Ex Parte Healey et al

Patent Trial and Appeal Board

Jan 10, 2013

11146745 (P.T.A.B. Jan. 10, 2013)

UNITED STATES PATENT AND TRADEMARK OFFICE
__________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
__________

Ex parte JENNIFER HEALEY and BETH T. LOGAN
__________

Appeal 2011-000368
Application 11/146,745
Technology Center 3700
__________

Before DEMETRA J. MILLS, MELANIE L. McCOLLUM, and
JACQUELINE WRIGHT BONILLA, Administrative Patent Judges.

BONILLA, Administrative Patent Judge.

DECISION ON APPEAL
This is an appeal under 35 U.S.C. § 134 involving claims directed to
computer implemented methods, apparatus, and programs for detecting atrial
fibrillation in an electrocardiogram (ECG) signal. The Examiner has
rejected the claims as obvious. We have jurisdiction under 35 U.S.C. § 6(b).
We reverse.

Appeal 2011-000368
Application 11/146,745

2
STATEMENT OF THE CASE
The Specification describes a computer implemented method for
automatically detecting atrial fibrillation in an ECG signal (Spec. 4). An
ECG typically has “a regular sequence of deflections (waves), labeled P,
QRS, T and U,” as shown in Figure 2 of the Specification (id. at 2). After
computing intervals between successive R peaks in QRS complexes (i.e., R-
R intervals), the method computes a variance of normalized intervals over a
sliding window in the ECG signal and compares that variance with a
threshold to provide an indication of whether atrial fibrillation is present in
the window (id. at 4-5). “An indication of whether atrial fibrillation is
present in a beat window in the ECG signal may be provided dependent on a
number of windows within the beat window in which atrial fibrillation has
been detected” (id. at 5). “In one embodiment, the window is 10 seconds,
the beat window is 600 beats and the settable threshold is 200” (id.)
Claims 1-4, 6-11, and 13-20 are on appeal. Independent claims 1 and
11 are representative and read as follows: (emphasis added):
1. A computer implemented method for automatically detecting atrial
fibrillation in an ECG signal comprising:
detecting QRS complexes in the ECG signal including
computing R-R intervals, wherein detecting QRS complexes is
morphology independent;
based on the computed R-R intervals in the detected QRS
complexes, normalizing the R-R intervals and computing a R-R
interval variance of the normalized intervals over a sliding window in
the ECG signal; and
comparing the computed R-R interval variance with a threshold
to provide an indication of whether atrial fibrillation is present in the
window.
Appeal 2011-000368
Application 11/146,745

3
11. A computer apparatus for automatically detecting atrial
fibrillation in an ECG signal comprising:
a QRS detector stored in a memory which detects QRS
complexes in the ECG signal and computes R-R intervals; and
a normalize routine which normalizes the computed R-R
intervals and computes a R-R interval variance of the normalized
intervals over a sliding window in the ECG signal, the normalize
routine comparing computed R-R interval variance with a threshold to
provide an indication of whether atrial fibrillation is present in the
window;
wherein the sliding window is less than or equal to 10 seconds.
Other independent claims similarly recite “detecting QRS complexes in the
ECG signal including computing R-R intervals, wherein detecting QRS
complexes is morphology independent,” and/or “computing a variance of
normalized intervals over a sliding window in the ECG signal.”
All claims on appeal stand rejected under 35 U.S.C. § 103(a) as
obvious over Lohman,
1
including by reference Adams,
2
in view of Zong,
3

and in further view of Greenhut.
4

Analysis
The Examiner states that “Lohman discloses the claimed invention
except for a morphologically independent QRS detector (claims 1, 4, 6, 7, 8,
11 and 13-20)” (Ans. 5). The Examiner finds that “Zong teaches QRS

1
Lohman et al., US 7,117,031 B2, issued Oct. 3, 2006.
2
Adams et al., US 5,350, 404, issued Sept. 27, 1994.
3
Zong et al., A Robust Open-source Algorithm to Detect Onset and
Duration of QRS Complexes, 30 COMPUTERS IN CARDIOLOGY 737-740
(2003).
4
Greenhut et al., US 5,480,413, issued Jan. 2, 1996.
Appeal 2011-000368
Application 11/146,745

4
complexes analysis using a robust open-source algorithm that is insensitive
to QRS morphology change for the purpose of detecting the onset and
duration of the QRS complexes” (id.). According to the Examiner, it would
have been obvious to an ordinary artisan to use the open-source algorithm
disclosed in Zong in the Lohman system (id.).
We agree with Appellants, and the Examiner concedes, that “Zong
discloses an algorithm that eliminates R peaks” (App. Br. 22; Ans. 8 (stating
that the “examiner also agrees that Zong et al. disclose an algorithm that
eliminated R-peaks”); Suppl. Ans. 3). As explained by Appellants, in Zong
“a QRS complex visible in the ECG signal (ECG), with Q, R, and S waves
labeled, is modified to a QS segment of an LT wave in the length transform
(LT) signal” (id. at 23). Thus, Zong‟s algorithm does not involve computing
or the use of R-R intervals.
The Examiner states that “use of the Zong et al. algorithm to
determine the cycle length and replace the less accurate system based on the
R-R interval is the precise intent of the combination” of Lohman and Zong,
so “that the Lohman et al. system is enhanced with a more precise indication
of the QRS complex and in turn the cardiac cycle length” (Ans. 8).
Appellants persuade us, however, that “in order to achieve
morphological independent when detecting QRS complexes, . . . Zong
expressly teaches the elimination of R peaks and consequently any
measurement of R-R intervals” (1
st
Reply Br. 6). Thus, the cited references
do not “teach or suggest the claimed method that achieves both the
computing of R-R intervals and, at the same time, morphological
independence” (2
nd
Reply Br. 9). In other words, even assuming one would
Appeal 2011-000368
Application 11/146,745

5
have had reason to use the algorithm of Zong for the reasons stated by the
Examiner, a method, apparatus or program code using that algorithm would
not have involved “computing R-R intervals,” as required by independent
claims 1, 4, 6-8, and 14-16 (and therefore dependent claims 2, 3, 9, and 10).
We note that independent claims 11, 13, and 17-20 do not recite
“morphology independent” as recited in other claims. Appellants further
argue, however, that the cited references do not teach or suggest “a
normalize routine which normalizes the computed R-R intervals and
computes a R-R interval variance of the normalized [R-R] intervals,” as
recited in claims 11 and 13 (App. Br. 29, 24-25), or “computing a variance
of normalized intervals,” as recited in claims 17-20 (id. at 30-31, 24-25). In
other words, Appellants argue that the cited reference do not teach or
suggest the use of normalized R-R intervals, as required in claims 11, 13,
and 17-20.
The Examiner finds Lohman discloses “a step to sub-band segments
of RR-interval, and normalize the data which serves as a smoothing routine
for the sub-banded segments (step 214)” (Ans. 4 (citing Lohman, col. 10, ll.
13-19)). The Examiner also states, however, that “modified Lohman [in
view of Zong] discloses the claimed invention except for normalizing the R-
R intervals (claims 1, 4, 6, 7, 8, 11 and 13-20)” (id. at 5). Thus, the
Examiner relies on Greenhut for disclosing normalized R-R intervals (id. at
6; Suppl. Ans. 4).
Greenhut discloses a pacemaker that detects and corrects for
ventricular rate instability that occurs during atrial fibrillation (Greenhut,
Appeal 2011-000368
Application 11/146,745

6
abstract, col. 2, ll. 18-22). As cited in relevant part by the Examiner (Ans.
6), Greenhut states:
Ventricular rate stability can be found using a number of
different methods. For example the interval between two R
waves (standard in the pacemaker field) can be measured and
the stability of this parameter may be defined in terms of
statistical variance, standard deviation, rate mean squared
difference, normalized mean absolute deviation, normalized
approximate interquartile range, autocorrelation, Markov
chains, coefficient of variation, histograms, using maximum,
average, and minimum values stored in a look-up table, or a
normalized mean absolute difference, . . . .
(Id. at col. 5, ll. 11-20 (emphasis added).)
Appellants argue that “Greenhut involves only three normalized
measures, namely, „normalized mean absolute deviation,‟ „normalized
approximate interquartile range,‟ and „normalized mean absolute
difference‟” (App. Br. 25). Thus, according to Appellants, Greenhut “does
not disclose or suggest normalizing the R-R intervals themselves, but instead
involves normalizing a deviation, a range, or a difference” (id.).
In response, the Examiner states: “While it is true Greenhut teaches
normalizing in terms of a deviation, a range and a difference, the data
normalized is the R-R interval data” (Ans. 9; see also Suppl. Ans. 4).
Greenhut teaches, however, that normalized deviation, interquartile range
and/or mean absolute difference may provide information about stability of
the R-R interval (Greenhut, col. 5, ll. 11-20). Such normalized deviation,
range and/or difference are not the same thing as normalized R-R intervals
themselves, as the claims require. As indicated in the Specification, the R-R
interval is normalized using the equation “RRnorm = RR/RRmean * 100” (Spec.
Appeal 2011-000368
Application 11/146,745

7
11). The Examiner does not adequately explain how normalizing deviation,
interquartile range and/or mean absolute difference render obvious
normalizing R-R intervals or the use of normalized R-R intervals.
Thus, Appellants persuade us that “while the deviation, range and
difference normalized are related to the R-R interval, they are not,
themselves, the R-R interval” (1
st
Reply Br. 7). We agree that the Examiner
has failed to establish that Greenhut “teaches normalizing the R-R intervals
themselves as opposed to normalizing other statistics about the R-R interval
as taught by Greenhut” (id. at 8).
SUMMARY
For the reasons discussed above, we reverse the rejection of claims 1-
4, 6-11, and 13-20 as obvious over Lohman, including by reference Adams,
in view of Zong, and in further view of Greenhut.

REVERSED

cdc