s “CardioSenC” harness for 12-lead ECG; has an
embedded earlier generation cell phone for slow
transmission of snapshot ECGs only. One squeezes
the RA and LA electrodes into one’ s armpits.
Tapuz Medical’s rubberized “Apron” for resting 12-lead
ECG; patient “stretches” arm-loop electrodes backward
SHL’s newer “SmartHeart”
version (left); has CE and
FDA clearance now, but for
SHL’s “telemedicine-center”
clients only (evidently).
Commwell Medical’s “Glove” for resting
12-lead ECG (Israeli designed, Illinois based)
10. What might a truly “patient-centered” automated
interpretive system for 12-lead ECG look like?
Probably
something
like this:
ECG data formatted
by Manufacturer X
ECG data formatted
by Manufacturer Y
ECG data formatted
by Manufacturer Z
Software (e.g., in EHR system) that
recognizes all of these different formats
and converts them to a common format
from which the interpretive
functionalities of all three manufacturers
can then be accessed
The unsure/querying clinician and
the patient benefit by choosing
the best/most likely automated
12-lead interpretations from
several manufacturers. Similar to
being able to access the “Poll the
Audience” lifeline on “Who Wants
to be a Millionaire?”
Manufacturer
X’s automated
interpretation
+
Manufacturer
Y’s automated
interpretation
+
Manufacturer
Z’s automated
interpretation
The Problem?
Multiple clinician researchers with expertise in
ECG have been trying to get Manufacturers
X, Y and Z to cooperate in this fashion for
decades, but no luck.
(A clash between patient-centered medicine
and manufacturer economic interest?)
So how might we assist clinicians and
patients, at least for now?
11. START HERE
Original 12-lead ECG
collected on machine of
any “Manufacturer X”
Here’s our effort to help:
New 12-lead ECG “DAC” Device:
END HERE
Same 12-lead ECG tracing but
interpreted by machine(s) of
See: Kothadia et al, New system for digital to
Manufacturer Y, Z, etc. instead)
analog transformation and reconstruction of 12- ”second opinion” automated dxs
lead ECGs. PLoS One. 2013 Apr 11;8(4):e61076.
A hardware/software device that nonengineers might regard as a “black box”
of sorts (prototype below)
Raw digital 12-lead
ECG data of known
format from manufacturer
X’s XML or similar file (or
live data stream)
Software + Board that
receives the digital 12-lead
ECG data (wirelessly if
desired), then reformats
those data into an optimized
binary format, then streams
them out at an optimal
sampling rate to an optimally
configured digital-to-analog
converter (DAC).
To any other ECG
manufacturer’s
machine(s), for “redigitization” and “reanalysis” there.
.
1
2
Level shifter chips that just allowed our
prototype Board (left) to communicate with
our prototype DAC (right), given that they
used different logic voltage levels.
3
Optimally configured DAC
that converts the digital data
streamed from the Board into
an optimized set of analog
signals that can be “re-fed” into
any manufacturer’s ECG
machine(s) for re-digitization
and re-analysis there.
12. Cloud or Server-based use of the DAC for patient-centered mHealth (e.g. with “Glove” or Harness)
Other advantages of new DAC:
- Fully patient centered (patient can initiate too)
- Savings on ECG capital equipment purchases
- Smaller, more mobile (if space constrainted)
- Improve all manufacturers’ diagnostic algorithms
Disadvantage:
- A pure software solution coming from cooperating
major manufacturers would be less cumbersome
(more practical), albeit not “universal” like the DAC
Inexpensive Android
Tablet or Smart Phone
digital ECG
data (in
optimized
format)
Remote bot
Reports
Multiple Reports
(conv. ECG + A-ECG)
Remote
physician
(if needed)
Secure Server/Cloud
New 12-lead
ECG “DAC”
device
Or…
To any
manufacturers’
ECG device(s)!
13. Advanced ECG Diagnostic Software (aka “A-ECG”)
A-ECG is suite of software applications that analyze, in one place, the results from
nearly every advanced & conventional 12-lead ECG parameter known to have
diagnostic or prognostic utility.
R&D philosophy: “Collect once, analyze everything of potential value”
A-ECG utilizes:
• A large and ever-growing database of 12-lead ECG information collected from
populations of both healthy and diseased individuals as defined by
accompanying, gold-standard cardiac imaging results, and
•
Both basic (e.g., logistic regression) and advanced (e.g., “pattern recognition”)
statistics, accompanied by branch-and-bound, jackknifing, bootstrapping, etc
procedures, to determine which of the 400+ advanced and conventional ECG
parameters studied are truly the most essential, especially in combination, for
detecting any given cardiac disease process;
To produce:
• Automated “A-ECG scores” (=“fingerprints” for each disease) that: 1) are
diagnostically more powerful than information gleaned from strictly
conventional ECGs; and that 2) can be applied forward to “new” patients to
allow for improved diagnosis plus continuous iteration and improvement
14. (Or Wireless mHealth 12-Lead ECG)
12-lead (8-channel) ECG data file or data stream
Large underlying database of patients with A-ECG
plus known “gold standard” clinical imaging results
15. Example Technique #1: Beat-to-beat QT interval variability (QTV)
A measurement of “temporal repolarization
heterogeneity”
e.g.,
In multiple studies performed by multiple different
investigator teams, increased QTV compared to
underlying RRV (i.e., increased “temporal
repolarization heterogeneity” or “lability” at rest)
strongly predicts the presence of cardiac disease and
the propensity for life-threatening cardiac events.
We employ several unique QTV parameters in A-ECG
From Starc and Schlegel, J. Electrocardiol
39: 358-367, 2006.
16. Example Technique #2: T-wave and QRS-wave complexity via “PCA” or “SVD”
“Complexity” by
Principal
Component
Analysis
(“PCA”) and/or
“Singular Value
Decomposition”
(“SVD”)
PC1, PC2 and PC3 are the first, second and third “principal components” (aka
“eigenvectors”) respectively, derived from “PCA” of the 8 independent ECG channels
that constitute the standard 12-lead ECG. (Thus one actually obtains PC1-PC8).
The vast majority of the “energy” will be in PC1
With “disease” compared to “health”, relatively more energy “shifts downstream”
PC1PC2…PC8
e.g., “PCA ratio” = PC2/PC1 x 100% (normal value is <25-30%, gender specific)
We employ several unique complexity-related parameters in A-ECG
17. Example Technique #3: “3D ECG”: e.g., Spatial QRS-T angle by 12-to-Frank lead transform
Normal: Spatial QRS-T angle = 61°
Abnormal: Spatial QRS-T angle = 111°
T loop
T loop
QRS loop
QRS loop
Figures from Cortez and
Schlegel, J. Electrocardiol.
2010 or Sakowski et al,
Aviat. Space Environ. Med.
2011.
Borderline: Spatial QRS-T angle = 88°
“Do you know what
your spatial QRS-T
angle is?”
is clinically a much
more important
question than:
“Do you know what
your cholesterol level
is?”
T loop
QRS loop
Increased spatial QRS-T
angle has more prognostic
value for fatal endpoints
than any classic CV risk
factor, including diabetes
(Kardys et al, Spatial QRS-T
angle predicts cardiac death
in a general population Eur.
Heart J. 2003;24:1357-1364).
It’s also the single best
known predictor of lifesaving appropriate ICD
discharges in patients with
ICDs (Borleffs et al, Circ. EP
2009; 2:548-554, 2009).
18. Rather than potentially bore an “informatics” audience with numerous charts and tables
outlining the details of our statistical procedures & results (for which our published
literature – appended to the presentation – can be referenced), here are some:
Case studies that demonstrate the
“informatics utility” of A-ECG compared
to strictly conventional ECG
All four case studies will have normal or “non-specifically abnormal”
conventional 12-lead ECGs
- Which of the four cases have heart disease, and which do not?
- And of those with heart disease, is LVSD (LVEF <50%) also present?
22. CASE 4: Disease or no Disease? (57 year-old male)
Poor R-wave progression? (Clinically non-specific)
Some “flattening” (maybe?) in the lateral T-waves?
23. A-ECG results from CASE 1 (the 34 year-old female):
Lead II
(SVD)
RR (ms)
ΔQT (ms)
PCA
normal
normal
All normal
References for QTV technique:
- Starc V. and Schlegel TT. J Electrocardiol 2006;39:358-367.
- Berger, R. et al. Circulation. 1997;96:1557-1565.
A-ECG results from CASE 2 (the 69 year-old male):
Lead II
(SVD)
RR (ms)
ΔQT (ms)
PCA
also normal
borderline
(but still
“normal”)
All normal
24. A-ECG results from CASE 3 (the 62 year-old male) (Pd=132 ms)
Lead II
PCA
(SVD)
ΔQT (ms)
RR (ms)
borderline
abnormal
All normal
A-ECG results from CASE 4 (57 year-old male, poor R-wave progression)
Lead II
(SVD)
RR (ms)
ΔQT (ms)
PCA
All abnormal
25. First the results from some simple but validated A-ECG
“logistic scores”:
(These are scores derived from results in the underlying database and take the general form of “aX + bY – cZ … +/- some
constant”. They provide simple, binary “yes vs. no” diagnostic calls for new patients. For statistical & other details, see
Schlegel, T.T., et al. BMC Cardiovascular Disorders 2010, 10:28 here: http://www.biomedcentral.com/1471-2261/10/28 )
Case#
Score 1:
Probability
of “Disease”
1. (34 F)
2. (69 M)
3. (62 M)
4. (57 M)
0.00068615
0.00121214
0.97871634
0.9996925
Score 1 call
Score 2:
Probability
of LVSD
“LVSD vs. no LVSD” Call
Healthy
Healthy
Disease
Disease
----0.15690945
0.78327565
No LVSD
No LVSD
No LVSD
LVSD
Thus according to our A-ECG logistic score results above:
Cases #1 and #2 are not diseased;
Case #3 is diseased, but does not have LVSD; and
Case #4 is most (severely) diseased, and also has LVSD
Next, let’s see what a follow-on Discriminant Analysis (DA) shows:
26. DA for common outpatient cardiac diseases
Case:
1.
2.
3.
4.
DA Call:
Healthy
Healthy
CAD
NICM
Probabilities:
0.99959715
0.87930201 CAD 0.12
0.98210161
0.99999984
CAD = coronary artery disease
LVH = left ventricular hypertrophy
HCM = hypertrophic cardiomyopathy
NICM = non-ischemic cardiomyopathy
ICM = ischemic cardiomyopathy
-- Thus the discriminant analysis results are very consistent with the A-ECG logistic score results.
-- And most importantly, both the A-ECG logistic score and discriminant results were also very consistent with the
gold-standard imaging results, which showed:
Cases 1 and 2: Normal imaging with no demonstrable disease;
Case 3: 2-vessel CAD, but with normal LVEF = 60%;
Case 4: Severe non-ischemic cardiomyopathy with LVEF = 25%
27. Same DA as the immediately prior slide, but now in “3D”:
Case 1Healthy; Case 2Slightly less Healthy, likely
“Healthy aging”; Case 3CAD; Case 4NICM
Canonical Plot 3D
Same style of DA as above, but when
one allows visualization of some of
the data points that constitute the
underlying background “spheres”
(“circles” in 2D), which themselves
represent areas of ~95% certainty for
the presence of the given condition.
28. DA for A-ECG “age” score
32.3 years
62.3 years
34
69
-2 (A-ECG age younger than chron. age)
-7 (A-ECG age younger than chron. age)
DA’s calls re: “ECG age”:
Case #
1. (34 F)
2. (69 M)
Predicted
decade
30s
60s
Probabilty
0.75
0.50
Remainder predicted
decade(s) & probability
20s 0.19
50s 0.14 70s 0.35
29. Bayesian A-ECG age scores
A statistically more robust way of
estimating “A-ECG age”
(many pages worth of this kind of
math in the background)
So one needs bright
statisticians to do this….
So for our two “healthy” cases:
Case
Case 1
Case 2
Bayesian predicted
A-ECG age
32.3 years
62.3 years
True chronological age
34 years
69 years
Delta
-2 years
-7 years
Conclusion
A-ECG age younger than chronological age
A-ECG age younger than chronological age
30. A-ECG Telemedicine in the 3rd World
Fiji: Oct 20-23rd 2013
(Snapshot A-ECG screening via
WiFi ECG + Android tablet/phone)
Conventional and A-ECG
(Data
Collection)
(Secure
Server/
Clinical
HQ in NZ)
A-ECG plus additional conventional ECG results
Raw conventional ECG data
(Analysis)
Conventional/advanced echo (validation)
31. Thanks to NASA, Dr. Walter Kulecz (programming), and all students, visiting faculty,
clinical collaborators and patients who continue to share the vision of A-ECG.
Special thanks to Dr. Patrick Gladding for NZ-related collaboration and to HINZ.
toddschlegel@rocketmail.com
32. References for beat-to-beat QT interval variability:
Atiga WL, Calkins H, Lawrence JH, Tomaselli GF, Smith JM, Berger RD. Beat-to-beat repolarization
lability identifies patients at risk for sudden cardiac death. J Cardiovasc Electrophysiol. 1998;9:899-908.
Berger RD, Kasper EK, Baughman KL, Marban E, Calkins H, Tomaselli GF. Beat-to-beat QT interval
variability: novel evidence for repolarization lability in ischemic and nonischemic dilated
cardiomyopathy. Circulation. 1997;96:1557-1565.
Murabayashi T, Fetics B, Kass D, Nevo E, Gramatikov B, Berger RD. Beat-to-beat QT interval variability
associated with acute myocardial ischemia. J Electrocardiol. 2002;35:19-25.
Haigney MC, Zareba W, Gentlesk PJ, et al. QT interval variability and spontaneous ventricular
tachycardia or fibrillation in the Multicenter Automatic Defibrillator Implantation Trial (MADIT) II patients.
J Am Coll Cardiol. 2004;44:1481-1487.
Starc V, Schlegel TT. Real-time multichannel system for beat-to-beat QT interval variability. J
Electrocardiol. 2006;39:358-367.
Piccirillo G, Magri D, Matera S, et al. QT variability strongly predicts sudden cardiac death in
asymptomatic subjects with mild or moderate left ventricular systolic dysfunction: a prospective study.
Eur Heart J. 2007;28:1344-1350.
Vrtovec B, Okrajsek R, Golicnik A, et al. Atorvastatin therapy may reduce the incidence of sudden
cardiac death in patients with advanced chronic heart failure. J Card Fail. 2008;14:140-144.
Baumert et al. Conventional QT Variability Measurement vs. Template Matching Techniques:
Comparison of Performance Using Simulated and Real ECG. PLoS One. 2012; 7:e41920. Freely available
online here: http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0041920
33. References for T-wave Complexity:
Priori SG, Mortara DW, Napolitano C, et al. Evaluation of the spatial aspects of T-wave complexity
in the long-QT syndrome. Circulation. 1997;96:3006-3012.
Malik M, Acar B, Gang Y, Yap YG, Hnatkova K, Camm AJ. QT dispersion does not represent
electrocardiographic interlead heterogeneity of ventricular repolarization. J Cardiovasc
Electrophysiol. 2000;11:835-843.
Zabel M, Acar B, Klingenheben T, Franz MR, Hohnloser SH, Malik M. Analysis of 12-lead T-wave
morphology for risk stratification after myocardial infarction. Circulation. 2000;102:1252-1257.
Zabel M, Malik M, Hnatkova K, et al. Analysis of T-wave morphology from the 12-lead
electrocardiogram for prediction of long-term prognosis in male US veterans. Circulation.
2002;105:1066-1070.
Okin PM, Devereux RB, Fabsitz RR, Lee ET, Galloway JM, Howard BV. Principal component
analysis of the T wave and prediction of cardiovascular mortality in American Indians: the Strong
Heart Study. Circulation. 2002;105:714-719.
Okin PM, Malik M, Hnatkova K, et al. Repolarization abnormality for prediction of all-cause and
cardiovascular mortality in American Indians: the Strong Heart Study. J Cardiovasc
Electrophysiol. 2005;16:1-7.
Rautaharju PM, Kooperberg C, Larson JC, LaCroix A. Electrocardiographic abnormalities that
predict coronary heart disease events and mortality in postmenopausal women: the Women's
Health Initiative. Circulation. 2006;113:473-480.
Batdorf BH, Feiveson AH, Schlegel TT. The effect of signal averaging on the reproducibility and
reliability of measures of T-wave morphology. J Electrocardiol. 2006;39:266-270.
34. References for 3D ECG (especially spatial mean QRS-T Angle)
Kardys I, Kors JA, van der Meer IM, Hofman A, van der Kuip DA, Witteman JC. Spatial QRS-T
angle predicts cardiac death in a general population. Eur Heart J. 2003;24:1357-1364.
de Torbal A, Kors JA, van Herpen G, et al. The electrical T-axis and the spatial QRS-T angle are
independent predictors of long-term mortality in patients admitted with acute ischemic chest
pain. Cardiology. 2004;101:199-207.
Yamazaki T, Froelicher VF, Myers J, Chun S, Wang P. Spatial QRS-T angle predicts cardiac death
in a clinical population. Heart Rhythm. 2005;2:73-78.
Fayn J, Rubel P, Pahlm O, Wagner GS. Improvement of the detection of myocardial ischemia
thanks to information technologies. Int J Cardiol. 2006.
Rautaharju PM, Kooperberg C, Larson JC, LaCroix A. Electrocardiographic abnormalities that
predict coronary heart disease events and mortality in postmenopausal women: the Women's
Health Initiative. Circulation. 2006;113:473-480.
Restier-Miron L, Fayn J, Millat G, et al. Spatiotemporal electrocardiographic characterization of
ventricular depolarization and repolarization abnormalities in long QT syndrome. J
Electrocardiol. 2007.
Borleffs et al. Predicting Ventricular Arrhythmias in Patients With Ischemic Heart Disease:
Clinical Application of the ECG-Derived QRS-T Angle. Circ Arrhythmia Electrophysiol.
2009;2:548-554.
Cortez, D.L., and Schlegel, T.T. When deriving the spatial QRS-T angle from the conventional 12lead ECG, which transform is more Frank: regression or inverse Dower? J. Electrocardiol. 2010;
43: 302–309.
35. Recent references for A-ECG that are freely available online:
Schlegel, T.T. , et al. Accuracy of advanced versus strictly conventional 12-lead ECG for detection and screening of
coronary artery disease, left ventricular hypertrophy and left ventricular systolic dysfunction. BMC Cardiovascular
Disorders 2010, 10:28; doi:10.1186/1471-2261-10-28.
Freely available online here:
http://www.biomedcentral.com/1471-2261/10/28
Or here:
http://www.medscape.com/viewarticle/725244
Starc, V. et al. Can functional cardiac age be predicted from the ECG in a normal healthy population? Computing in
Cardiology 2012; 39:101-104.
Freely available online here:
http://cinc.org/archives/2012/pdf/0101.pdf
Gladding et al. Open access integrated therapeutic and diagnostic platforms for personalized cardiovascular
medicine. J. Pers. Med. 2013, 3(3), 203-237; doi:10.3390/jpm3030203
Freely available online here:
http://www.mdpi.com/2075-4426/3/3/203
(A-ECG sections are on pages 209-211, 220-221 and 223-226)