Ex Parte Aleksovski

Patent Trial and Appeal Board

Dec 18, 2018

13786524 (P.T.A.B. Dec. 18, 2018)

UNITED STA TES p A TENT AND TRADEMARK OFFICE
APPLICATION NO. FILING DATE FIRST NAMED INVENTOR
13/786,524 03/06/2013 Zharko Aleksovski
24737 7590 12/20/2018
PHILIPS INTELLECTUAL PROPERTY & STANDARDS
465 Columbus A venue
Suite 340
Valhalla, NY 10595
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
ATTORNEY DOCKET NO. CONFIRMATION NO.
2011P02041US01 6809
EXAMINER
CHOW,DOONY
ART UNIT PAPER NUMBER
2174
NOTIFICATION DATE DELIVERY MODE
12/20/2018 ELECTRONIC
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.
Notice of the Office communication was sent electronically on above-indicated "Notification Date" to the
following e-mail address(es):
patti. demichele@Philips.com
marianne.fox@philips.com
katelyn.mulroy@philips.com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte ZHARKO ALEKSOVSKI
Appeal2017-001040 1
Application 13/786,524
Technology Center 2100
Before HUNG H. BUI, AMBER L. HAGY, and MICHAEL J. ENGLE,
Administrative Patent Judges.
ENGLE, Administrative Patent Judge.
DECISION ON APPEAL
Appellant appeals under 35 U.S.C. § 134(a) from a final rejection of
claims 1, 5, and 12-19, which are all of the claims pending in the
application. We have jurisdiction under 35 U.S.C. § 6(b ).
We REVERSE.
Technology
The application relates to a document management system providing
visualizations of medical records. Spec. 1.
Illustrative Claim
Claim 1 is illustrative and reproduced below with certain limitations at
issue emphasized:
1 According to Appellant, the real party in interest is Koninklijke Philips,
N.V. App. Br. 1.
Appeal2017-001040
Application 13/786,524
1. A medical records visualization system comprising:
a medical records database which stores medical records
of a plurality of patients relating to a plurality of medical
conditions, the medical records including natural language text;
an electronic processing system including:
an extraction unit for processing a first medical
record in the medical records database to extract an
indicator from the natural language text of the first
medical record, wherein the indicator is indicative of a
reference to a second medical record in the medical
record database, wherein the indicator includes a date and
a patient identifier associated with the first medical record,
a control unit for causing the extraction unit to
process a plurality of the medical records of the medical
records database to extract indicators from the natural
language text of the medical records, wherein the
indicators are indicative of references to other medical
records and include a date and a patient identifier in the
natural language text of the second medical record,
a reference determiner for matching the date and/ or
the patient identifier in the natural language text of the first
medical record with the date and/or patient identifier
associated with the second medical record and/or a
structured data field of the second medical record,
a concept determiner for determining at least one
concept addressed in the matched medical records, the
concept including a medical condition,
a clustering unit for clustering the medical records
based on the indicators, the concepts, and the patient
identifier to generate clusters of medical records relating
to a common medical condition of the identified patient
wherein each of the clustered medical records references
at least one other of the clustered medical records, and
a visualization unit for generating a visualization
comprising (a) a visual representation of the first medical record,
the second medical record, additional medical records, the
2
Appeal2017-001040
Application 13/786,524
visualized medical records being arranged in the clusters, (b)
lines or arrows linking the visualized medical records of each
cluster that reference each other, and ( c) a timeline.
Examiner's Rejections & References
Claims 1, 5, 12-14, 16, 17, and 19 stand rejected under 35 U.S.C.
§ 103 (a) as obvious over the combination of Morita (US 2008/0208631 A 1;
Aug. 28, 2008), Banks (US 2006/0190817 Al; Aug. 24, 2006), Hally (US
2004/0207666 Al; Oct. 21, 2004), and Soules (US 2010/0115001 Al; May
6, 2010). Final Act. 4.
Claim 15 stands rejected under 35 U.S.C. § I03(a) as obvious over the
combination of Morita, Banks, Hally, Soules, and Mittal (US 8,612,411 Bl;
Dec. 17, 2013). Final Act. 14--15.
Claim 18 stands rejected under 35 U.S.C. § I03(a) as obvious over the
combination of Morita, Banks, Hally, Soules, and Baker (US 8,015,136 Bl;
Sept. 6, 2011 ). Final Act. 15. 2
ISSUE
Did the Examiner err in finding Banks teaches or suggests an
extraction unit "to extract an indicator from the natural language text of the
first medical record, wherein the indicator is indicative of a reference to a
second medical record in the medical record database, wherein the indicator
includes a date and a patient identifier associated with the first medical
record," as recited in claim 1?
ANALYSIS
Claim 1 recites an extraction unit "to extract an indicator from the
natural language text of the first medical record, wherein the indicator is
2 The rejection under 35 U.S.C. § 101 was withdrawn. Ans. 2.
3
Appeal2017-001040
Application 13/786,524
indicative of a reference to a second medical record in the medical record
database, wherein the indicator includes a date and a patient identifier
associated with the first medical record."
Claim 1 further recites a clustering unit for "clustering the medical
records based on the indicators ... wherein each of the clustered medical
records references at least one other of the clustered medical records" and a
visualization unit for generating a visualization comprising "lines or arrows
linking the visualized medical records ... that reference each other."
Based on this claim language, an indicator is language in the natural
language text of a first medical record-such as a date or name-that is
indicative of a reference in the first medical record to a second medical
record. Examples of the claimed "indicators" are shown in Figures 6 and 7,
and examples of the claimed "visualization" are shown in Figures 4 and 5.
Figure 7 is reproduced below.
Fig 7
Figures 6 and 7 illustrate "two documents 601 and 602" that are
"highly related to one another" because document 602 ( on the right)
discusses the results of an x-ray ordered in document 601 ( on the left).
4
Appeal2017-001040
Application 13/786,524
Spec. 12:23-27. But the relationship between documents 601 and 602 may
not be known "to a doctor who looks at a plain list of documents." Id. at
12:27-28. The present application proposes to address this problem:
Using a system as described herein, the relevant indicators may
be extracted from the follow-up document 602, and the reference
to the original document 601 may be resolved. The system first
extracts the indicators "Dr. John Smith", "09.Nov.2003",
and/or "breast cancer". Then, it looks up corresponding
information in other documents, including document 601. In the
document 601, the system finds corresponding information of the
person who wrote the document "Dr. John Smith", letter date
"09-Nov-2003", and "breast cancer", as indicated by means of
the arrows in Fig. 7. Accordingly, the system will establish that
document 602 refers to document 601, and will show an
indication 405 of this reference when generating the visualization
[as shown in Figure 4].
Spec. 12:28-13:2 (emphasis added).
The Examiner finds Morita teaches Appellant's claimed "medical
records visualization system," including, inter alia, "a medical records
database which stores medical records." Final Act. 4 ( citing Morita ,r,r 23-
26, 28, 77, Fig. 7). The Examiner then relies on Banks for teaching the
claimed indicator and extraction of an indicator. Id. at 4---6 ( citing Banks
,r,r 21-26, Fig. 1 ). In particular, the Examiner finds that Banks teaches a
user entering a keyword ( such as the letter "a") and the system determining
which documents contain that keyword. Id. at 5---6 ( citing Banks ,r 30). "In
other words, the letter 'a' is an indicator that is indicative of a reference to at
least a second item (i.e., a second medical record) in the subset of items (i.e.,
a medical record database)." Id.
Although we do not agree with all of Appellant's arguments, we do
agree with Appellant that Bank's filtering documents by keyword, by itself,
5
Appeal2017-001040
Application 13/786,524
fails to teach or suggest "to extract an indicator from the natural language
text of the first medical record" that "is indicative of a reference to a second
medical record," as required by claim 1. For example, George Orwell's
Animal Farm and the children's song Old MacDonald Had a Farm both
contain the keyword "farm," but that does not mean that the word "farm" in
Old MacDonald is indicative of a reference to Animal Farm, or vice versa.
The Examiner is correct that an "indicator" does not require an express link
(see Ans. 13); nonetheless, the plain language of the claim requires more
than mere circumstantial use of the same word.
Claims 12-14 recite commensurate limitations with respect to "an
indicator ... indicative of a reference to a second medical record."
Accordingly, we are constrained to reverse the Examiner's rejection
of claims 1, 5, and 12-19.
DECISION
For the reasons above, we reverse the Examiner's decision rejecting
claims 1, 5, and 12-19.
REVERSED
6