Ex Parte Huo et al

Patent Trial and Appeal Board

Mar 8, 2013

11414759 (P.T.A.B. Mar. 8, 2013)

UNITED STATES PATENT AND TRADEMARK OFFICE
UNITED STATES DEPARTMENT OF COMMERCE
United States Patent and Trademark Office
Address: COMMISSIONER FOR PATENTS
P.O. Box 1450
Alexandria, Virginia 22313-1450
www.uspto.gov
APPLICATION NO. FILING DATE FIRST NAMED INVENTOR ATTORNEY DOCKET NO. CONFIRMATION NO.
11/414,759 04/28/2006 Zhimin Huo 90539 2856
70523 7590 03/08/2013
Carestream Health, Inc.
ATTN: Patent Legal Staff
150 Verona Street
Rochester, NY 14608
EXAMINER
BRUTUS, JOEL F
ART UNIT PAPER NUMBER
3777
MAIL DATE DELIVERY MODE
03/08/2013 PAPER
Please find below and/or attached an Office communication concerning this application or proceeding.
The time period for reply, if any, is set in the attached communication.
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
__________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
__________

Ex parte ZHIMIN HUO and XU LIU
__________

Appeal 2011-011041
Application 11/414,759
Technology Center 3700
__________

Before MELANIE L. McCOLLUM, JEFFREY N. FREDMAN, and
JACQUELINE WRIGHT BONILLA, Administrative Patent Judges.

McCOLLUM, Administrative Patent Judge.

DECISION ON APPEAL
This is an appeal under 35 U.S.C. § 134 involving claims to a method
for determining a candidate lesion region in a digital ultrasound medical
image. The Examiner has rejected the claims as obvious. We have
jurisdiction under 35 U.S.C. § 6(b). We reverse.
STATEMENT OF THE CASE
Claims 1, 2, and 4-13 are pending and on appeal (App. Br. 1). We
will focus on claim 1, which reads as follows:
Appeal 2011-011041
Application 11/414,759

2
1. A method for determining a candidate lesion region in a digital
ultrasound medical image of anatomical tissue, the method comprising using
a processor or computer to perform automatically steps of:
accessing the digital ultrasound medical image of anatomical tissue;
applying an anisotropic diffusion filter to the ultrasound image to
generate a filtered ultrasound image;
performing a normalized cut operation on the filtered ultrasound
image to partition the filtered ultrasound image into a plurality of regions;
merging the plurality of regions into merged regions based on
predetermined threshold values; and
selecting, from the merged regions, at least one region as a candidate
lesion region.
Claims 1, 2, and 4-13 stand rejected under 35 U.S.C. § 103(a) as
obvious over Kupinski
1
in view of Madabhushi
2
and Shi
3
(Ans. 3-4).
The Examiner relies on Kupinski for disclosing “methods and an
associated system for detecting a lesion in an ultrasound image of
anatomical tissue” (id. at 4). The Examiner finds that Kupinski teaches “a
computer readable medium to perform the steps of the method
automatically” (id.).
With regards to a normalized cut, the Examiner notes that Appellants’
Specification “disclose[s] normalized cut to partition images into plurality of
regions . . . and segmenting spatially contiguous pixels” (id.). The Examiner
finds that Kupinski teaches “partitioning of spatially contiguous pixels of

1
Kupinski et al., US 6,138,045, Oct. 24, 2000.
2
Anant Madabhushi and Dimitris N. Metaxas, Combining Low-, High-Level
and Empirical Domain Knowledge for Automated Segmentation of
Ultrasonic Breast Lesions, 22 IEEE TRANSACTIONS ON MEDICAL IMAGING
155-169 (2003).
3
Jianbo Shi and Jitendra Malik, Normalized Cuts and Image Segmentation,
22 IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE
INTELLIGENCE 888-905 (2000).
Appeal 2011-011041
Application 11/414,759

3
similar intensity . . . in which a lesion will not exhibit vast variation in pixel
values, which is interpreted to mean that the partitioned regions have
substantially similar intensity patterns” (id.). The Examiner also finds that
Kupinski teaches “segmenting the image into a plurality of regions
according to intensity patterns; selecting candidate lesion regions as those
having a lower intensity value than a threshold value; and classifying the
candidate lesion regions” (id.). In addition, the Examiner finds that
Kupinski discloses that “the task of a lesion segmentation algorithm is to
partition a set (I) into two sets . . . and further mention[s] pixel values of all
images are normalized” (id.).
The Examiner relies on Madabhushi for disclosing an “anisotropic
diffusion filter to remove speckle (or noise) . . . ; a method to merge
neighboring regions into larger regions and then appl[y] an averaging filter
to each region” (id. at 5). The Examiner relies on Shi for teaching
“normalized cut” (id.). The Examiner concludes:
[O]ne with ordinary skill in the art at the time the invention was
made would have been motivated to combine Kupinski et al
with Madabhushi et al by using an anisotropic diffusion filter
on the image and Shi et al by using normalized cut operation;
for the purpose of enhancing the signal to noise ratio; thus
improving visualization.
(Id.)
FINDINGS OF FACT
1. Kupinski discloses:
A method for the automated segmentation of an abnormality in
a medical image, including acquiring first image data
representative of the medical image; locating a suspicious site
at which the abnormality may exist; establishing a seed point
Appeal 2011-011041
Application 11/414,759

4
within the suspicious site; and preprocessing the suspicious site
with a constraint function to produce second image data in
which pixel values distant of the seed point are suppressed. . . .
The method further includes applying plural thresholds to the
second image data to partition the second image data at each
threshold; identifying corresponding first image data for the
partitioned second image data obtained at each respective
threshold; determining a respective index for each of the
partitioned first image data; and determining a preferred
partitioning, for example that partitioning leading to a
maximum index value, based on the indices determined at each
threshold, and segmenting the lesion based on the partitioning
established by the threshold resulting in the maximum index.
(Kupinski, Abstract.)
2. Kupinski also discloses that, “while the above discussion relates
largely to detection of lesions in mammograms, . . . segmentation of the
abnormality can be preformed [sic, performed] also on 3-dimensional
datasets. . . . Examples of such abnormalities for segmentation include
masses in 3-dimensional medical images (magnetic resonance imaging or
ultrasound imaging) of the breast.” (Id. at col. 12, ll. 39-57.)
3. Madabhushi discloses:
Speckle is a particular kind of noise which affects ultrasound
images, and can significantly reduce their quality and
diagnostic usefulness. . . . Guo discussed an adjacency graph
method to merge neighboring regions into larger regions and
then applying an averaging filter to each region. Other filters
such as anisotropic diffusion filter . . . have also been applied to
sonographic images to remove speckle.
(Madabhushi 158.)
4. Shi discloses:
We propose a novel approach for solving the perceptual
grouping problem in vision. Rather than focusing on local
Appeal 2011-011041
Application 11/414,759

5
features and their consistencies in the image data, our approach
aims at extracting the global impression of an image. We treat
image segmentation as a graph partitioning problem and
propose a novel global criterion, the normalized cut, for
segmenting the graph. The normalized cut criterion measures
both the total dissimilarity between the different groups as well
the total similarity within the groups.
(Shi, Abstract.)
ANALYSIS
First, we agree with Appellants that the Examiner has not set forth a
prima facie case that Kupinski teaches, either expressly or inherently, a
normalized cut operation (App. Br. 6). The evidence of record indicates that
a normalized cut operation is a specific way to partition an image (Finding
of Fact (FF) 4). Thus, although Kupinski partitions an image (FF 1), we
conclude that the Examiner has not shown that Kupinski uses a normalized
cut operation.
In addition, we agree with Appellants that the Examiner has not
adequately explained why it would have been obvious to incorporate Shi’s
normalized cut operation into Kupinski’s method (App. Br. 8). As noted
above, the Examiner concludes that it would have been obvious “to combine
Kupinski et al with Madabhushi et al by using an anisotropic diffusion filter
on the image and Shi et al by using normalized cut operation; for the
purpose of enhancing the signal to noise ratio; thus improving visualization”
(Ans. 5). However, the Examiner does not explain why a normalized cut
operation would be expected to enhance the signal to noise ratio. In fact, it
appears that this reasoning applies only to the combination of Madabhushi
with Kupinski.
Appeal 2011-011041
Application 11/414,759

6
We also agree with Appellants that the Examiner has not set forth a
prima facie case that it would have been obvious “to combine the claimed
performing, merging and selecting steps” (App. Br. 9). The Examiner notes
that Madabhushi teaches “a method to merge neighboring regions into larger
regions” (Ans. 5; see also FF 3). However, as noted by the Examiner,
Madabhushi “then applie[s] an averaging filter to each region” (id.). The
Examiner has not adequately explained why it would have been obvious to
merge a plurality of regions formed by a normalized cut operation.
CONCLUSION
The Examiner has not set forth a prima facie case that Kupinski,
Madabhushi, and Shi suggest the method of claim 1. We therefore reverse
the obviousness rejection of claim 1 and of claims 2 and 4-13, which depend
from or otherwise incorporate the features of claim 1.

REVERSED

cdc