Ex Parte Schroder et al

Patent Trial and Appeal Board

Mar 27, 2013

11775915 (P.T.A.B. Mar. 27, 2013)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
____________

Ex parte TOBIAS SCHRODER, MARCOS F. VIDAL MELO,
GUIDO MUSCH, ROBERT SCOTT HARRIS, JOSE G. VENEGAS,
and TILO WINKLER
____________

Appeal 2011-009862
Application 11/775,915
Technology Center 3700
____________

Before DEMETRA J. MILLS, MELANIE L. McCOLLUM, and
ULRIKE W. JENKS, Administrative Patent Judges.

Opinion for the Board filed by Administrative Patent Judge JENKS.

Concurring Opinion filed by Administrative Patent Judge McCOLLUM.

JENKS, Administrative Patent Judge

DECISION ON APPEAL
This is an appeal under 35 U.S.C. § 134 involving claims directed to a
method of calculating an input function for determining metabolic activity in
in a subject using positron emission tomographic images. The Examiner has
Appeal 2011-009862
Application 11/775,915

2
rejected the claims for obviousness. We have jurisdiction under 35 U.S.C.
§ 6(b). We reverse.

STATEMENT OF THE CASE
The Specification is directed to determining the metabolic activity of
tissue located in a PET image. “[T]he image frames and the input function
are employed to produce an uptake image. For each image voxel the tracer
uptake curve for that voxel as revealed in the series of acquired image
frames and as determined using the calculated input function.” (Spec. ¶
0033.)
[The] method [is] based on a two-parameter model of the input
function which separates the ROI's [region of interest] time-
activity curve into a blood and an activity spillover
compartment. These two parameters are derived using an
iterative calculation and two blood samples that are used to
calibrate raw PET-derived activity data acquired from a
selected ROI in a PET image.
(Spec. ¶ 0008.)
Claims 1-5 are on appeal, and can be found in the Claims Appendix of
the Appeal Brief (App. Br. A1-2). Claim 1 is the sole independent claim,
and reads as follows (emphasis added):
1. A method for calculating an input function for use in the
assessment of lung activity in a subject, the steps comprising:
a) administering a radionuclide tracer to the subject;
b) acquiring a series of image frames with a positron emission
tomographic (PET) system;
c) acquiring two blood samples from the subject during the
performance of step b);
Appeal 2011-009862
Application 11/775,915

3
d) designating a blood pool region of interest (ROI) in the image
frames; and
e) calculating the input function using information in the two blood
samples, using values in the image frames corresponding to the
designated blood pool ROI, and using a two-parameter model of the
ROI time-activity curve.
The following grounds of rejection are before us for review:
The Examiner has rejected claims 1-3 and 5 under 35 U.S.C. § 103(a)
as unpatentable over Bishop
1
in view of Convert.
2

The Examiner has rejected claim 4 under 35 U.S.C. § 103(a) as
unpatentable over Bishop in view of Convert and further in view of Hoeft.
3

The Issue
The Examiner finds that Bishop disclosed a method of calculating an
input function (Ans. 3). In particular, the Examiner finds that Bishop “does
not specifically suggest a step (c) of actually acquiring the two blood
contributions, or samples, from the subject during the performance of step
(b) PET imaging” (id. at 4), while Convert teaches “obtaining samples of
blood during PET imaging” (id.). The Examiner concludes that “it would
have been prima facie obvious to modify Bishop et al. with
Convert et al., so that one could possible [sic] get accurate diagnosis before
the tracer completely flushes from the blood volume (Bishop et al.: [0053]);
something that is possible with the use of Convert.” (Id.)

1
Harry Bishop et al., US 2002/0068864 A1, published Jun. 6, 2002.
2
Laurence Convert et al., US 2006/0279724 A1, published Dec. 14, 2006.
3
Andreas Hoeft, US 5,865,757, issued Feb. 2, 1999.
Appeal 2011-009862
Application 11/775,915

4
Appellants contend, in relevant part, that neither Bishop nor Convert
“teach or suggest calculating an input function using information in two
blood samples, using values in [PET] image frames corresponding to a
designated blood pool ROI [region of interest], and using a two-parameter
model of an ROI time-activity curve.” (App. Br. 9, 10; Reply Br. 3.)
The dispositive issue is whether the combination of references
calculates an input function “using a two-parameter model of the ROI time-
activity curve” as required by claim 1. As the same issue is dispositive for
both rejections, we will discuss them together.

Findings of Fact
1. Bishop disclosed that “the images [of the heart] are obtained with a
high temporal resolution gamma camera.” (Bishop ¶ 0009; Ans. 4.)
2. Bishop disclosed obtaining a time-activity curve after injection of a
radioactive tracer:
This time-activity curve contains two contributions, one from
tracer flowing through the left ventricle and another from tracer
flowing through the coronary circulation system. The time
activity curve can be represented by the equation:
AT= VLVCLV(t)+ VC(D)CC(t, D) (1)
wherein AT is the measured activity, VLV is the volume of the
left ventricle, CLV is the concentration of tracer in the left
ventricle, which depends on time, t, VC is the volume of the
coronary circulation system which depends on the disease state,
D, and CC is the concentration of tracer in the coronary
circulation system, which depends on both time, t, and the
disease state, D.
(Bishop ¶¶ 0051-0052; Ans. 4.)
3. Convert disclosed:
Appeal 2011-009862
Application 11/775,915

5
The blood counting device 29 comprises a main unit 5, a
pumping unit 7 and a detector assembly 3. The main unit 5
incorporates the electronics to control the pumping unit 7 and
the detector assembly 3, and to communicate with the personal
computer 6 or the PET scanner 27, which are both equipped
with software for remote control, data analysis and display.
(Convert ¶ 0023; Ans. 4.)
4. Convert disclosed that “the device draws blood from a subject, it can
be easily coupled to an automated sampling device to collect micro-volumes
of blood as a time-activity curve is being measured.” (Convert ¶ 0051;
Ans. 4.)
5. The Specification provides that:
[A] first blood sample Cp(t1) is acquired as indicated at process
block 312 from 3.5 to 25 minutes into the data acquisition
phase, and the second sample Cp(t2) is acquired at
approximately 50 minutes into the acquisition phase. Each
blood sample is acquired by drawing 1 ml of venous blood,
which is spun down to obtain the plasma fraction. The activity
concentration is measured in a well counter that has been cross
calibrated with the PET camera.
(Spec. ¶ 0028; see also Fig. 4.)
6. The Specification provides that the “two-parameter model of the input
function which separates the ROI's time-activity curve into a blood and an
activity spillover compartment.” (Spec. ¶ 0008.)
7. The Specification provides:
The PET signal from a blood pool ROI (CPET(t)) is
modeled by two compartments. One accounts for
18
F-FDG
activity from blood plasma (Cp(t)), and the other for spillover
activity from tracer that accumulates in tissue surrounding the
region of interest (ROI). "RC" is defined as a constant recovery
coefficient quantifying a geometrical partial volume effect, and
"SC" as a constant spillover coefficient from tracer
Appeal 2011-009862
Application 11/775,915

6
accumulating in the surrounding tissue. The PET signal CPET(t)
is expressed as:

where the integral describes the tracer accumulation in
surrounding tissue as a function of
18
F-FDG availability in
blood.
(Spec. ¶ 0016.)
8. The Specification provides that the input function is calculated by
using:
[T]he ROI image voxels in all of the reconstructed image
frames as well as the two acquired blood samples. The process
is repeated until the image derived input function corresponds
to the measured blood samples at time t1 and t2. The iterative
process for calculating the input function has been found to
converge to an accurate input function curve in about six
iterations.
(Spec. ¶ 0031.)

Principle of Law
“The Patent Office has the initial duty of supplying the factual basis
for its rejection. It may not . . . resort to speculation, unfounded assumptions
or hindsight reconstruction to supply deficiencies” in the cited references. In
re Warner, 379 F.2d 1011, 1017 (CCPA 1967).

Analysis
Claim 1 requires, in pertinent part, the step of “calculating the input
function using information in the two blood samples, using values in the
Appeal 2011-009862
Application 11/775,915

7
image frames corresponding to the designated blood pool ROI, and using a
two-parameter model of the ROI time-activity curve.” The Specification
provides that the two-parameter model of the ROI time-activity is separated
into a blood compartment and a spillover compartment (FF 6). The blood
compartment looks at the radiotracer activity in the blood plasma and the
spillover activity looks at accumulation of the radiotracer in the tissue
surrounding the ROI (FF 7). The model requires that the constant recovery
coefficient (RC) and the constant spillover coefficient (SC) must be
estimated from the data obtained using the two blood samples (FF 7; Spec.
¶ 0008).
We agree with the Appellants that the Examiner has not set out a
prima facie case of obviousness based on Bishop and Convert. Specifically,
the Examiner has not sufficiently articulated how the ordinary artisan would
arrive at “the two parameter model of the ROI time activity curve” as
presently claimed. The Examiner takes the position that Bishop disclosed all
parameters of claim 1 except “actually acquiring the two blood
contributions, or samples, from the subject during the performance of step
(b) PET imaging.” (Ans. 4.) However, a review of Bishop finds that more
is missing from the reference than just “acquiring two blood samples” from
the subject during PET imaging.
A time-activity curve for the radio tracer measures the concentration
of radioactive tracer in the blood over time; this can be achieved by
measuring the radioactivity in the blood directly (FF 4) or it can be achieved
by measuring the gamma particles in the circulation (FF 2). In Bishop, the
time-activity curve calculation requires two contributions for the calculation
one is the tracer flowing through the left ventricle, VLVCLV(t), which is part
Appeal 2011-009862
Application 11/775,915

8
of the circulation system, and the other is from the tracer flowing through
the remainder of the circulation system, VC(D)CC(t,D) as measured by the
gamma camera (FF 2). Thus, Bishop’s time activity calculation only
measures the tracer in the circulatory system over time and does not look at
the spillover compartment, surrounding ROI, as required by the presently
claimed two-parameter model of the ROI time-activity curve. Convert, on
the other hand, disclosed directly measuring a time-activity curve for the
radiotracer in the blood of an animal. In Convert the device takes samples of
blood and it passes them through a counter (FFs 3, 4). Thus, both Bishop
and Convert arrive at their time-activity curve by directly measuring the
radiotracer signal at various time points after the administration of the
radiotracer. The presently claimed “two-parameter model” separates the
ROI's time-activity curve into a blood compartment and an activity spillover
compartment, that is, the tracer that enters the tissue surrounding the region
of interest (FFs 6, 7). This is not suggested by either Bishop or Convert.
The Examiner takes the position that it was obvious to combine
Bishop with Convert “so that one could possible [sic] get accurate diagnosis
before the tracer completely flushes from the blood volume (Bishop et al.:
[0053]); something that is possible with the use of Convert et al. (Convert et
al.: Abstract).” (Ans. 4.) We agree with the Examiner’s findings, and
Appellants do not dispute, that combining Bishop and Convert would allow
the ordinary artisan to arrive at steps (a)-(d) of the method as presently
claimed. However, what is missing from the Examiner's analysis is
evidence that an artisan would have a reason to arrive at step (e) as presently
claimed. See KSR Int’l v. Teleflex Inc., 550 U.S 398, 418 (2007)
(obviousness rejections require “some articulated reasoning with some
Appeal 2011-009862
Application 11/775,915

9
rational underpinning”). Neither Bishop nor Convert look to account for the
radioactivity found in the spillover compartment, a factor that is required in
order to calculate the two-parameter model of the ROI time-activity curve as
claimed. The Examiner points to no teaching in either reference, or in the
knowledge generally available to an ordinary artisan, suggesting that it
would have been desirable to look at the spillover compartment, and to use
this information in the ROI time-activity curve calculation. “[O]bviousness
requires a suggestion of all limitations in a claim.” CFMT, Inc. v. Yieldup
Intern. Corp., 349 F.3d 1333, 1342 (Fed. Cir. 2003).
As the Examiner has failed to adequately articulate a rationale
explaining how the combination of references discloses “calculating the
input function using information in the two blood samples, using values in
the image frames corresponding to the designated blood pool ROI, and using
a two-parameter model of the ROI time-activity curve,” we are constrained
to reverse the rejections that rely on the combination of Bishop and Convert.

SUMMARY
We reverse the Examiner’s rejection of claims 1-3 and 5 under
35 U.S.C. § 103(a) over Bishop in view of Convert.
We reverse the Examiner’s rejection of claim 4 under 35 U.S.C.
§ 103(a) over Bishop in view of Convert and further in view of Hoeft.

REVERSED

cdc
Appeal 2011-009862
Application 11/775,915

1

McCOLLUM, Administrative Patent Judge, concurring.

As noted by Appellants, Bishop is directed to “a method of screening
patients for an early stage of CAD [coronary artery disease]” (Bishop
¶ [0006]). “The method involves . . . injecting a patient with a radioactive
bolus, . . . measuring the time variation in radioactivity from the bolus in a
left ventricle region of the patient’s body, and . . . analyzing the time
variation to screen the patient for an early stage of CAD” (id. at ¶ [0023]).
The Examiner finds that Bishop teaches calculating the input function
using information in two blood samples (Ans. 4 & 6). However, the
Examiner has not clearly articulated where Bishop teaches the use of
acquired blood samples. In the absence of this teaching, I conclude that the
Examiner has not provided an adequate explanation as to why one of
ordinary skill in the art would have calculated the input function as recited in
step (e) of claim 1. As noted in Convert, non-invasive radioactivity counting
techniques, such as using a “PET imaging system by selecting a blood cavity
as a region of interest” is an alternative to blood sampling (Convert,
¶ [0004]). I conclude that the Examiner has not adequately explained how
the applied references suggest combining these two approaches to calculate
the input function. I therefore agree with my colleagues that the obviousness
rejections should be reversed.