Best Medical International, Inc.

Patent Trials and Appeals Board

Apr 23, 2021

IPR2020-00071 (P.T.A.B. Apr. 23, 2021)

Trials@uspto.gov Paper 70
571-272-7822 Entered: April 23, 2021
UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
____________

VARIAN MEDICAL SYSTEMS, INC. and ELEKTA INC.,
Petitioner,

v.

BEST MEDICAL INTERNATIONAL, INC.,
Patent Owner.
____________

IPR2020-000711
Patent 6,393,096 B1
____________

Before KARL D. EASTHOM, JOHN A. HUDALLA, and
AVELYN M. ROSS, Administrative Patent Judges.

HUDALLA, Administrative Patent Judge.

JUDGMENT
Final Written Decision
Determining Some Challenged Claims Unpatentable
Denying-in-Part and Dismissing-in-Part Petitioner’s Motion to Exclude
35 U.S.C. § 318(a); 37 C.F.R. § 42.64

Varian Medical Systems, Inc. (“Varian”) filed a Petition (Paper 2,
“Pet.”) requesting an inter partes review of claims 1 and 18 of U.S. Patent
No. 6,393,096 B1 (Ex. 1001, “the ’096 patent”). Patent Owner, Best

1 Petitioner from IPR2020-00971 has joined this proceeding.
IPR2020-00071
Patent 6,393,096 B1
2
Medical International, Inc. (“Patent Owner”), filed a Preliminary Response
(Paper 6, “Prelim. Resp.”). Varian also filed a Preliminary Reply (Paper 8),
and Patent Owner filed a Preliminary Sur-reply (Paper 9). Taking into
account the arguments presented in these papers, we determined that the
information presented in the Petition established that there was a reasonable
likelihood that Varian would prevail with respect to its unpatentability
challenges. Pursuant to 35 U.S.C. § 314, we instituted this proceeding on
May 1, 2020, as to all challenged claims and all grounds of unpatentability.
Paper 14 (“Dec. on Inst.”).
Elekta Inc. (“Elekta”) subsequently filed a similar petition and motion
for joinder in IPR2020-00971. See IPR2020-00971, Papers 2, 3. We
instituted an inter partes review and joined Elekta as a party to this
proceeding in a limited capacity. See IPR2020-00971, Paper 7. Henceforth,
we refer collectively to Varian and Elekta as “Petitioner.”
During the course of trial, Patent Owner filed a Patent Owner
Response (Papers 33,2 34,3 “PO Resp.”), and Petitioner filed a Reply to the
Patent Owner Response (Paper 47, “Pet. Reply”). Patent Owner also filed a
Sur-reply. Paper 52 (“PO Sur-reply”). An oral hearing was held on
January 28, 2021, and a transcript of the hearing is included in the record.
Paper 68 (“Tr.”).

2 This paper was sealed in accordance with the Protective Order entered in
this case. See Paper 32, Attach. A (copy of protective order); Paper 43
(entering protective order).
3 This paper is the public version.
IPR2020-00071
Patent 6,393,096 B1
3
Petitioner filed declarations of Kenneth P. Gall, Ph.D. with its Petition
(Ex. 1002) and its Reply (Ex. 1043). Petitioner also filed declarations of
Sylvia Hall-Ellis, Ph.D. (Ex. 1003) and Christopher Butler (Ex. 1004) with
its Petition. Patent Owner filed declarations of Daniel J. Chase with its
Preliminary Response (Ex. 2002) and with its Response (Ex. 2037). Patent
Owner also filed declarations of Dr. Mark P. Carol (Ex. 2043), Merle E.
Romesberg III (Ex. 20444), and Thomas Rowden (Ex. 2045) with its
Response. The parties also filed transcripts of the depositions of Dr. Gall
(Ex. 2041), Mr. Chase (Exs. 1044, 1045), Dr. Carol (Ex. 1046), Mr. Rowden
(Ex. 1047), and Mr. Romesberg (Exs. 1048,5 1049, 10506).
Petitioner filed a motion to exclude certain testimony from Mr. Chase
and Mr. Romesberg. Paper 56 (“Exclude Mot.”). Patent Owner filed an
opposition (Paper 58, “Exclude Opp.”), and Petitioner filed a reply
(Paper 62, “Exclude Reply”).
We have jurisdiction under 35 U.S.C. § 6. This decision is a Final
Written Decision under 35 U.S.C. § 318(a) as to the patentability of claims 1
and 18 of the ’096 patent. For the reasons discussed below, Petitioner has
demonstrated by a preponderance of the evidence that claim 1 of the ’096
patent is unpatentable. Petitioner has not demonstrated by a preponderance
of the evidence that claim 18 of the ’096 patent is unpatentable.

4 The record includes sealed and public versions of this exhibit in
accordance with the protective order in this case.
5 This exhibit is the public version of the transcript at Exhibit 1050.
6 This exhibit was sealed in accordance with the Protective Order entered in
this case.
IPR2020-00071
Patent 6,393,096 B1
4

I. BACKGROUND
A. Real Parties-in-Interest
Varian identifies Varian Medical Systems, Inc., VMS International
AG, VMS International Holdings, Inc., VMS Netherlands Holdings, Inc.,
and VMS Nederland BV as real parties-in-interest. Pet. 3. Elekta identifies
Elekta Limited (UK), Elekta Holdings U.S., Inc., and Elekta AB as real
parties-in-interest. IPR2020-00971, Paper 2, 4. Patent Owner identifies
Best Medical International, Inc. as the real party-in-interest. Paper 3, 1.

B. Related Proceedings
The parties identify the following proceedings related to the
’096 patent (Pet. 4; Paper 3, 1–2; IPR2020-00971, Paper 2, 4):
Best Med. Int’l, Inc. v. Elekta Inc., No. 1:19-cv-03409-MLB (N.D.
Ga.);
Best Med. Int’l, Inc. v. Elekta AB, No. 1:18-cv-01600-MN (D. Del.);
Best Med. Int’l, Inc. v. Varian Med. Sys., Inc., No. 1:18-cv-01599 (D.
Del.); and
Varian Med. Sys., Inc. v. Best Med. Int’l, Inc., IPR2020-00072, which
challenges claims 43, 44, and 46 of the ’096 patent. We issue a final written
decision in IPR2020-00072 concurrently herewith.
We also note that another petitioner filed a petition requesting an inter
partes review of the ’096 patent in IPR2020-00074. We denied institution in
that case.
The parties belatedly notified us of a pending ex parte reexamination
of claims 1, 3–8, 18, 21–24, and 37–42 of the ’096 patent in Reexamination
IPR2020-00071
Patent 6,393,096 B1
5
Control No. 90/014,424. PO Sur-reply 1; Paper 65, 2–3; Paper 67, 1;
Ex. 1051 (reexamination request); Ex. 1052 (order and decision granting
request). As part of the reexamination, Patent Owner has canceled claim 1
in an amendment. PO Sur-reply 1; Ex. 1053 (excerpts from Nov. 23, 2020,
amendment), 1, 4.7 Although Patent Owner contends that, as a result,
claim 18 “is the only claim now at issue” (PO Sur-reply 1), we note that the
reexamination is still pending, and that Patent Owner has not filed a
statutory disclaimer of claim 1. Accordingly, because claim 1 has not yet
been canceled by any final action, we address the patentability of claim 1
below.
We additionally note that Petitioner challenged other patents owned
by Patent Owner in IPR2020-00053, IPR2020-00075, IPR2020-00076, and
IPR2020-00077. We denied institution in all of these cases except for
IPR2020-00076, for which we recently issued a final written decision.

C. The ’096 patent
The ’096 patent is directed to “determining an optimized radiation
beam arrangement for applying radiation to a tumor target volume while
minimizing radiation of [another] structure volume in a patient.” Ex. 1001,
Abstr. Optimized treatment plans are created using a computational method
(such as simulated annealing radiotherapy planning (SARP)) based on an
objective cost function that attributes costs of radiation of various portions of
both the tumor and surrounding tissues/structures. Id. at 3:17–22, 5:3–10.
Nevertheless, the ’096 patent alleges that the cost functions in then-existing

7 We refer to Petitioner’s added page numbers in Exhibit 1053.
IPR2020-00071
Patent 6,393,096 B1
6
methods relied merely on costs related to discrete points within the structure,
and did not account for the structure volumes as a whole or for the relative
importance of varying surrounding structure types. Id. at 3:25–29. Further,
the ’096 patent alleges that then-existing methods did not allow physicians
to utilize Cumulative Dose Volume Histogram (CDVH) curves in
establishing desired dose distributions. Id. at 3:48–51.
The ’096 patent describes a treatment planning system that accounts
for multiple treatment parameters for both a target and multiple surrounding
structure types. Ex. 1001, 5:54–56. The system arrives at an optimal beam
arrangement “by computationally increasing the proposed beam weight
iteratively [and] incorporating cost functions to ensure that an iterative
change in the beam weight would not result in an unacceptable exposure to
the volumes of tissue or other structures being subjected to the proposed
dose.” Id. at 5:39–44. The system includes a modified cost function that
allows a physician to use conventional CDVHs to establish a desired dose
for both the target volume and each involved structure; the CDVHs are used
as input for the treatment planning system. Id. at 5:57–64.
Figures 3 and 4 of the ’096 patent are reproduced below.
IPR2020-00071
Patent 6,393,096 B1
7

Figures 3 and 4 show composite CDVH curves 10, 20, respectively. Id. at
8:64–65. In Figure 3, composite CDVH curve 10 includes desired target
CDVH curve 100 and proposed CDVH target curve 101, the latter of which
reflects the effect of a prescription proposed by the system during a given
iteration of plan optimization. Id. at 6:40–44, 8:60–64. In Figure 4,
composite CDVH curve 20 includes desired structure CDVH curve 2008 and
proposed CDVH structure curve 201, the latter of which again reflects the

8 In Figure 4, the callout arrow for reference numeral 200 appears displaced
slightly from the CDVH curve it references (which is a solid line).
IPR2020-00071
Patent 6,393,096 B1
8
effect of a prescription proposed by the system during a given iteration of
plan optimization. Id. Certain control points or regions N, N′, Q, Q′, X, and
X′ of composite CDVH curves 10, 20 may be identified as being more
important for a particular type of target or structure. Id. at 8:67–9:3. Each
control point or control region value is used as an input variable to a
parameterized influence function for each target or structure. Id. at 10:40–
44. The resultant values from the influence function calculation for each
control point or control region value are summed to produce a final cost of
the proposed beam weights reflected by proposed CDVH curve 101, 201
during a given iteration. Id. at 10:44–50.
The ’096 patent issued from an application that was filed May 27,
1999, which claims priority to a provisional application filed on May 27,
1998. Ex. 1001, codes (22), (60). As discussed below, Petitioner attempts
to establish that, at a minimum, its asserted references qualify as prior art
relative to the May 27, 1998, filing date of the provisional application. See
Pet. 28.

D. Illustrative Claim
Of the challenged claims, claim 1 is independent. Claim 18 is a
multiple dependent claim that depends from, inter alia, claim 1. Claim 1 is
illustrative of the challenged claims:
1. A method of determining an optimized radiation beam
arrangement for applying radiation to a tumor target volume
IPR2020-00071
Patent 6,393,096 B1
9
while minimizing radiation of a structure volume in a patient,
comprising the steps of:
using a computer to computationally obtain a proposed
radiation beam arrangement;
using a computer to computationally change the proposed
radiation beam arrangement iteratively,
incorporating a cost function at each iteration to approach
correspondence of a CDVH associated with the proposed
radiation beam arrangement to a CDVH associated with a
predetermined desired dose prescription;
comparing the dose distribution to a prescribed dose for
the tumor volume and surrounding tissue structures, and
increasing or decreasing radiation beam intensity if the
change of the proposed beam arrangement leads to a greater
correspondence to the desired dose prescription to obtain an
optimized radiation beam arrangement.
Ex. 1001, 16:39–57.

E. Prior Art
Petitioner relies on the following prior art:
Viggars D.A., et al., “The Objective Evaluation of
Alternative Treatment Plans III: The Quantitative Analysis of
Dose Volume Histograms,” International Journal of Radiation
Oncology • Biology • Physics, 23:419–27 (1992) (Ex. 1015,
“Viggars”);
Oldham, M. et al., “A comparison of conventional
‘forward planning’ with inverse planning for 3D conformal
radiotherapy of the prostrate,” Radiotherapy and Oncology,
35:248–62 (1995) (Ex. 1019, “Oldham”);
Carol, M.P., Chapter 2 – IMRT: Where We Are Today, The
Theory & Practice of Intensity Modulated Radiation Therapy
17–36 (1997) (Ex. 1020, “Carol-2”);
IPR2020-00071
Patent 6,393,096 B1
10
Carol, M.P., Chapter 17 – Where We Go From Here: One
Person’s Vision, The Theory & Practice of Intensity Modulated
Radiation Therapy 243–52 (1997) (Ex. 1021, “Carol-17”); and
Morrill, S.M. et al., “Treatment planning optimization
using constrained simulated annealing,” Phys. Med. Biol.,
36(10):1341–61 (1991) (Ex. 1022, “Morrill-1991”).

F. The Instituted Grounds
We instituted inter partes review of claims 1 and 18 of the ’096 patent
on the following grounds (Dec. on Inst. 35), which are all the grounds
presented in the Petition (Pet. 7):
Claim(s) Challenged 35 U.S.C. § References
1 103(a)9 Oldham, Viggars
18 103(a) Oldham, Viggars, Morrill-1991
1, 18 103(a) Carol-2, Carol-17
18 103(a) Carol-2, Carol-17, Morrill-1991

II. ANALYSIS
A. Legal Standards
A claim is unpatentable under 35 U.S.C. § 103(a) if the differences
between the claimed subject matter and the prior art are such that the subject
matter, as a whole, would have been obvious at the time the invention was
made to a person having ordinary skill in the art to which said subject matter
pertains. See KSR Int’l Co. v. Teleflex Inc., 550 U.S. 398, 406 (2007).

9 The Leahy-Smith America Invents Act (AIA), Pub. L. No. 112-29, 125
Stat. 284, 287–88 (2011), amended 35 U.S.C. § 103. Because the
’096 patent was filed before March 16, 2013 (the effective date of the
relevant amendment), the pre-AIA version of § 103 applies.
IPR2020-00071
Patent 6,393,096 B1
11
The question of obviousness is resolved on the basis of underlying factual
determinations, including (1) the scope and content of the prior art; (2) any
differences between the claimed subject matter and the prior art; (3) the level
of skill in the art; and (4) where in evidence, so-called secondary
considerations.10 See Graham v. John Deere Co., 383 U.S. 1, 17–18 (1966).
We also recognize that prior art references must be “considered together
with the knowledge of one of ordinary skill in the pertinent art.” In re
Paulsen, 30 F.3d 1475, 1480 (Fed. Cir. 1994) (quoting In re Samour,
571 F.2d 559, 562 (CCPA 1978)).

B. Level of Ordinary Skill in the Art
The level of ordinary skill in the art is “a prism or lens through which
. . . the Board views the prior art and the claimed invention” to prevent
hindsight bias. Okajima v. Bourdeau, 261 F.3d 1350, 1355 (Fed. Cir. 2001).
In determining the level of ordinary skill, various factors may be considered,
including the “types of problems encountered in the art; prior art solutions to
those problems; rapidity with which innovation are made; sophistication of
the technology; and educational level of active workers in the field.” In re
GPAC Inc., 57 F.3d 1573, 1579 (Fed. Cir. 1995) (internal quotation and
citation omitted). Generally, it is easier to establish obviousness under a
higher level of ordinary skill in the art. Innovention Toys, LLC v. MGA
Entm’t, Inc., 637 F.3d 1314, 1323 (Fed. Cir. 2011) (“A less sophisticated
level of skill generally favors a determination of nonobviousness . . . while a
higher level of skill favors the reverse.”).

10 The record does not contain any evidence of secondary considerations of
nonobviousness.
IPR2020-00071
Patent 6,393,096 B1
12
Citing testimony from Dr. Gall, Petitioner contends a person having
ordinary skill in the art (POSA) would have been “a medical physicist with a
Ph.D. (or similar advanced degree) in physics, medical physics, or a related
field,” and would have had “two or more years of experience in radiation
oncology physics, treatment planning, treatment plan optimization related to
radiation oncology applications, and computer programming associated with
treatment plan optimization.” Pet. 20–21 (citing Ex. 1002 ¶ 16). Patent
Owner cites testimony from Mr. Chase and contends an ordinarily skilled
artisan would have “earned at least a master’s or doctoral degree in radiation
dosimetry, physics, medical physics, or medicine, or equivalent disciplines”
and would have had “three years of clinical experience in radiation treatment
planning.” PO Resp. 15–16 (citing Ex. 2037 ¶¶ 60, 67–72); see also Prelim.
Resp. 18–19 (same definition). For purposes of our Decision on Institution,
we adopted Patent Owner’s definition. Dec. on Inst. 9.
The parties’ proposed definitions do not differ greatly as to the level
and type of formal education, and the parties have not raised any dispute
related to formal education. Thus, as to that aspect, we continue to apply
Patent Owner’s articulation that a person of ordinary skill in the art would
have earned “at least a master’s or doctoral degree in radiation dosimetry,
physics, medical physics, or medicine, or equivalent disciplines.” See PO
Resp. 15–16.
Nevertheless, the parties have put forth great effort disputing the type
of experience that an ordinarily skilled artisan would have possessed.11

11 We note that the parties’ definitions do not differ meaningfully as to the
number of years’ experience that would have been possessed by an
IPR2020-00071
Patent 6,393,096 B1
13
Specifically, Petitioner criticizes Patent Owner’s definition insofar as
“Patent Owner’s purported POSA is unable to design and write the computer
programming necessary to understand and implement the teachings of the
’096 patent and asserted prior art.” Pet. Reply 2 (citing Ex. 1043 ¶¶ 5–11).
Petitioner also argues that Patent Owner’s definition would result in “a
POSA [who] did not [have] the knowledge or skill to be able to make the
inventions disclosed and claimed in the ’096 patent.” Id. at 2–3 (citing
Ex. 1043 ¶ 8; Ex. 1045, 10:7–17).
Citing testimony from Mr. Chase, Patent Owner draws a distinction
regarding Petitioner’s requirement for computer programming experience
between “the use and manipulation of programs by users” and “the writing
of the underlying software code by the system designer and author of such
programs.” PO Sur-reply 1–2 (citing Ex. 2002 ¶ 69). Patent Owner also
notes that the challenged claims are method claims directed to optimization
planning, and, as such, do not require software programming. Id. at 2–3
(citing Ex. 1001, 18:23–26).
At the outset, we note that in determining the level of ordinary skill in
the art, we consider the inventions of the ’096 patent as a whole, and not just
inventions related to individual claims. See Hologic, Inc. v. Minerva
Surgical, Inc., 764 F. App’x 873, 879 (Fed. Cir. 2019) (considering broader
teachings in a challenged patent’s specification, and not just the more
narrowly drawn claims, when evaluating the Board’s determination of a
level of ordinary skill). At the oral hearing, Patent Owner’s counsel
acknowledged that this was the proper approach. Tr. 61:8–63:11. Thus, we

ordinarily skilled artisan, i.e., “two or more” for Petitioner versus “three” for
Patent Owner. As mentioned below, we apply two years of experience.
IPR2020-00071
Patent 6,393,096 B1
14
do not agree with Patent Owner’s attempt (PO Sur-reply 2–3) to limit the
level of ordinary skill based on the fact that the challenged claims here are
method claims. We instead consider the totality of the ’096 patent
disclosure when determining the level of ordinary skill.
The ’096 patent “relates to a method and apparatus for conformal
radiation therapy of tumors with a radiation beam having a pre-determined,
constant beam intensity.” Ex. 1001, 1:10–12. The disclosed methods and
apparatus are based on known SARP methods, where “[t]he optimal beam
arrangement is arrived at by computationally increasing the proposed beam
weight iteratively, incorporating cost functions to ensure that an iterative
change in the beam weight would not result in an unacceptable exposure to
the volumes of tissue or other structures being subjected to the proposed
dose.” Id. at 5:39–44; see also id. at 8:37–39 (“A SARP technique is
utilized to do this optimization”). According to the specification of the
’096 patent,
[t]he system of the present invention includes . . . a modified
cost function, which allows a physician to use conventional
cumulative dose volume histographs (“CDVH”s) to establish a
desired prescription of dosage to both the target volume, or
target, and each involved structure volume, or structure, which
will then be used as input for the system for determining the
proposed dose distribution for delivery to a patient. The
optimization method may be carried out using conventional
equipment, including a conventional linear accelerator
(“LINAC”) 300, as shown in FIG. 1, having a rotatable gantry,
a conventional computer or set of computers, and plan
optimization software, which utilizes the optimization method
of the present invention.
Id. at 5:54–6:2. Therefore, the ’096 patent describes a different way of
optimizing a treatment plan for use with “conventional” radiation therapy
IPR2020-00071
Patent 6,393,096 B1
15
equipment. The main difference from the prior art is the modified cost
function that is used.
The specification goes on to describe the cost function with respect to
“Plan Optimization step 803” of Figure 2. The mathematical formulas for
two exemplary cost/influence functions (INF1 and INF2) are provided at
column 10, line 51 through column 11, line 39. According to the
specification,
[a] value is calculated for each control point value . . . of each
CDVH curve of each target and structure according to the
influence function INF1 or INF2. The total cost for the
proposed dose represented by the proposed CDVH curve may
then be obtained by summing each value of INF1 or INF2 for
each control point value of each CDVH curve of each target
and structure.
Id. at 11:40–47. As such, the modified cost functions are based on CDVH
curve conformity, which is different from the prior art, but they are applied
iteratively just as in prior art SARP methods. See id. at 5:39–53.
The specification also describes exemplary computer hardware for
implementing the disclosed invention:
A suitable computer is utilized in performing the Plan
Optimization step 802 [sic, 803] (FIG. 2), as well as the other
steps of the radiation planning system. For illustration purposes
only, a programmable 150 Mhz pentium computer with four
symmetric multiprocessors, running the Sun Solaris operating
system, and having 256 megabytes RAM could be utilized in
performing the Plan Optimization step 802 [sic, 803] (FIG. 2).
Ex. 1001, 8:52–59.
As indicated by the specification of the ’096 patent, the disclosed
invention is implemented with computers and involves iterative calculations
of cost functions, just like prior art SARP methods. See Ex. 1001, 3:17–22
IPR2020-00071
Patent 6,393,096 B1
16
(“Existing methods and apparatus utilize a computational method of
establishing optimized treatment plans based on an objective cost function
that attributes costs of radiation of various portions of both the tumor and
surrounding tissues, or structures. One such computational method is known
in the art as simulated annealing.”). This supports Petitioner’s contention
that a person of ordinary skill in the art would have had experience in
programming computers in the context of treatment plan optimization. See
Pet. 20–21. This type of experience is consistent with implementing
different cost functions using SARP techniques.
In contrast, Patent Owner’s proposed type of experience—“clinical
experience in radiation treatment planning” (PO Resp. 15–16)—does not
reflect the type of skills necessary to implement different types of cost
functions in an iterative SARP-based plan optimization system. We are not
aware of any record evidence that a clinician in 1998 would have been able
to program different types of cost functions in a treatment planning system.
And, even if certain clinicians might have possessed these skills at the time,
we find Petitioner’s proposed level of experience to better reflect the
relevant skills necessary to implement various cost functions in a treatment
planning system, thereby changing the way a system iteratively computes
costs as part of an optimization method. For this reason, we reject Patent
Owner’s proposed type of experience and adopt Petitioner’s formulation:
two years of experience in radiation oncology physics, treatment planning,
treatment plan optimization related to radiation oncology applications, and
computer programming associated with treatment plan optimization.
We additionally reject Patent Owner’s and Mr. Chase’s suggestion
that “computer programming” as used in Petitioner’s proposed level of
IPR2020-00071
Patent 6,393,096 B1
17
ordinary skill “could refer . . . to the use and manipulation of programs by
users, including medical physicists such as [Mr. Chase].” PO Sur-reply 2
(citing Ex. 2002 ¶ 69); see also Exclude Opp. 3–4 (same argument). Mere
use and manipulation of existing treatment planning system programs is not
the same as programming such systems in the first instance. In making this
determination, we have considered certain instances in the ’096 patent
specification where the word “program” describes the entry of data into
existing treatment planning systems by a clinician. See Ex. 1001, 1:35–43,
13:30–35 (both using the word “program” in the sense of a clinician using
existing systems). Although the specification sometimes uses the word
“program” in the sense that a clinician enters a desired radiation prescription
or selects influence parameters in an existing system (see id.), our
articulation of ordinary skill is meant to reflect “formal computer
programming, i.e., designing and writing underlying computer code” (see
Ex. 2002 ¶ 69), as would have been necessary to implement different cost
functions in an iterative optimization method.
For these reasons, and based on the entire trial record, we determine
that an ordinarily skilled artisan (1) would have earned at least a master’s or
doctoral degree in radiation dosimetry, physics, medical physics, or
medicine, or equivalent disciplines, and (2) would have had two years of
experience in radiation oncology physics, treatment planning, treatment plan
optimization related to radiation oncology applications, and computer
programming associated with treatment plan optimization.

IPR2020-00071
Patent 6,393,096 B1
18
C. Claim Interpretation
In an inter partes review, we construe each claim “in accordance with
the ordinary and customary meaning of such claim as understood by one of
ordinary skill in the art and the prosecution history pertaining to the patent.”
37 C.F.R. § 42.100(b) (2019). Accordingly, our claim construction standard
is the same as that of a district court. See id. Under the standard applied by
district courts, claim terms are generally given their plain and ordinary
meaning as would have been understood by a person of ordinary skill in the
art at the time of the invention and in the context of the entire patent
disclosure. Phillips v. AWH Corp., 415 F.3d 1303, 1313 (Fed. Cir. 2005)
(en banc). “There are only two exceptions to this general rule: 1) when a
patentee sets out a definition and acts as his own lexicographer, or 2) when
the patentee disavows the full scope of a claim term either in the
specification or during prosecution.” Thorner v. Sony Comput. Entm’t Am.
LLC, 669 F.3d 1362, 1365 (Fed. Cir. 2012).

1. “Computer” Limitations
Petitioner contends that the limitations “a computer to
computationally obtain a proposed radiation beam arrangement” and “a
computer to computationally change the proposed radiation beam
arrangement iteratively” in claim 1 should be construed as means-plus-
function limitations under pre-AIA 35 U.S.C. § 112 ¶ 6. Pet. 21–27.
Petitioner contends an ordinarily skilled artisan would not have understood
these terms to have a sufficiently definite meaning as a name for structure.
Id. at 23 (citing Ex. 1002 ¶ 60), 26 (citing Ex. 1002 ¶ 65); see also Pet.
Reply 4–5 (same argument).
IPR2020-00071
Patent 6,393,096 B1
19
Patent Owner disputes that treatment under § 112 ¶ 6 should apply.
PO Resp. 16–17; see also Prelim. Resp. 19–22. In its Preliminary Response,
Patent Owner argued that the lack of the words “means for” in these
limitations creates a rebuttable presumption that § 112 ¶ 6 does not apply.
Prelim. Resp. 19–20 (citing Phillips, 415 F.3d at 1310). Patent Owner also
contended that an ordinarily skilled artisan would have readily understood
what a computer is. Id. at 21–22 (citing Ex. 2002 ¶ 78); see also Pet. 27–28
(Petitioner’s alternate position that its grounds sufficiently identify the
recited computer). In its Response, Patent Owner notes that the district court
in the underlying litigation did not treat these limitations under § 112 ¶ 6.
PO Resp. 17 (citing Ex. 2042, 2, 6–8).
In our Decision on Institution, we agreed with Patent Owner’s
reasoning that § 112 ¶ 6 does not apply to these limitations. Dec. on Inst.
10. The parties’ positions have not changed since the time of institution, and
we discern no reason to change our initial construction. Therefore, we apply
the plain and ordinary meaning to these limitations, and we do not apply
§ 112 ¶ 6.

2. The Effect of the Transitional Word “Comprising” on Claims 1
and 18
The parties also put forth arguments about how the limitation
“incorporating a cost function at each iteration” of claim 1 should be
construed in light of the transitional word “comprising.” These arguments
arise in the context of claim 18, which depends from and includes the
IPR2020-00071
Patent 6,393,096 B1
20
limitations of claim 1.12 PO Resp. 38–42; Pet. Reply 16–23; PO Sur-reply
6–7. Petitioner argues that the transitional word “comprising” in claim 1 is
open-ended and allows for additional steps. Pet. Reply 16 (citing In re
Affinity Labs of Tex., LLC, 856 F.3d 902, 907 (Fed. Cir. 2017). As a result,
Petitioner argues that we should reject any interpretation of “incorporating a
cost function at each iteration” that requires a cost function to be calculated
after each and every iteratively proposed beam arrangement of an
optimization. Id. According to Petitioner, any such interpretation “is
improperly narrow and would be trivial to design around.” Id. at 17 (citing
Ex. 1043 ¶ 38). Patent Owner disputes Petitioner’s argument and contends
that Petitioner’s argument based on the word “comprising” would allow it to
entirely omit the added limitation in claim 18. PO Sur-reply 7.
Although we do not agree with Patent Owner’s exact reasoning, we do
agree with Patent Owner that Petitioner’s proposed construction based on
the word “comprising” goes too far. The issue is whether—even despite the
transitional word “comprising”—the method of claim 1 must always
incorporate a cost function for each cycle of the recited method (i.e., “at
each iteration”). We determine that it must. Our reviewing court considered
a similar situation in a case styled Dippin’ Dots, Inc. v. Mosey, 476 F.3d
1337 (Fed. Cir. 2007). The plaintiff brought an infringement action based
on a method claim for making a form of cryogenically prepared novelty ice
cream product in the form of “beads.” Id. at 1339–40. The steps of the
method recited various actions with respect to the “beads,” and the trial court
construed “beads” to require droplets with spherical appearance at the

12 The parties do not raise any disputes regarding these limitations with
respect to Petitioner’s analysis of claim 1.
IPR2020-00071
Patent 6,393,096 B1
21
exclusion of irregular or odd-shaped particles. Id. at 1342–43. The plaintiff
argued that the transition word “comprising” in the claim extended the scope
of the claimed method beyond a beads-only process. Id. at 1343. The
Dippin’ Dots court disagreed, noting that “‘[c]omprising’ is not a weasel
word with which to abrogate claim limitations.” Id. (internal quotation
omitted). The court reasoned that, although the word “comprising”
indicated “that an infringing process could practice other steps in addition to
the ones mentioned,” the steps of the method claim “must, however, all be
practiced as recited in the claim for a process to infringe.” Id. Importantly,
the court also reasoned that “[t]he presumption raised by the term
‘comprising’ does not reach into each of the [method] steps to render every
word and phrase therein open-ended.” Id.
Petitioner advances the same argument here based on the transition
word “comprising.” Following the reasoning of Dippin’ Dots, we reject
Petitioner’s proposed construction because it would abrogate the limitation
“incorporating a cost function at each iteration.” See id. In other words,
Petitioner is wrong to suggest that the presence of the word “comprising”
allows it to interpret the “incorporating” limitation as only requiring
incorporation of a cost function at some iterations of the method. See id.
Thus, we determine that the limitation “incorporating a cost function at each
iteration” requires incorporating a cost function for each and every cycle of
the recited method.

3. Other Terms
We determine that no other terms require explicit construction. See,
e.g., Nidec Motor Corp. v. Zhongshan Broad Ocean Motor Co., 868 F.3d
IPR2020-00071
Patent 6,393,096 B1
22
1013, 1017 (Fed. Cir. 2017) (“[W]e need only construe terms ‘that are in
controversy, and only to the extent necessary to resolve the controversy’
. . . .” (quoting Vivid Techs., Inc. v. Am. Sci. & Eng’g, Inc., 200 F.3d 795,
803 (Fed. Cir. 1999))).

D. Mr. Chase’s Testimony and Petitioner’s Motion to Exclude Portions
of Mr. Chase’s Declaration
Patent Owner puts forth a declaration of Mr. Chase as an expert
witness in the field of the ’096 patent. See Ex. 2037. Petitioner moves to
exclude paragraphs 104–110, 113–115, 128, and 129 of Mr. Chase’s
declaration (id.) under Federal Rule of Evidence 702 “because Mr. Chase is
unqualified to render opinions directed to subject matter beyond his limited
expertise in the use of commercially available treatment planning systems.”
Exclude Mot. 1. Petitioner further contends that Mr. Chase lacks
“specialized knowledge or experience on the design and development of
treatment planning systems and/or software as required by Fed. R. Evid.
702.” Id. at 2 (internal quotation omitted).
Petitioner argues that, “[i]n a patent case, a witness is only permitted
to testify as an expert on issues of noninfringement or invalidity if ‘qualified
as an expert in the pertinent art.’” Exclude Mot. 3 (quoting Sundance, Inc.
v. DeMonte Fabricating Ltd., 550 F.3d 1356, 1363 (Fed. Cir. 2008)).
According to Petitioner, a witness who does not qualify as an ordinarily
skilled artisan may not testify on obviousness, including whether there is a
motivation to combine references. Id. at 3–4 (citing Sundance, 550 F.3d at
1364). Petitioner purports to define the “relevant art” under Sundance as not
being “limited to a clinical application,” but as “extend[ing] more broadly to
treatment planning system design in the field of conformal radiation
IPR2020-00071
Patent 6,393,096 B1
23
therapy.” Id. at 4 (citing Ex. 1001, 1:10–12, 9:6, 11:54, 13:11–25). By this
standard, Petitioner contends that “Mr. Chase is unversed in this specific
subject matter of the ’096 patent.” Id. at 5. Petitioner also contends that we
should exclude Mr. Chase’s opinions “with respect to the design and
development of treatment planning systems and/or software” because he
“cannot implement the teachings of the ’096 patent or asserted prior art, and
he asserted that extraordinary skill is required to implement the teachings of
the ’096 patent.” Id. at 5–6 & n.5 (citing Ex. 1044, 33:12–20). Petitioner
further notes that Mr. Chase is not a person of ordinary skill under
Petitioner’s proposed level of ordinary skill in the art, though Petitioner
contends we need not resolve this issue in order to grant Petitioner’s motion.
Id. at 3 n.4.
Patent Owner contends that there is an “adequate relationship”
between Mr. Chase’s experience and the claimed inventions disclosed in the
’096 patent. Exclude Opp. 8 (quoting SEB S.A. v. Montgomery Ward & Co.,
594 F.3d 1360, 1373 (Fed. Cir. 2010)); see also id. at 7–8 (discussing
Mr. Chase’s experience as a medical physicist with “extensive clinical
experience including extensive work with linear accelerators (‘LINACs’),
treatment planning systems and their associated software,” among other
things). Patent Owner also argues that an expert witness’s expertise need
not be coextensive with the scope of the patent. Id. at 6 (citing Board cases).
As stated above, the ’096 patent relates to a method and apparatus for
conformal radiation therapy of tumors. Ex. 1001, 1:10–12. We agree with
Patent Owner (Exclude Opp. 7–8) that Mr. Chase’s background in nuclear
engineering and experience with treatment planning systems is adequately
related to the disclosed methods and apparatus for conformal radiation
IPR2020-00071
Patent 6,393,096 B1
24
therapy of tumors in the ’096 patent. See SEB, 594 F.3d at 1373; Ex. 2037
¶¶ 7–12. In addition, Mr. Chase is not required to have the exact experience
in our adopted definition of the level of ordinary skill, namely “computer
programming associated with treatment plan optimization.” See Patent Trial
and Appeal Board Consolidated Trial Practice Guide 34 (Nov. 2019)
(“Consolidated Trial Practice Guide”), available at https://www.uspto.gov/
sites/default/files/documents/tpgnov.pdf (“There is . . . no requirement of a
perfect match between the expert’s experience and the relevant field.”); see
also Mytee Prods., Inc. v. Harris Research, Inc., 439 F. App’x 882, 886–87
(Fed. Cir. 2011) (nonprecedential) (upholding admission of the testimony of
an expert who “had experience relevant to the field of the invention,” despite
admission that he was not a person of ordinary skill in the art). Thus,
Mr. Chase is qualified to provide expert testimony in this proceeding under
Federal Rule of Evidence 702.
We also note that the Board acts as both the gatekeeper and the
weigher of evidence. Similar to a district court in a bench trial, the Board,
sitting as a non-jury tribunal with administrative expertise, is well positioned
to determine and assign appropriate weight to evidence presented, including
giving it no weight. See, e.g., Donnelly Garment Co. v. NLRB, 123 F.2d
215, 224 (8th Cir. 1941) (“One who is capable of ruling accurately upon the
admissibility of evidence is equally capable of sifting it accurately after it
has been received . . . .”). For these reasons, we deny Petitioner’s motion to
exclude paragraphs 104–110, 113–115, 128, and 129 of Mr. Chase’s
declaration. Rather than excluding evidence that is allegedly confusing,
misleading, unsupported, and/or irrelevant, we will simply not rely on it or
give it little or no probative weight, as appropriate, in our analysis.
IPR2020-00071
Patent 6,393,096 B1
25
Despite our denial of Petitioner’s motion to exclude with respect to
Mr. Chase, we consider the facts underlying Petitioner’s motion to be
significant in our weighing of Mr. Chase’s testimony vis-à-vis the testimony
of Petitioner’s declarant Dr. Gall. According to Mr. Chase, a person of
ordinary skill in the art would have been a mere clinician who would not
have been able to write computer code or design a computer system that
performed the steps of claim 1. See Ex. 1045, 10:7–17; Ex. 2035 ¶ 67; see
also Ex. 1044, 59:24–60:3 (Mr. Chase testifying that a person of ordinary
skill in the art would not have been able to make the radiation planning
system that is described in the flowchart at Figure 2 of the ’096 patent).
Moreover, Mr. Chase testified that a person of ordinary skill in the art would
not have known how to write code that calculated a CDVH from a dose
distribution. Ex. 1044, 46:21–49:10. In fact, Mr. Chase testified that the
people making and developing the systems claimed and disclosed in the
’096 patent would have had extraordinary skill in the art. Id. at 33:12–20.
Mr. Chase likewise testified that the authors of Viggars exhibited
extraordinary skill in integrating a treatment planning system with a CDVH-
based evaluation program. Id. at 66:16–67:2. Yet it strains credulity to say
that an ordinarily artisan working in the art of the ’096 patent could not have
developed methods and systems like those disclosed in the ’096 patent itself.
Importantly, Mr. Chase acknowledged that he is not an expert in the
field of designing and developing treatment planning systems. Ex. 1044,
59:10–13, 67:18–68:5. Mr. Chase’s analysis also is distorted insofar as he
bases his opinions on the perspective of an artisan who had familiarity only
with commercial systems and not with “homemade systems,” i.e., those
being researched and developed in the art. See Ex. 1045, 21:25–24:4. This
IPR2020-00071
Patent 6,393,096 B1
26
perspective does not square with the notion that the ’096 patent allegedly
advanced the existing art. Ex. 1001, 3:53–4:9 (noting that “the art ha[d]
sought a method and apparatus for conformal radiation therapy . . . which
utilizes CDVH curves in establishing the desired dose distributions for each
target tumor volume and tissue and structure types” and that no such method
or apparatus had existed before the ’096 patent). In addition, the asserted
prior art is not necessarily limited to commercial methods and systems. See,
e.g., Ex. 1015, 420, 422–23 (Viggars describing scoring functions for
evaluating dose-volume histograms that were developed for the OSCAR
computer program, which was integrated with “conventional software”);
Ex. 1019, 249–50 (Oldham describing the development of cost functions for
use in a radiotherapy treatment plan optimization algorithm). This indicates
that an ordinarily skilled artisan would not have been confined to
commercialized systems as suggested by Mr. Chase. For these reasons, we
discount Mr. Chase’s testimony related to system development in our
obviousness analysis below.

E. Obviousness Ground Based on Oldham and Viggars
Petitioner contends the subject matter of claim 1 would have been
obvious over the combination of Oldham and Viggars. Pet. 28–47; Pet.
Reply 5–15. Patent Owner disputes Petitioner’s contentions. PO Resp. 21–
35; PO Sur-reply 8–12.

1. Oldham
Oldham is a paper directed to a radiotherapy treatment plan
optimization algorithm that uses a cost function to achieve a homogenous
IPR2020-00071
Patent 6,393,096 B1
27
dose for a planning target volume and to minimize the integral dose to
organs at risk. Ex. 1019, 248. The algorithm is based on fast simulated
annealing. Id. Beam weights are independently perturbed by adding a
“grain” of beam weight until the algorithm finds beam weight sets that
successively converge to the minimum of the cost function. Id. at 249.
Oldham’s cost function is segmented into component terms for
different regions: the target (PTV), organs-at-risk (OAR), and all other
tissue (BODY). Id. at 250. Equations (2)–(4) of Oldham are reproduced
below.

Equations (2)–(4) reflect the desired clinical dose to each region (CPTV,
COAR, and CBODY) where Di is the dose to the ith cubic voxel of each
segmented region. Id. These component terms are merged linearly into a
total cost function (Equation (5)), which is reproduced below.

For this total cost function CTOTAL in Equation (5), each term is weighted by
an “importance factor” (i.e., “WEIGHT”) to define its relative importance at
the start of the optimization. Id. The importance factors were implemented
by making “informed importance factor set ‘guesses,’” which were then
evaluated. Id. at 253. “Minimising the cost function CTOTAL thus
corresponds to minimising the integral dose in the OAR and BODY regions,
IPR2020-00071
Patent 6,393,096 B1
28
while attempting to achieve a uniform dose of 100% in the PTV.” Id.
at 250.
Petitioner contends Oldham qualifies as prior art under 35 U.S.C.
§ 102(b). Pet. 28–29 (citing, inter alia, Ex. 1003 ¶¶ 65–70). In support of
Oldham’s status as prior art, Petitioner includes testimony from Sylvia Hall-
Ellis, Ph.D., a professor with experience in the field of library science.
Ex. 1003 ¶¶ 6–8. She testifies that Oldham “was publicly accessible as early
as September 1, 1995, and in any event, more than one year before the
May 27, 1998 priority date,” based on a record of Oldham in the National
Library of Medicine. Id. ¶¶ 65–70. The journal in which Oldham appears
(Radiotherapy and Oncology: Journal of the European Society for
Therapeutic Radiology and Oncology) is dated June 1995, and it includes a
1995 copyright date. Ex. 1019, 1–3. The journal also includes stickers from
the National Library of Medicine including the date “09/01/95.” Id. at 1.
Dr. Gall additionally testifies that this journal is a “well-known and long-
standing scientific journal[] in the field of radiotherapy.” Ex. 1002 ¶ 73.
In our Decision on Institution, we credited Dr. Hall-Ellis’s testimony
regarding public accessibility and various indicators of publication
appearing on the front matter of the journal in which Oldham appears. Dec.
on Inst. 13. Thus, at that stage, we determined that Oldham qualified as
prior art under 35 U.S.C. § 102(b). Id. Since the time of institution, neither
party has put forth further arguments about the prior art status of Oldham, so
we perceive no reason to change our determination from that stage.
Accordingly, we determine that Oldham qualifies as prior art
under 35 U.S.C. § 102(b) because Oldham’s publication date in 1995 is
more than one year before the earliest possible effective filing date of the
IPR2020-00071
Patent 6,393,096 B1
29
challenged claims, which is May 27, 1998. Ex. 1001, code (60); Ex. 1003
¶¶ 65–70; Ex. 1019, 1–3.

2. Viggars
Viggars is a paper directed to the OSCAR computer program, which
“evaluates dose-volume histograms in a consistent way for use in
3-dimensional treatment planning.” Ex. 1015, 419. Viggars states that
“[d]ose volume histograms (DVH) are a convenient way of summarizing the
information in a 3-dimensional dose distribution.” Id. The aim of Viggars is
to use DVHs to compare and evaluate alternative plans objectively and
consistently such that DVHs may be used in defining and ensuring
adherence to a treatment protocol. Id.
According to Viggars, the quality of a proposed treatment plan may be
judged by how far its cumulative dose volume histogram (CDVH) departs
from the ideal histograms, and a dose prescription can be defined by
specifying the maximum acceptable deviations from the ideal shape.
Ex. 1015, 420. Such deviations are referred to as “regret.” Id. A set of
score functions may be used to compare the actual deviations of a plan from
the ideal CDVH with the maximum deviations allowed by the dose
prescription. Id. at 422.
“For each dose volume limit [Di,Ri(max)] in the prescription, the
score function is derived from a ratio ri,” which is defined in the equation
reproduced below. Ex. 1015, 422.
ri = Ri(Di) Ri(max)⁄
This ratio ri is then used in a score function Si, which is reproduced below.
Id. at 423.
IPR2020-00071
Patent 6,393,096 B1
30
Si = 10[1− ri]
This score function Si results in “10 for an ideal distribution, zero at the limit
of acceptability, and [a] negative [value] when the dose-volume limit is
violated.” Id. An optimal plan could, in principle, be selected by assigning
weights to each score to derive an overall objective function. Id. at 425.
Petitioner contends Viggars qualifies as prior art under 35 U.S.C.
§ 102(b). Pet. 28–29 (citing, inter alia, Ex. 1003 ¶¶ 54–59). In support of
Viggars’s status as prior art, Dr. Hall-Ellis testifies that Viggars “was
publicly accessible as early as June 10, 1992, and in any event, more than
one year before the May 27, 1998 priority date,” based on a record of
Viggars in the University of California San Diego. Ex. 1003 ¶¶ 54–59. The
journal in which Viggars appears (International Journal of Radiation
Oncology • Biology • Physics) is dated 1992, and it includes a 1992 copyright
date. Ex. 1015, 1–3. The journal also includes a sticker from the University
of California San Diego that states “Received on: 06-10-92.” Id. at 1.
Dr. Gall additionally testifies that this journal is a “well-known and long-
standing scientific journal[] in the field of radiotherapy.” Ex. 1002 ¶ 73.
In our Decision on Institution, we credited Dr. Hall-Ellis’s testimony
regarding public accessibility and various indicators of publication
appearing on the front matter of the journal in which Viggars appears. Dec.
on Inst. 15. Thus, at that stage, we determined that Viggars qualified as
prior art under 35 U.S.C. § 102(b). Id. Since the time of institution, neither
party has put forth further arguments about the prior art status of Viggars, so
we perceive no reason to change our determination from that stage.
Accordingly, we determine that Viggars qualifies as prior art under
35 U.S.C. § 102(b) because Viggars’s publication date in 1992 is more than
IPR2020-00071
Patent 6,393,096 B1
31
one year before the earliest possible effective filing date of the challenged
claims, which is May 27, 1998. Ex. 1001, code (60); Ex. 1003 ¶¶ 54–59;
Ex. 1015, 1–3.

3. Claim 1
a. Preamble and Claim Limitations
The preamble of claim 1 recites “[a] method of determining an
optimized radiation beam arrangement for applying radiation to a tumor
target volume while minimizing radiation of a structure volume in a patient,
comprising the steps of[.]” Ex. 1001, 16:39–42. Petitioner cites Oldham’s
simulated annealing optimization method and its teaching of a cost function
used to find beam weights that achieve a homogenous dose in the target
volume while minimizing the dose to organs at risk. Pet. 38 (citing
Ex. 1019, 248–49). Petitioner also cites Viggars’s CDVH-based cost
function for evaluating DVHs, which includes “overdose and underdose
limits for the radiation applied to the target, as well as dose-volume limits on
the radiation received by the organs-at-risk and non-target tissue.” Id. at 39
(citing Ex. 1015, 420–21).
Patent Owner does not dispute Petitioner’s analysis of the preamble of
claim 1. Neither party addresses whether the preamble is limiting. Because
Petitioner has shown that the combination of Oldham and Viggars teaches
the preamble, we need not determine whether the preamble is limiting. See
Nidec, 868 F.3d at 1017.
Claim 1 further recites “using a computer to computationally obtain a
proposed radiation beam arrangement” (“first ‘using’ limitation”). Ex. 1001,
16:43–44. Petitioner cites Oldham’s disclosure of the COVIRAOPT
IPR2020-00071
Patent 6,393,096 B1
32
computer program, which uses a fast simulated annealing algorithm that
converges on a beam-weight set that corresponds to a minimum of the cost
function. Pet. 39–40 (citing Ex. 1019, 249, 261). Patent Owner does not
dispute Petitioner’s analysis of the first “using” limitation. We are
persuaded that Oldham’s use of a computer-implemented fast simulated
annealing algorithm in COVIRAOPT teaches the first “using” limitation.
See, e.g., Ex. 1019, 249, 261.
Claim 1 further recites “using a computer to computationally change
the proposed radiation beam arrangement iteratively” (“second ‘using’
limitation”). Ex. 1001, 16:45–46. Petitioner again cites Oldham’s fast
simulated annealing algorithm and Oldham’s teaching that the algorithm is
iterative, wherein “at each iteration all beam-weights are independently
perturbed by adding a ‘grain’ of beam-weight.” Pet. 40 (quoting Ex. 1019,
249). Patent Owner does not dispute Petitioner’s analysis of the second
“using” limitation. We are persuaded that Oldham’s perturbation of beam-
weights teaches the second “using” limitation. See, e.g., Ex. 1019, 249.
Claim 1 further recites “incorporating a cost function at each iteration
to approach correspondence of a CDVH associated with the proposed
radiation beam arrangement to a CDVH associated with a predetermined
desired dose prescription.” Ex. 1001, 16:47–50. Petitioner cites Oldham’s
algorithm, which iteratively perturbs beam weights and then evaluates a cost
function. Id. at 41–42 (citing Ex. 1019, 249). According to Petitioner, the
algorithm successively converges to a minimum value of the cost function.
Id. at 42 (citing Ex. 1019, 249). Petitioner also cites Oldham’s “total cost
function that is segmented into component terms for each of the target
(PTV), organs-at-risk (OAR), and surrounding tissue (BODY).” Id. (citing
IPR2020-00071
Patent 6,393,096 B1
33
Ex. 1019, 250). Petitioner notes the component terms are weighted by
“importance factor” and merged linearly to form the total cost function. Id.
Petitioner contends “[i]t would have been obvious to a[n ordinarily
skilled artisan] to incorporate the segmented score functions of Viggars for
the target, organs-at-risk, and non-target tissue into an overall cost function
that replicates the merged and weighted total cost function of Oldham.”13
Pet. 42–43 (citing Ex. 1002 ¶ 107). In particular, Petitioner contends
Viggars’s “segmented score functions merged into the overall weighted cost
function compare the actual deviations of a plan from the ideal CDVH with
the maximum deviations allowed by the dose prescription.” Id. at 43 (citing
Ex. 1002 ¶ 108; Ex. 1015, 422–23) (internal quotation omitted).
Petitioner explains its proposed combination as follows. Petitioner
starts with equation (5) of Oldham, which is reproduced below.

Equation (5) of Oldham is a “total cost function” CTOTAL, which linearly
merges segmented cost functions for the target (CPTV), organs-at-risk (COAR),
and surrounding tissue (CBODY). Ex. 1019, 250. Each of the segmented cost
functions is “weighted by an ‘importance factor’ to define its relative
importance at the start of the optimisation.” Id. The subscript “ST” denotes
the starting value of a term. Id.

13 We further discuss the parties’ arguments regarding Petitioner’s rationale
for combining Viggars with Oldham below. See infra II.E.3.b.
IPR2020-00071
Patent 6,393,096 B1
34
Petitioner proposes replacing Oldham’s segmented cost functions with
Viggars’s score function Si, which “compare[s] the actual deviations of a
plan from the ideal CDVH with the maximum deviations allowed by the
dose prescription.” Pet. 42–43 (citing, inter alia, Ex. 1015, 422). In
particular, Petitioner cites Viggars’s score function Si, which is reproduced
below.
Si = 10[1− ri]
This score function Si results in “10 for an ideal distribution, zero at the limit
of acceptability, and [a] negative [value] when the dose-volume limit is
violated.” Ex. 1015, 423. Petitioner notes the value ri “is the measure of the
plan’s deviation from the ideal dose prescription CDVH.” Pet. 43 (citing
Ex. 1002 ¶ 109); see also Ex. 1015, 422 (defining ri).
Petitioner proposes combining these two teachings into the following
combined cost function CTOTAL.

In Petitioner’s proposed equation reproduced above, the combined cost
function CTOTAL represents “the overall cost function of Viggars to determine
an optimal treatment plan using Oldham’s segmented-cost method.” Pet. 44
(citing Ex. 1002 ¶ 113). Petitioner notes the sign has been changed in the
Starget, SOAR, and SBODY terms to achieve minimization by Oldham’s fast
simulated annealing algorithm. Id. at 43–44 (citing Ex. 1002 ¶ 111).
Petitioner contends such a change would have been trivial and readily
IPR2020-00071
Patent 6,393,096 B1
35
apparent to a person of ordinary skill in the art. Id. Petitioner further
contends that
[d]etermining the three suitable weighting factors to achieve a
clinical objective could easily be arrived at using the
straightforward trial-and-error approach taught by Oldham and
guided by the clinician’s judgment concerning the relative
importance of applying the appropriate dose to the target versus
the dose tolerated by the organ-at-risk or body tissue.
Id. at 44–45 (citing Ex. 1002 ¶ 114).
Patent Owner criticizes Petitioner’s analysis of the “incorporating”
step as being “premised on the fallacy that Viggars discloses an ‘overall
CDVH-based cost function.’” PO Resp. 27. Rather, Patent Owner calls
Viggars’s OSCAR scoring system “a standalone scoring program” that
“does not teach any use of, or how to use, these scores in a cost function.”
PO Sur-reply 8. Patent Owner also characterizes Viggars as stating that
“(1) the authors did not attempt to derive an overall cost function, (2) it was
unclear how to do that, and (3) it was preferable not to do that.” PO
Resp. 27 (citing Ex. 1015, 425–26; Ex. 2037 ¶¶ 100–101); see also PO
Sur-reply 8 (same argument). In particular, Patent Owner highlights
Viggars’s statement on assigning weights for a single figure of merit that
“[i]t is not clear how such weighting should be carried out, and it seems
preferable at present to retain the separate scores.” PO Resp. 27 (quoting
Ex. 1015, 425–26) (emphasis omitted). Patent Owner argues that “Viggars’
suggestion that one could, in principle, assign weights to each score to
derive an overall objective function was negated by the remaining discussion
in Viggars.” Id. at 28.
We do not agree with Patent Owner’s arguments. Viggars expressly
states that its CDVH-based score functions could, in principle, be weighted
IPR2020-00071
Patent 6,393,096 B1
36
“to derive an overall objective function.” Ex. 1015, 425 (emphasis added).
In light of this, Petitioner’s characterization of Viggars teaching “an overall
CDVH-based cost function” is apt. Indeed, Viggars teaches judging “how
far [a proposed plan’s] CDVH departs from the ideal histograms.” Id. at
420. We also do not agree with Patent Owner’s argument (PO Resp. 28)
that Viggars’s teachings “negate” the idea that an ordinarily skilled artisan
would have assigned weights as part of a cost function. Although the
authors of Viggars did not themselves create a single figure of merit
(Ex. 1015, 426), they certainly suggested it could be done. Id. at 425 (“[A]n
optimal plan could, in principle, be selected by assigning weights to each
score to derive an overall objective function.”).
Importantly, we must consider Viggars “for everything it teaches by
way of technology,” which is “not limited to the particular invention it is
describing.” EWP Corp. v. Reliance Universal Inc., 755 F.2d 898, 907 (Fed.
Cir. 1985). This includes Viggars’s express teaching that one could
“assign[] weights to each score to derive an overall objective function.”
Ex. 1015, 425. We also note record evidence that, in general, weighting is
patient-dependent, and no universal set of weights applies across all
treatment plans. See Ex. 1015, 426; Ex. 1043 ¶¶ 24–25, 27; Ex. 1045,
87:23–89:19. We further note that Petitioner additionally relies on
Oldham’s express teachings of assigning weights via a trial-and-error
approach. See Pet. 36–37 (citing Ex. 1002 ¶ 92; Ex. 1019, 253); Pet.
Reply 8–9 (citing Ex. 1043 ¶ 24). All of these teachings support Dr. Gall’s
testimony “that it would have been a routine and straightforward exercise to
determine the relative values for the weights of the overall CDVH-based
cost function” (Ex. 1043 ¶ 24), which we credit.
IPR2020-00071
Patent 6,393,096 B1
37
Therefore, we are persuaded by Petitioner’s analysis of the
“incorporating” step. Petitioner relies on a modified version of Oldham’s
combined cost function CTOTAL that incorporates Viggars’s teaching of a
score function that compares actual deviations of a plan from the ideal
CDVH. See Pet. 42–44. As such, Petitioner establishes that its proposed
version of Oldham’s combined cost function CTOTAL—as modified to
incorporate Viggars’s CDVH score functions—teaches the “incorporating”
limitation. See Pet. 44.
Claim 1 further recites “comparing the dose distribution to a
prescribed dose for the tumor volume and surrounding tissue structures.”
Ex. 1001, 16:51–52. Petitioner cites Viggars’s teaching of objective cost
functions, “which quantify the deviation of the dose distribution from the
dose prescription,” and notes this functionality is included in Petitioner’s
proposed combined cost function (discussed above). Pet. 45 (citing
Ex. 1002 ¶ 116; Ex. 1015, 420). Petitioner also contends its combined cost
function “‘provide[s] a quantitative measure of how well a proposed
treatment plan conforms to the dose prescription,’ and thus compares the
dose distribution to a prescribed dose for the tumor volume and surrounding
tissue structures.” Id. (quoting Ex. 1015, 422). Petitioner additionally cites
Viggars’s teachings of visual displays for CDVHs, dose limits, histograms
of regret, isodose charts, and images of regret. Id. at 45–46 (citing Ex. 1015,
419–20, 422). Patent Owner does not dispute Petitioner’s analysis of the
“comparing” limitation. We are persuaded that Viggars’s CDVH-based
score functions to quantify conformity—as implemented in Petitioner’s
proposed overall cost function based on Oldham and Viggars—teach the
“comparing” limitation. See, e.g., Ex. 1015, 422–23.
IPR2020-00071
Patent 6,393,096 B1
38
Claim 1 further recites “increasing or decreasing radiation beam
intensity if the change of the proposed beam arrangement leads to a greater
correspondence to the desired dose prescription to obtain an optimized
radiation beam arrangement.” Ex. 1001, 16:54–57. Petitioner cites
Oldham’s teachings of iteratively adding (or subtracting) “a ‘grain’ of beam-
weight” to determine the effect on a cost function. Pet. 46–47 (citing
Ex. 1002 ¶ 117; Ex. 1019, 249). According to Petitioner, this process
successively converges on a minimum of the cost function, which is
associated with correspondence to a desired dose prescription. Id. Patent
Owner does not dispute Petitioner’s analysis of the “increasing or
decreasing” limitation. We are persuaded that Oldham’s iterative method of
adding or subtracting “grain[s] of beam-weight” in an effort to minimize a
cost function teaches the “increasing or decreasing” limitation. See, e.g.,
Ex. 1019, 249.
Based on the entire trial record, Petitioner has established that the
combination of Oldham and Viggars teaches all limitations of claim 1.

b. Reasons for the Combination
As part of its rationale for the combination, Petitioner notes Viggars’s
teaching of “an overall CDVH-based cost function to determine an optimal
treatment plan.” Pet. 33 (citing Ex. 1015, 425). Petitioner also notes
Oldham’s teaching of “how to determine suitable weights for the individual
costs associated with an overall cost function.” Id. (citing Ex. 1019, 253).
In light of these teachings, Petitioner contends an ordinarily skilled artisan
“would have been motivated by Oldham to construct and incorporate the
overall cost function disclosed in Viggars within Oldham’s optimization
IPR2020-00071
Patent 6,393,096 B1
39
algorithm.” Id. (citing Ex. 1002 ¶ 85); see also id. at 35 (citing Viggars
(Ex. 1015, 426) for teaching the ease of interpreting “a single figure of merit
for a treatment plan”). This would have resulted in using “the cost function
expressly disclosed in Viggars in order to perform computer-implemented
optimization of a treatment plan of Oldham to implement the same CDVH-
based evaluations of proposed treatment plans that were already being
performed by the physician.” Id. at 33–34 (citing Ex. 1002 ¶ 87). Petitioner
cites the advantage of “being able to effectively and efficiently screen a vast
set of different beam configurations with the SARP algorithm of Oldham to
arrive at a more optimal treatment configuration.” Id. at 34 (citing Ex. 1002
¶ 88). As another advantage, Petitioner cites the ability to “account[] for
dose-volume limits associated with partial volumes identified in the
physician’s dose prescription,” which Oldham’s cost function alone cannot
do. Id. at 34–35 (citing Ex. 1002 ¶ 88).
Petitioner contends it would have been “obvious to try the overall
CDVH-based cost function suggested by Viggars with Oldham’s SARP
algorithm in order to determine an ‘optimal plan’” based on “Oldham’s
teaching of how to assign suitable weights to the individual components.”
Pet. 36 (citing Ex. 1002 ¶ 92). In particular, Petitioner notes that SARP
optimization methods were known to “permit[] the straightforward
utilization of any objective function,” which leads to expected success. Id.
(quoting Ex. 1014, 135) (citing Ex. 1002 ¶ 90). Petitioner also cites
Oldham’s teaching of weighting the target, organ, and tissue based on an
“importance factor,” wherein Oldham teaches making “informed importance
factor set ‘guesses’” and then evaluating them. Id. at 36–37 (citing Ex. 1002
¶ 92; Ex. 1019, 253). Thus, in light of Oldham’s teachings, Petitioner
IPR2020-00071
Patent 6,393,096 B1
40
contends that an ordinarily skilled artisan would have reasonably expected
success in identifying the weights to be applied to the target, organs, and
tissue. Id. at 36–37 (citing Ex. 1002 ¶¶ 91–92, 94).
Patent Owner argues that Viggars expressly states a preference against
incorporating its score functions into an overall cost function, and instead
suggests the scores should “be kept separate and used by clinicians.” PO
Resp. 28 (citing Ex. 1037 ¶ 106). As discussed above with respect to the
“incorporating” step of claim 1, however, Viggars states expressly that “an
optimal plan could, in principle, be selected by assigning weights to each
score to derive an overall objective function.” Ex. 1015, 425. Viggars also
states that a clinician could make a final decision on the relative importance
of the scores “in accordance with the needs of individual patients.” Id. at
426. This is consistent with the idea that the assignment of weights to
targets and structures is unique to each patient. Ex. 1043 ¶¶ 24–25, 27;
Ex. 1045, 87:23–89:19. We are persuaded that an ordinarily skilled artisan
considering these passages from Viggars would have known how to assign
weights during the trial-and-error inverse planning process based on
Oldham’s teachings. See, e.g., Ex. 1002 ¶¶ 85, 91–94; Ex. 1043 ¶¶ 18–20.
Therefore, we do not agree that Viggars’s other statements about its authors
not implementing such an overall cost function—or even preferring “at
present to retain the separate scores” (id. at 426 (emphasis added))—would
have discouraged an ordinarily skilled artisan from making Petitioner’s
proposed combination.
Patent Owner argues that an ordinarily skilled artisan “would not have
been motivated to complicate Oldham’s relatively simple cost function with
the more complex score functions of Viggars.” PO Resp. 29 (citing
IPR2020-00071
Patent 6,393,096 B1
41
Ex. 2037 ¶ 103). Patent Owner also argues that “[i]t would not have been
obvious that the more complicated score functions of Viggars would yield
more optimal results than Oldham’s Equations.” Id.
Contrary to Patent Owner’s arguments, Petitioner does provide a
reason why an ordinarily skilled artisan would have added Viggars’s more
complex CDVH scoring to Oldham’s cost function CTOTAL: to create a
single figure of merit for a treatment plan that incorporates Viggars’s
accounting for dose-volume limits associated with partial volumes identified
in the physician’s dose prescription. See Pet. 34–35 (citing Ex. 1002 ¶¶ 88–
89). Petitioner’s rationale is persuasive given that Oldham acknowledged a
shortcoming of its cost function as failing “to model complicated volume
effects.” Ex. 1019, 250. In addition, Oldham already taught “plan
evaluation tool[s]” such as DVHs (see PO Resp. 31 (citing Ex. 1019, 254–
56)), albeit in a way that was not quantified in a single cost function.
Correspondingly, Viggars touts the ability of the OSCAR system to “select[]
and improv[e] a treatment plan . . . without the ongoing intervention of a
radiation oncologist.” Ex. 1015, 425. We agree with Petitioner that this
would have provided motivation to automate Viggars’s CDVH-based
evaluation of beam configurations. Pet. 34 (citing Ex. 1002 ¶ 88). As such,
Petitioner’s combination remedies Oldham’s acknowledged shortcoming
and systematizes the use of DVHs as an evaluation tool.
Patent Owner questions whether the “score functions of Viggars
would yield more optimal results than Oldham’s Equations.” PO Resp. 29
(citing Ex. 2037 ¶ 103). Mr. Chase offers supporting testimony (Ex. 2037
¶ 103) in which he purports to compare the effectiveness of Petitioner’s
combination with that of Oldham alone. But these arguments miss the point
IPR2020-00071
Patent 6,393,096 B1
42
of the combination, which is to automate the optimization process by
creating a single figure of merit that accounts for dose-volume limits. See
Pet. 34–35. Petitioner establishes that its proposed combination would have
been a desirable improvement regardless of whether a clinician might still
use Oldham’s scoring functions alone while effectively optimizing a given
treatment plan. Id.; Ex. 1002 ¶¶ 88–89.
Patent Owner argues that this modification would have come at a
significant cost, namely, increased computation time. PO Resp. 30–31
(citing, inter alia, Ex. 2037 ¶¶ 104–105). Patent Owner contends the
combination would eliminate the time savings of Oldham, which is allegedly
based on a “computational short cut” whereby “voxel summation need only
be done once, for each beam, at the start of the optimisation.” Id. (quoting
Ex. 1019, 250) (citing Ex. 2037 ¶ 104). In reply, Petitioner argues any
additional computational time associated with Petitioner’s combination
“would not have outweighed or discouraged a POSA from pursuing the
recognized benefits of the combination.” Pet. Reply 12 (citing Ex. 1043
¶ 30). We agree with Petitioner that additional computational time would
not have discouraged the combination. See Allied Erecting & Dismantling
Co. v. Genesis Attachments, LLC, 825 F.3d 1373, 1381 (Fed. Cir. 2016)
(“[A] given course of action often has simultaneous advantages and
disadvantages, and this does not necessarily obviate motivation to combine.”
(internal quotation omitted)). As noted by Petitioner, Petitioner’s proposed
combination of Oldham and Viggars obtains benefits that are not possible
using Oldham’s cost function alone, namely, modeling dose volume effects
and systematizing the use of CDVHs as a plan evaluation tool. PO Resp. 12
(citing Ex. 1043 ¶ 30). In this way, Patent Owner’s attempt to compare the
IPR2020-00071
Patent 6,393,096 B1
43
processing speed of Oldham alone with that of Petitioner’s combination is
not apt. Thus, regardless of increased processing time, we are persuaded
that an ordinarily skilled artisan would have been motivated to obtain the
benefits of Petitioner’s combination. See Ex. 1002 ¶¶ 88–89; Ex. 1043 ¶ 30.
Furthermore, the record contains evidence that the computational time of
Petitioner’s proposed combination would have been commensurate with
typical inverse treatment planning optimization programs from the time of
the invention. Ex. 1006, 60; Ex. 1019, 248, 250; Ex. 1043 ¶ 29; Ex. 1045,
91:21–93:1; Ex. 2037 ¶ 90.
The fact that Oldham already mentions dose value histograms as a
plan evaluation tool (see PO Resp. 31 (citing Ex. 1019, 254–56; Ex. 2037
¶ 111)) likewise does not undermine the combination. Oldham’s teachings
do not purport to test fidelity to a dose value histogram in an automated way,
as is proposed by Petitioner. See Ex. 1019, 254–56.
Patent Owner additionally argues that it would not have been obvious
how to incorporate the score functions of Viggars into the objective function
of Oldham and that an ordinarily skilled artisan would not have had a
reasonable expectation of doing it successfully. PO Resp. 32–35. Patent
Owner notes that Viggars does not teach how to assign weights to its score
functions and that the authors stated that they preferred not to do so. Id. at
32 (citing Ex. 1015, 425–26; Ex. 2037 ¶ 106). Patent Owner further
criticizes Petitioner’s reliance on Oldham for teaching “guesses” and a trial-
and-error approach to assigning weights because these teachings pertain to
the simpler objective function of Oldham. Id. at 32–33 (citing Ex. 2037
¶ 109). In particular, Patent Owner argues that such an approach does not
establish a reasonable expectation of success in making the combination. Id.
IPR2020-00071
Patent 6,393,096 B1
44
at 33 (citing Ex. 2037 ¶¶ 108–109). Patent Owner also cites a portion of an
article (Ex. 1008, “Webb-1993”) by a co-author of Oldham for the
proposition that “[i]t may not however be entirely clear how to clinically set
the dose limits or how to interpret the regret if dose limits cannot be met.”
Id. (quoting Ex. 1008, 20–21). Patent Owner argues that “[t]his suggests
that a POSA would not have had a reasonable expectation of success in
using Viggars’ regret scores to drive a planning method.” Id. (citing
Ex. 2037 ¶ 110).
We do not agree with Patent Owner that Petitioner has failed to
establish a reasonable expectation of success in making Petitioner’s
proposed combination. Record evidence, including cross-examination
testimony from Mr. Chase, reflects that Petitioner’s proposed trial-and-error
process for determining weights was routine in the art. Ex. 1002 ¶¶ 85, 91–
94; Ex. 1019, 253; Ex. 1043 ¶¶ 18, 31–32, Ex. 1045, 19:15–22:11.
Petitioner also establishes that there is an intuitive correspondence between
the weights assigned to components in a cost function, i.e., target, organ, and
tissue, and the resulting dose distribution. Pet. 36 (citing Ex. 1029, 887). In
addition, Dr. Gall testifies that Viggars’s score functions had a “convenient
numerical form and linear nature” that makes them simple to weight.
Ex. 1002 ¶ 91; Ex. 1043 ¶ 32. Dr. Gall explains that “it would have been a
trivial matter to have given greater relative weight to a structure when the
target CDVH is clinically acceptable but a structure CDVH is not, and vice
versa, as part of a trial-and-error process towards a clinically acceptable
plan.” Ex. 1043 ¶ 32. We credit Dr. Gall’s testimony over Mr. Chase’s
testimony for the reasons discussed above. See supra § II.D. All of these
factors support Petitioner’s argument that
IPR2020-00071
Patent 6,393,096 B1
45
[a] POSA reading Viggars would have understood, based on the
express teachings of Oldham and the understanding by
clinicians regarding how to assign weights during the trial-and-
error inverse planning process, that it would have been a routine
and straightforward exercise to determine the relative values for
the weights of the overall CDVH-based cost function.
Pet. Reply 8–9 (citing Ex. 1043 ¶ 24); see also Pet. 44–45 (citing Ex. 1002
¶ 114).
We also have considered Patent Owner’s arguments (PO Resp. 33)
based on the quote from Webb-1993 that “[i]t may not however be entirely
clear how to clinically set the dose limits or how to interpret the regret if
dose limits cannot be met.” Ex. 1008, 21. We agree with Petitioner that this
statement, which concerns Viggars’s score functions, only applies if dose
limits cannot be met by a given plan. Pet. Reply 14 (citing Ex. 1043 ¶ 43).
As noted by Petitioner, its proposed combination “screens thousands of
beam configurations to find an optimal plan that would, in fact, satisfy all of
the dose-volume limits of the prescription.” Id. Petitioner additionally notes
that a sentence in the same paragraph of Webb-1993 extols a situation where
Viggars’s score function leads to a positive outcome. Id. (quoting Ex. 1008,
21) (“Provided the treatment planner derives a scheme with positive score
function, the plan can be considered acceptable . . . the prescription having
been met.”). Given this context, we do not agree with Patent Owner (PO
Resp. 33) that Webb-1993’s commentary “contradicts Petitioner’s
obviousness argument.”
Patent Owner also argues that an ordinarily skilled artisan “would not
have reasonably expected success in combining the forward planning tool of
Viggars with the inverse planning tool of Oldham.” PO Resp. 34.
According to Patent Owner, “[f]orward planning and inverse planning were
IPR2020-00071
Patent 6,393,096 B1
46
fundamentally different techniques with their own processes and
terminologies, that did not have common terms of evaluation.” Id. (citing
Ex. 2037 ¶¶ 113–114). Nevertheless, Patent Owner’s counsel acknowledged
at the oral hearing that Viggars’s CDVH-based cost function is an evaluation
tool that can be applied to inverse planning treatment plans. See Tr. 84:18–
86:7. And simulated annealing optimization methods (such as Oldham’s)
“permit[] the straightforward utilization of any objective function.” Pet. 36
(quoting Ex. 1013, 135) (citing Ex. 1002 ¶ 90) (emphasis added). Thus, we
do not agree with Patent Owner’s arguments.
Having considered the entire trial record, we are persuaded by
Petitioner’s showing that an ordinarily skilled artisan would have had
reasons to combine Oldham and Viggars.

c. Conclusion Regarding Claim 1
Petitioner has persuasively shown that the combination of Oldham
and Viggars teaches all the limitations of claim 1. Petitioner also has put
forth persuasive reasons for combining these references and has established
that an ordinarily skilled artisan would have had a reasonable expectation of
success in making the combination. On the entire trial record, we determine
Petitioner has shown, by a preponderance of the evidence, that the subject
matter of claim 1 would have been obvious over the combination of Oldham
and Viggars.

F. Obviousness Ground Based on Oldham, Viggars, and Morrill-1991
Petitioner contends the subject matter of claim 18 would have been
obvious over the combination of Oldham, Viggars, and Morrill-1991.
IPR2020-00071
Patent 6,393,096 B1
47
Pet. 47–55; Pet. Reply 15–23. Patent Owner disputes Petitioner’s
contentions. PO Resp. 35–42; PO Sur-reply 6–7, 12–14.

1. Morrill-1991
Morrill-1991 is a paper directed to “[a] variation of simulated
annealing optimization called ‘constrained simulated annealing’ [that] is
used with a simple annealing schedule to optimize beam weights and angles
in radiation therapy treatment planning.” Ex. 1022, 1341. According to
Morrill-1991, “[t]he use of dose-volume information effectively removes the
dependence of the optimized solution upon the position of any single dose
constraint point.” Id. at 1344. Morrill-1991 describes an objective function
called “maximize dose with dose-volume limits” (MDVL) that “maximizes
the dose to isocentre, subject to target volume dose heterogeneity limits as
well as maximum dose and dose-volume limits on the normal organs.” Id.
at 1345.
Table 2 of Morrill-1991 is reproduced below.

IPR2020-00071
Patent 6,393,096 B1
48
Table 2 depicts “[d]ose-volume constraints for the normal organs used by
the MDVL objective function in the optimization of a treatment plan for a
pancreatic tumour.” Id. at 1347.

2. Claim 18
Claim 18 depends from claim 1, 2, or 14 and further recites “the step
of allowing a radiation limit on the tissue structure to be exceeded by a set
amount if such excess allows better conformation to the desired target
CDVH curve.” Ex. 1001, 18:23–26. The analysis that follows addresses
only the direct dependency of claim 18 from claim 1, which reflects the
broadest scope of claim 18. Building on its obviousness contentions for
claim 1 in the Oldham–Viggars ground, Petitioner proposes applying
Morrill-1991’s constrained simulated annealing approach within Oldham’s
fast simulated annealing algorithm and Viggars’s CDVH-based cost
function. Pet. 51 (citing Ex. 1002 ¶ 127; Ex. 1022, 1343). Addressing the
“radiation limit” language of claim 18 in particular, Petitioner cites “[t]he
maximum dose constraints placed on the treatment plan within the
constrained variation of simulated annealing taught by Morrill-1991.” Id. at
52 (citing Ex. 1002 ¶ 130). Petitioner provides an example keyed to the
“Spinal cord” entry in Table 2 of Morrill-1991 (reproduced above). See
Ex. 1022, 1347. According to Petitioner, this table “indicates that the spinal
cord has a maximum dose of 45 Gy, a maximum volume of 50%, and a
volume dose of 40 Gy,” which “means that the dose to the spinal cord is
subject to a radiation limit of 40 Gy, but up to 50% of the spinal cord
volume can exceed this radiation limit by a set amount of 5 Gy to the
maximum dose of 45 Gy.” Pet. 53 (citing Ex. 1002 ¶ 133). Petitioner
IPR2020-00071
Patent 6,393,096 B1
49
explains that allowing the radiation limit on a normal organ to be exceeded
can result in better conformation to the target CDVH, albeit at the expense
of excess radiation to the normal organ up to the maximum dose constraint.
Id. at 54 (citing Ex. 1002 ¶ 133).
Petitioner contends an ordinarily skilled artisan would have been
motivated to combine the dose-volume constraints of Morrill-1991’s
constrained simulated annealing method with the optimization algorithm of
Oldham and the cost function of Viggars. Pet. 50 (citing Ex. 1002 ¶ 125).
Petitioner notes that Oldham and Morrill-1991 both use simulated annealing
algorithm and cost functions for the optimization of beam arrangement for
conformal radiotherapy. Id. Petitioner cites several advantages to
Morrill-1991’s constrained simulated annealing method, including “the
flexibility to specify individual organ dose-volume limits (especially to the
spinal cord and kidneys) which implement the clinician’s personal treatment
methodology.” Id. (quoting Ex. 1022, 1354). Petitioner notes that using
constraints rather than weighted penalties in a cost function “is
computationally more efficient.” Id. (quoting Ex. 1022, 1343) (citing
Ex. 1002 ¶ 126). Petitioner also cites the ability “to place strict limits on
maximum doses not to be violated by a treatment plan in accordance with a
dose prescription.” Id. at 49–50 (citing Ex. 1002 ¶ 126). Petitioner
additionally contends an ordinarily skilled artisan would have expected
success in making the combination because introducing Morrill-1991’s
constraints “would merely require a step within [Oldham’s] SARP algorithm
that checks whether the constraints are satisfied with every sample beam
arrangement configuration at each iteration of the simulated annealing
algorithm.” Id. at 51 (citing Ex. 1002 ¶ 128; Ex. 1013, 139; Ex. 1022,
IPR2020-00071
Patent 6,393,096 B1
50
1343). Petitioner also cites Morrill-1991’s own teaching that “[d]ifferent
objective functions, constraints and annealing schedules are straightforward
to implement.” Id. (quoting Ex. 1022, 1358).
Patent Owner argues that Morrill-1991’s constrained simulated
annealing method rejects sample configurations outright if they fail to satisfy
the dose-volume constraints. See PO Resp. 39–41 (citing, inter alia,
Ex. 1022, 1343; Ex. 2037 ¶¶ 132, 134); see also PO Sur-reply 6–7 (same
argument). According to Patent Owner, “the steps of evaluating the cost
function and accepting or rejecting the proposed beam weights are entirely
skipped” in such a scenario. PO Resp. 41 (citing Ex. 2037 ¶ 135). In our
Decision on Institution, we found that this teaching conflicted with the
method of claim 18, which—via its dependency from claim 1—seeks
conformance to a desired CDVH curve by evaluating a cost function at each
iteration. Dec. on Inst. 28.
In reply, Petitioner argues that its obviousness arguments are based on
“iterations where the constraints are not violated and the Viggars’s overall
CDVH-based cost function is evaluated.” Pet. Reply 16. According to
Petitioner, “[t]here would be numerous iterations, including many sequential
iterations, in which the constraints would be met, a cost function
incorporated, and an optimal plan determine[d].” Id. at 17–18 (citing
Ex. 1043 ¶¶ 39–40). Petitioner argues that these iterations are sufficient to
establish obviousness in light of the open-ended transition “comprising” in
claim 1. Id. at 16, 18. Petitioner also criticizes Patent Owner’s arguments to
the extent that Patent Owner conflates a “radiation limit” with a “dose-
volume limit.” Id. at 22–23 (citing PO Resp. 41). According to Petitioner,
the cost function is evaluated in its proposed combination when the radiation
IPR2020-00071
Patent 6,393,096 B1
51
limit for a structure is exceeded so long as the dose-volume constraints are
satisfied. Id. (citing Ex. 1043 ¶ 18).
In its Sur-reply, Patent Owner argues that Petitioner is attempting to
omit limitations in claims 1 and 18 based on the transitional word
“comprising.” See PO Sur-reply 6–7.
Our disposition of claim 18 turns on the interplay of the transitional
word “comprising” with the “incorporating” limitation in claim 1, from
which claim 18 ultimately depends. As discussed above, we reject
Petitioner’s arguments based on the transitional word “comprising” and
determine that the “incorporating a cost function at each iteration” in claim 1
requires incorporating a cost function for each and every cycle of the recited
method. See supra § II.C.2.
In its analysis for claim 18, Petitioner proposes combining the dose-
volume constraints of Morrill-1991’s constrained simulated annealing
method with the optimization algorithm of Oldham and the cost function of
Viggars. See Pet. 50. As a reason for combining Morrill-1991 with Oldham
and Viggars, Petitioner touts Morrill-1991’s use of constraints as being
“computationally more efficient” than merely using weighted penalties. Id.
(quoting Ex. 1022, 1343) (citing Ex. 1002 ¶ 126). Yet Morrill-1991
provides a specific explanation why the use of constraints results in
computational efficiency: “If the sample configuration fails to satisfy these
additional constraints, it is rejected outright (i.e. the [cost function]
algorithm is not called).” Ex. 1022, 1343. Thus, in contrast to the Oldham–
Viggars ground where the cost function is always evaluated, Petitioner’s
proposed optimization process of the instant ground (with Morrill-1991’s
added constraints) rejects some proposed beam configurations outright
IPR2020-00071
Patent 6,393,096 B1
52
because they fail to meet constraints, in which case the cost function is not
called. Petitioner tacitly acknowledges as much when it asks us to focus on
“the undisputed iterations where the constraints are not violated and the
Viggars’s overall CDVH-based cost function is evaluated.” Pet. Reply 16;
see also id. at 16–18. But Petitioner’s argument belies our construction of
the “at each iteration” limitation of claim 1. See supra § II.C.2. Therefore,
because Petitioner’s proposed obviousness combination in this ground does
not incorporate a cost function for each and every cycle of the recited
method, it does not teach the “incorporating a cost function at each iteration”
limitation of claim 1.
Thus, we determine Petitioner has not shown, by a preponderance of
the evidence, that the subject matter of claim 18 would have been obvious
over the combination of Oldham, Viggars, and Morrill-1991.

G. Obviousness Ground Based on Carol-2 and Carol-17
Petitioner contends the subject matter of claims 1 and 18 would have
been obvious over the combination of Carol-2 and Carol-17. Pet. 55–66;
Pet. Reply 23–34. Patent Owner disputes Petitioner’s contentions. PO
Resp. 42–66; PO Sur-reply 14–33. We already have found claim 1 to be
unpatentable over the combination of Oldham and Viggars (see supra
§ II.E.3), so we confine our analysis to claim 18 for this ground. See SAS
Inst. Inc. v. Iancu, 138 S. Ct. 1348, 1359 (2018) (holding a petitioner “is
entitled to a final written decision addressing all of the claims it has
challenged”); Boston Sci. Scimed, Inc. v. Cook Grp. Inc., 809 F. App’x 984,
990 (Fed. Cir. 2020) (non-precedential) (recognizing that the “Board need
not address issues that are not necessary to the resolution of the
IPR2020-00071
Patent 6,393,096 B1
53
proceeding” and, thus, agreeing that the Board has “discretion to decline to
decide additional instituted grounds once the petitioner has prevailed on all
its challenged claims”).

1. Carol-2
Carol-2 is a book chapter from 1997 that sets forth “an overview of
the current status of the clinical implementation of IMRT.”14 Ex. 1020, 17.
Carol-2 describes computer-based inverse treatment planning operations,
including simulated annealing. Id. at 20. Simulated annealing is an iterative
process that “proceeds by randomly changing beam weights, then evaluating
the effect of each change on the dose distribution. The acceptability of a
change is determined by a cost function which is a mathematical
quantification of how conflicting goals will be resolved.” Id.
Carol-2 also describes the PEACOCK Plan treatment planning
system, in which “the target cost function is the mean-squared difference
between realized dose and prescribed dose” and the structure cost function
“is the mean-squared difference between realized dose and zero dose.” Id. at
21. The overall calculated cost is based on the weight assigned to each
structure and target. Id. Accordingly, PEACOCK Plan “uses an interface
which involves assigning graded weights and priorities to the structures and
targets in order to achieve the desired result.” Id.

14 Although Patent Owner disputes the prior art status of Carol-2 (PO
Resp. 42–60; PO Sur-reply 14–32), we do not reach those arguments
because we resolve the remaining grounds on the merits.
IPR2020-00071
Patent 6,393,096 B1
54
2. Carol-17
Carol-17 is a book chapter that attempts to anticipate areas of IMRT
that would be investigated in the years after its publication in 1997.15
Ex. 1021, 243. Carol-17 describes a system for an inverse planning called
CORVUS that “uses partial volume information for each structure out of
which CDVH curves are generated and used as the goal by the optimizer.”
Id. at 247. Carol-17 describes the process:
For each target, the user enters: goal, minimum dose, maximum
dose and percent volume which is allowed to be underdosed.
For each structure, the user enters: desired limit, minimum
dose, maximum dose and percent volume which can be greater
than limit. The system creates CDVH curves for the targets and
structures from these entries which are used by the optimizer as
a representation of the desired dose distribution.
Id. “After a CDVH is constructed from user-entered partial volume values,
the system divides the CDVH into regions and automatically assigns a
relative weight to each,” which are “used to resolve conflicts between the
various CDVH regions defined by the target goals and structure limits.” Id.

3. Claim 18
Claim 18 recites “the step of allowing a radiation limit on the tissue
structure to be exceeded by a set amount if such excess allows better
conformation to the desired target CDVH curve.” Ex. 1001, 18:23–26.
Petitioner’s entire analysis for claim 18 from the Petition is reproduced
below.

15 Although Patent Owner disputes the prior art status of Carol-17 (PO
Resp. 42–60; PO Sur-reply 14–32), we do not reach those arguments
because we resolve the remaining grounds on the merits.
IPR2020-00071
Patent 6,393,096 B1
55
Carol-17 teaches that “weights are used to resolve conflicts
between the various CDVH regions defined by the target goals
and structure limits;” “[t]he default weights favor structures
over targets when such conflicts exist; all structure limits, no
matter how severe, will be met before target goals are met;” and
“[t]he user has the option of selecting, on a target-by-target
basis, whether target goals or structure limits will prevail.”
(Carol-17 at 247.) When the user sets the weights to favor
targets over structures, the “target goals” prevail over the
“structure limits” and the step of claim 18 is satisfied.
(Ex. 1002 ¶159.)
Pet. 66.
Patent Owner argues that Petitioner fails to explain how “the step of
claim 18 is satisfied” when “weights assigned to CDVH regions are set to
permit target goals to ‘prevail’ over structure limits.” PO Resp. 61 (citing
Pet. 66; Ex. 1002 ¶ 159; Ex. 2037 ¶ 151). Patent Owner also argues that
Carol-17 does not disclose whether or how its disclosed system, CORVUS,
“identifies any ‘set amount’ of excess radiation that the tissue structure can
be allowed to receive.” Id. at 62 (citing Ex. 2037 ¶ 153). Patent Owner
further argues that Petitioner does not explain “how Carol-17 discloses or
suggests whether or how CORVUS permits excess radiation only if doing so
‘allows better conformation to the desired target CDVH curve.’” Id. (citing
Ex. 2037 ¶ 154).
In reply, Petitioner argues that an ordinarily skilled artisan would have
known that one of the “fundamental and inherent features of a SARP
optimization” is “exceeding a radiation limit on a tissue structure by a set
amount.” Pet. Reply 31 (citing Ex. 1043 ¶ 53). Petitioner provides the
following explanation:
A POSA would have understood that when a user sets the
weights to favor greater correspondence to the target limits at
IPR2020-00071
Patent 6,393,096 B1
56
the sacrifice of allowing structure limits to be exceeded, the
simulated annealing algorithm would guide the beam
arrangements toward configurations in which the structure
radiation limit is exceeded by a set amount as a trade-off for
better conformation to the prioritized target CDVH. The set
amount by which the radiation limit on the structure can be
exceeded is based on the magnitude of the relative weights
applied to the target and structures.
Pet Reply 32 (citing Ex. 1043 ¶ 55).
In its Sur-Reply, Patent Owner argues that we should ignore
Petitioner’s theory regarding inherent features of simulated annealing as
being an improper new argument under 37 C.F.R. § 42.23(b). PO
Sur-reply 32–33.
We agree with Patent Owner that Petitioner’s analysis from the
Petition fails to show how Carol-17 teaches the limitation of claim 18.
Importantly, the Petition does not show how Carol-17 teaches that any
particular “radiation limit” is “exceeded by a set amount.” Nor does the
Petition show how Carol-17 teaches exceeding a “radiation limit” if such
excess “allows better conformation to the desired target CDVH curve.” We
also have considered Petitioner’s statement that “[w]hen the user sets the
weights to favor targets over structures, the ‘target goals’ prevail over the
‘structure limits’ and the step of claim 18 is satisfied.” Pet. 66 (citing
Ex. 1002 ¶ 159). Dr. Gall’s initial declaration includes the same conclusory
statement. Ex. 1002 ¶ 159. We do not credit these statements because they
are devoid of supporting evidence or reasoning and they do not address the
specific limitation in claim 18. Thus, Petitioner’s arguments from the
Petition are insufficient to establish obviousness based on the combination
of Carol-2 and Carol-17.
IPR2020-00071
Patent 6,393,096 B1
57
We additionally agree with Patent Owner that Petitioner’s reply
arguments about “fundamental and inherent features” of simulated annealing
are improper reply arguments under 37 C.F.R. § 42.23(b). A reply may only
respond to arguments raised in the preceding brief, and, importantly,
“‘[r]espond,’ in the context of 37 C.F.R. § 42.23(b), does not mean proceed
in a new direction with a new approach as compared to the positions taken in
a prior filing.” Consolidated Trial Practice Guide 74. In its Reply,
Petitioner contends that Patent Owner ignores “relevant evidence and fails to
acknowledge the fundamental and inherent features of a SARP
optimization” (Pet. Reply 31), but Petitioner provides wholly new arguments
and testimony about the alleged “fundamental and inherent features.” See
id. at 31–33. In particular, Petitioner bases its analysis almost entirely on
extensive new testimony from Dr. Gall. See id.; Ex. 1043 ¶¶ 53–58. And,
for the first time, Petitioner maps certain recitations of claim 18 to the
teachings of Carol-17. See, e.g., Pet. Reply 32 (citing Ex. 1043 ¶ 55)
(addressing the “exceeded by a set amount” recitation in the first instance).
Thus, contrary to Petitioner’s argument, Patent Owner could not have
“ignore[d] relevant evidence” (Pet. Reply 31) where Petitioner did not
explain its position in the Petition. Because we find Petitioner’s reply
arguments about “fundamental and inherent features of a SARP
optimization” and its mapping of the “exceeded by a set amount” recitation
to be a new direction with a new approach as compared to the positions
taken in the Petition, we do not consider Petitioner’s new reply arguments.
See Intelligent Bio-Sys., Inc. v. Illumina Cambridge Ltd., 821 F.3d 1359,
1369–70 (Fed. Cir. 2016) (holding that the Board did not err in refusing the
reply brief as improper under 37 C.F.R. § 42.23(b) because petitioner relied
IPR2020-00071
Patent 6,393,096 B1
58
on an entirely new rationale to explain why one of skill in the art would have
combined the references at issue).
Thus, we determine Petitioner has not shown, by a preponderance of
the evidence, that the subject matter of claim 18 would have been obvious
over the combination of Carol-2 and Carol-17.

H. Obviousness Ground Based on Carol-2, Carol-17, and Morrill-1991
Petitioner contends the subject matter of claim 18 would have been
obvious over the combination of Carol-2, Carol-17, and Morrill-1991.
Pet. 66–69; Pet. Reply 23–31, 34–35. Patent Owner disputes Petitioner’s
contentions. PO Resp. 42–60, 66–67; PO Sur-reply 14–33.
As with the Oldham–Viggars–Morrill-1991 ground above, Petitioner
cites Morrill-1991 in this ground for teaching the added limitation of
claim 18. See Pet. 66–69. Also similar to above, Petitioner proposes
applying the constrained simulated annealing approach of Morrill-1991 in
the simulated annealing methodology Carol-2 with the CDVH-based cost
functions of Carol-17. Id. at 67 (citing Ex. 1002 ¶ 161). Petitioner relies on
the same rationale for combining Morrill-1991 with the other references as
in the Oldham–Viggars–Morrill-1991 ground. Compare Pet. 67–68, with id.
at 50–52. Petitioner additionally relies on the same arguments that it puts
forth in reply for the Oldham–Viggars–Morrill-1991 ground. Pet. Reply 34–
35.
Patent Owner argues that Petitioner’s obviousness challenges for this
ground should fail based on the same arguments Patent Owner makes with
respect to the Oldham–Viggars–Morrill-1991 ground. PO Resp. 66–67; PO
Sur-reply 33. We agree with Patent Owner. Petitioner’s proposed
IPR2020-00071
Patent 6,393,096 B1
59
combination for this ground incorporates the use of Morrill-1991’s
constraints. Pet. 67. Radiation beam configurations that fail the constraints
would be rejected outright (see Ex. 1022, 1343), so a cost function would
not be incorporated in such a scenario. That means that Petitioner’s
combination does not teach “incorporating a cost function at each iteration”
for the same reasons discussed above. See supra § II.F.2. Thus, we
determine Petitioner has not shown, by a preponderance of the evidence, that
the subject matter of claim 18 would have been obvious over the
combination of Carol-2, Carol-17, and Morrill-1991.

I. Petitioner’s Motion to Exclude Portions of Mr. Romesberg’s
Declaration
Petitioner moves to exclude paragraphs 13–18 from Mr. Romesberg’s
declaration (Exhibit 2044) under Federal Rules of Evidence 602 and 702.
Exclude Mot. 1–2, 9–15. Because we do not rely upon Exhibit 2044 in
rendering this Decision, we dismiss as moot Petitioner’s motion to exclude
paragraphs 13–18 of Mr. Romesberg’s declaration.
III. CONCLUSION
Petitioner has shown, by a preponderance of the evidence, that the
subject matter of claim 1 would have been obvious over the combination of
Oldham and Viggars. Petitioner has not shown, by a preponderance of the
evidence, that (1) the subject matter of claim 18 would have been obvious
over the combination of Oldham, Viggars, and Morrill-1991; (2) that the
subject matter of claim 18 would have been obvious over the combination of
IPR2020-00071
Patent 6,393,096 B1
60
Carol-2 and Carol-17; and (3) the subject matter of claim 18 would have
been obvious over the combination of Carol-2, Carol-17, and Morrill-1991.16

IV. ORDER
Accordingly, it is
ORDERED that claim 1 of the ’096 patent is unpatentable;
FURTHER ORDERED that claim 18 of the ’096 patent is not
unpatentable;
FURTHER ORDERED that Petitioner’s motion to exclude paragraphs
104–110, 113–115, 128, and 129 of Exhibit 2037 is denied;
FURTHER ORDERED that Petitioner’s motion to exclude paragraphs
13–18 of Exhibit 2044 is dismissed as moot; and
FURTHER ORDERED that, because this is a Final Written Decision,
parties to this proceeding seeking judicial review of our decision must
comply with the notice and service requirements of 37 C.F.R. § 90.2.

16 Should Patent Owner wish to pursue amendment of the challenged claims
in a reissue or reexamination proceeding subsequent to the issuance of this
Decision, we draw Patent Owner’s attention to the April 2019 Notice
Regarding Options for Amendments by Patent Owner Through Reissue or
Reexamination During a Pending AIA Trial Proceeding. See 84 Fed. Reg.
16,654 (Apr. 22, 2019). If Patent Owner chooses to file a reissue application
or a request for reexamination of the challenged patent, we remind Patent
Owner of its continuing obligation to notify the Board of any such related
matters in updated mandatory notices. See 37 C.F.R. § 42.8(a)(3), (b)(2).
IPR2020-00071
Patent 6,393,096 B1
61

In summary:

17 As explained above, we do not reach claim 1 for this ground. See supra
§ II.G.
Claims 35 U.S.C. § Reference(s)/Basis
Claims
Shown
Unpatentable
Claims
Not shown
Unpatentable
1 103(a) Oldham, Viggars 1
18 103(a) Oldham, Viggars,
Morrill-1991
18
1, 18 103(a)17 Carol-2, Carol-17 18
18 103(a) Carol-2, Carol-17,
Morrill-1991
18
Overall
Outcome
1 18
IPR2020-00071
Patent 6,393,096 B1
62
For PETITIONER:

Heidi Keefe
Xiaozhen YU
COOLEY LLP
hkeefe@cooley.com
syu@cooley.com

For PATENT OWNER:

Barry J. Coyne
Kent E. Baldauf, Jr.
Bryan P. Clark
THE WEBB LAW FIRM
bcoyne@webblaw.com
kbaldaufjr@webblaw.com
bclark@webblaw.com