`
`
`
`
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`
`EXOCAD GMBH AND EXOCAD AMERICA, INC.
`Petitioner,
`v.
`3SHAPE A/S,
`Patent Owner.
`
`
`Patent No. 9,336,336
`Issue Date: May 10, 2016
`Title: 2D IMAGE ARRANGEMENT
`
`
`
`Inter Partes Review No. IPR2018-00788
`
`
`SECOND DECLARATION OF JOSEPH L. MUNDY, Ph.D.
`
`exocad Ex. 1023
`exocad v. 3Shape, IPR2018-00788
`
`Page 1 of 37
`
`
`
`
`
`Table of Contents
`
`B.
`
`C.
`
`INTRODUCTION ........................................................................................... 1
`I.
`CLAIM CONSTRUCTION ............................................................................ 2
`II.
`III. SACHDEVA ANTICIPATES CLAIMS 1-14, 16-20 AND 22-30 ................. 3
`Sachdeva Meets the “2D Image” Limitation Because It
`A.
`Discloses that the Morphable Face Model May Be 2D ........................ 4
`Sachdeva Discloses that the Morphable Face Model and 3D
`Tooth Model “Remain Separate Representations After Being
`Arranged” .............................................................................................. 8
`Sachdeva Discloses “Either Virtually Cut[ting] at Least a Part
`of Teeth Out of the at Least One 2D Image or Render[ing] a
`Part of the at Least One 2D Image that Includes Teeth Partly or
`Wholly Transparent” ........................................................................... 13
`IV. SACHDEVA ANTICIPATES THE DEPENDENT CLAIMS ..................... 17
`A. Dependent Claims 6-8 ......................................................................... 17
`B. Dependent Claim 9 .............................................................................. 17
`V. ALTERNATIVELY, CLAIMS 1-14, 16-20 AND 22-30 WOULD
`HAVE BEEN OBVIOUS BASED ON SACHDEVA COMBINED
`WITH KOPELMAN ...................................................................................... 18
`A. A POSITA Would Have Replaced the Morphable Model in
`Sachdeva with a 2D Image in Kopelman ............................................ 18
`B. Kopelman Discloses that the 2D Image and 3D Model “Remain
`Separate Representations After Being Arranged” .............................. 20
`VI. WIEDMANN IN VIEW OF SACHDEVA RENDERS CLAIMS 1-14,
`16-20 AND 22-30 OBVIOUS ....................................................................... 23
`A. Wiedmann Discloses a “Virtual 3D Space” ........................................ 23
`B. Wiedmann Discloses that the 2D Image and 3D Model
`“Remain Separate Representations After Being Arranged” ............... 26
`It Would Have Been Obvious to Substitute a Post-Restoration
`Model of the Patient’s Oral Cavity with a Pre-Restoration
`Model ................................................................................................... 27
`
`C.
`
`
`
`i
`
`exocad Ex. 1023
`exocad v. 3Shape, IPR2018-00788
`
`Page 2 of 37
`
`
`
`D.
`
`To the Extent a Virtual 3D Space Is Not Disclosed by
`Wiedmann, It Would Have Been Obvious to Substitute
`Wiedmann’s (Alleged) 2D Optimum Tooth Shape with the 3D
`Tooth Model of Sachdeva in a 3D Space ............................................ 29
`IN THE ALTERNATIVE TO OBVIOUSNESS BASED ON
`SACHDEVA, SACHDEVA/KOPELAMN, OR
`WIEDMANN/SACHDEVA, CLAIMS 6-8 ARE OBVIOUS BY
`FURTHER COMBINING WITH MACDOUGALD ................................... 32
`
`VII.
`
`
`
`
`
`ii
`
`exocad Ex. 1023
`exocad v. 3Shape, IPR2018-00788
`
`Page 3 of 37
`
`
`
`
`I.
`
`INTRODUCTION
`1.
`I have been retained as an expert in the field of computer science
`
`technology, including working with 3D and 2D images and models in CAD
`
`software, which includes digital dentistry software, by Foley & Lardner LLP,
`
`which represents exocad GmbH and exocad America, Inc. (collectively “exocad”)
`
`in this matter. I previously prepared a declaration in this inter partes review,
`
`which was submitted as Exhibit 1002. In this Second Declaration, I have been
`
`asked to prepare a written declaration replying to the statements and assertions
`
`made by the Patent Owner and Patent Owner’s expert, Dr. Saber, as well as certain
`
`portions of the Board’s Institution Decision. My opinions are detailed below.
`
`2.
`
`The documents that I have considered in developing my opinions set
`
`forth in this declaration are referenced in my prior declaration and this declaration,
`
`including the declaration of Eli Saber, Ph.D, Patent Owner’s Response to the
`
`Petition and exhibits cited therein, the deposition transcript from the deposition of
`
`Dr. Saber (Ex. 1026), and additional exhibits 1024-25.
`
`3. My curriculum vitae was submitted previously as Exhibit 1003. My
`
`compensation for working on issues in this matter continues to be based on a rate
`
`of $350 per hour for consulting and $400 per hour for testimony in deposition or
`
`trial, plus reimbursement for reasonably incurred expenses. My compensation
`
`
`
`1
`
`exocad Ex. 1023
`exocad v. 3Shape, IPR2018-00788
`
`Page 4 of 37
`
`
`
`does not depend upon the outcome of this matter or the related litigation, the
`
`opinions I express, or my testimony.
`
`4.
`
`Additional information may become available which would further
`
`support or modify the conclusions that I have reached to date. Accordingly, I
`
`reserve the right to modify and/or enlarge this opinion or the bases thereof upon
`
`consideration of any further discovery, testimony, or other evidence, including any
`
`issues raised by any expert or witness of the Patent Owner, 3Shape A/S
`
`(“3Shape”), or based upon interpretations of any claim term by the Patent Office
`
`different than those proposed in this declaration.
`
`5. My qualifications are set forth in my prior declaration and curriculum
`
`vitae.
`
`II. CLAIM CONSTRUCTION
`6. My analysis of various claim constructions were set forth in my
`
`previous declaration. I respond to certain points argued by Patent Owner and Dr.
`
`Saber here.
`
`7.
`
`“3D virtual model of at least part of an oral cavity of the patient.” I
`
`understand that the Board disagreed with my proposed construction in its
`
`Institution Decision. According to the Board, this phrase refers only to the
`
`patient’s cavity pre-restoration and excludes using an oral cavity of the patient
`
`
`
`
`
`2
`
`exocad Ex. 1023
`exocad v. 3Shape, IPR2018-00788
`
`Page 5 of 37
`
`
`
`post-restoration. Institution Decision at 9-10. While I disagree with than
`
`construction, I will apply it in this declaration.
`
`8.
`
`“Remain separate representations after being arranged.” I agree
`
`with Patent Owner that the term “after being arranged” is part of the claim
`
`language, and thus in my opinion, my previous proposed construction should be
`
`used, but with the phrase “after being arranged” added. In that case, the
`
`construction should be: “the 2D image and the 3D virtual model remain in their
`
`respective formats and are not merged into a single representation after being
`
`arranged.”
`
`9.
`
`I note also that the Board correctly construed the phrase such that
`
`there is no temporal limitation placed on how long the 2D image and the 3D virtual
`
`model remain separate. Institution Decision at 8-9 (Board’s construction “does not
`
`preclude subsequent merging or fusing together of the separate data representations
`
`after alignment, provide the 2D image and the 3D model remain separate at least
`
`momentarily after having been ‘aligned.’”).
`
`III. Sachdeva Anticipates Claims 1-14, 16-20 and 22-30
`10.
`I explained in my first declaration how Sachdeva anticipates claims 1-
`
`14, 16-20 and 22-30 of the ’336 patent. Ex. 1002 ¶¶466-571 (citing back to
`
`various disclosures of Sachdeva in paragraphs 287-445). Here, I address some of
`
`the arguments set forth by Patent Owner and Dr. Saber.
`
`
`
`
`
`3
`
`exocad Ex. 1023
`exocad v. 3Shape, IPR2018-00788
`
`Page 6 of 37
`
`
`
`A.
`
`11.
`
`Sachdeva Meets the “2D Image” Limitation Because It Discloses
`that the Morphable Face Model May Be 2D
`I previously explained that Sachdeva meets this limitation, including
`
`because the “morphable model 102” in Sacheva may be 2D. Ex. 1002 ¶¶302-07.
`
`12.
`
`I note that Sachdeva includes language in the specification and
`
`Abstract allowing for the virtual patient model to be 2D, which means a POSITA
`
`would understand Sachdeva to disclose that the morphable model 102 may also
`
`sometimes be 2D, whether combined with 2D or 3D tooth model. See Ex. 1005 at
`
`5:63-66 (“two dimensional and/or three-dimensional virtual patient model”); see
`
`also Abstract (broadly stating that the “image data” may be “3D image data and/or
`
`2 image data”).
`
`13. Sachdeva actually very broadly discloses that the virtual patient model
`
`consists of combining a “first set of data” and a “second set of data” to create it
`
`(Ex. 1005 at 7:8-19), and this portion of Sachdeva does not specify whether those
`
`data sets are 2D or 3D, which a POSITA would thus read as being either, both or
`
`one of each.
`
`14. Furthermore, it was well known to a POSITA that morphable models
`
`can be 2D, and thus referring to a “morphable model” without specifying
`
`dimensions necessarily includes both 2D and 3D. For example, 2D morphable
`
`models are disclosed in Lu, Xiaguang, Image Analysis for Face Recognition, in
`
`Proceedings of Personal Notes, Department of Computer Science and Engineering,
`
`
`
`
`
`4
`
`exocad Ex. 1023
`exocad v. 3Shape, IPR2018-00788
`
`Page 7 of 37
`
`
`
`Michigan State University (2003) (available at http://www.face-rec.org/interesting-
`
`papers/). Ex. 1024 at p. 16 and p. 20.
`
`15.
`
`In Figure 7 of Sachdeva, the morphable model 102 is also shown as
`
`2D. The 2D scale can be seen in the lower left corner, in contrast to the 3D scale
`
`for the 3D tooth model on the other part of the Figure
`
`
`
`Indication of 2D
`
`Indication of 3D
`
`
`
`
`
`16. As I mentioned in my previous declaration, the other Sachdeva patent,
`
`U.S. Patent No. 7,234,937 (Ex. 1013) (the “Sachdeva ’937 patent”), further
`
`demonstrates that a POSITA would read Sachdeva to include 2D morphable
`
`
`
`
`
`5
`
`exocad Ex. 1023
`exocad v. 3Shape, IPR2018-00788
`
`Page 8 of 37
`
`
`
`models. In the Sachdeva ’937 patent, which has basically the same description of
`
`the “morphable model 102,” the “tooth model 104” and the virtual patient model,
`
`claim 1 provides for a “first set of data” without specifying whether that data is 2D
`
`or 3D. Ex. 1013 at 28:7-9. Then, dependent claim 2 provides that the
`
`representation combining the first and second data sets may be “a three-
`
`dimensional representation. Id. at 28:40-42. Additionally, dependent claim 33
`
`provides that the combined digital representation comprises “superposition of a
`
`two dimensional representation and a three dimensional representation.” Id. at
`
`30:45-48. These claims, which are supported by the disclosure in the specification,
`
`thus clearly include the morphable model being 2D, with a 2D-to-3D
`
`superposition. In my opinion, a POSITA would not read Sachdeva inconsistent
`
`with how the same disclosure must be read and claimed by the same inventor in
`
`this other nearly identical (in relevant parts) Sachdeva patent.
`
`17.
`
`I understand that Patent Owner cites to portions of Sachdeva in which
`
`Sachdeva discusses that the morphable model can be created using two-
`
`dimensional color pictures of the face to create a 3D model. 3Shape Resp. at 30
`
`and 32-33. I do not believe a POSITA would read that description as excluding the
`
`2D morphable model, both for the reasons previously stated and also because those
`
`are cited as just an “alternative,” or “example” or one embodiment. See Ex. 1005
`
`at 10:57-11:11 (describing it as “[o]ne such alternative” at 10:60); 7:29-38 (“For
`
`
`
`
`
`6
`
`exocad Ex. 1023
`exocad v. 3Shape, IPR2018-00788
`
`Page 9 of 37
`
`
`
`example the virtual patient model could be created by a superposition of the
`
`following data sets: … and color photographs of the face, that are combined in the
`
`computer to form a 3D morphable face model.”); 14:18-20 (“the morphable model
`
`102 is obtained, for example, from color photographs using the techniques
`
`described previously”). Instead, Sachdeva discloses that “morphable models can
`
`be built on various known approaches …..” Id. at 11:12-13.
`
`18.
`
`I understand that Patent Owner cites to Sachdeva’s citing of a Blanz
`
`article. 3Shape Resp. at 30 (citing Ex. 2003, Blanz article). I do not read Blanz as
`
`excluding 2D morphable models, however, and there is no statement to that effect
`
`in the article. Sachdeva merely citing to the Blanz article thus does not provide
`
`any basis to conclude that Sachdeva excludes 2D morphable models, in my
`
`opinion.
`
`19. Finally, I understand that Patent Owner contends that the disclosure of
`
`a 2D virtual model in Sachdeva means that Sachdeva is only disclosing 2D-to-2D
`
`and 3D-to-3D superposition, but not 2D-to-3D superposition. 3Shape Resp. at 31.
`
`I disagree, and various parts of Sachdeva demonstrate this is wrong. Sachdeva
`
`repeatedly discloses 2D-to-3D superpositions. See, e.g., Ex. 1005 at Fig. 5 (13:41-
`
`14:3); Fig. 28 (28:64-17).
`
`
`
`
`
`7
`
`exocad Ex. 1023
`exocad v. 3Shape, IPR2018-00788
`
`Page 10 of 37
`
`
`
`B.
`
`20.
`
`Sachdeva Discloses that the Morphable Face Model and 3D Tooth
`Model “Remain Separate Representations After Being Arranged”
`I understand that Patent Owner contends that this limitation is not
`
`disclosed by Sachdeva, including based on the word “composite” in Sachdeva.
`
`3Shape Resp. at 34-42. That is not how a POSITA would read Sachdeva,
`
`however. A POSITA reading a reference disclosing superpositioning of an image
`
`and a model would not assume, unless being told so explicitly, that the image and
`
`model can no longer remain separate representations after being arranged/aligned.
`
`And thus absent such a statement in Sachdeva, a POSITA would assume the
`
`morphable model and tooth model can remain separate representations.
`
`21. This is further supported by the fact that Sachdeva discloses the
`
`ability to move the morphable model 102 and tooth model 104 to align them by
`
`human interaction. Ex. 1005 at 15:8-9. When being moved into such alignment,
`
`there is at least some time period in which the morphable model and tooth model
`
`are aligned and separate representations, and by remaining separate
`
`representations, further adjustments are possible.
`
`22.
`
`I also disagree that the word “composite” in Sachdeva has a meaning
`
`requiring that the morphable model and tooth model can no longer remain separate
`
`representations when part of a composite. That is not how a POSITA would
`
`understand the word, which is defined more generally, for example, as “made up of
`
`distinct components”. Ex. 1025 (definition of “composite”).
`
`
`
`
`
`8
`
`exocad Ex. 1023
`exocad v. 3Shape, IPR2018-00788
`
`Page 11 of 37
`
`
`
`23.
`
`I note that Patent Owner has cited to an article that discusses some
`
`types of “image compositing.” See Ex. 2005. That article is irrelevant to
`
`understanding what the word “composite” means, however. The article never
`
`states that the concept of a “composite” model cannot include the model consisting
`
`of separate representations. The methods for layering different 2D image data
`
`described in Ex. 2005 are also very different from what is done in Sachdeva, for
`
`example.
`
`24. Additionally, there are other portions of Sachdeva that further
`
`explicitly indicate that the morphable model and tooth model may remain separate
`
`representations. For example, Sachdeva defines the word “superimpose” to mean
`
`“registered to each other via software in the workstation.” Ex. 1005 at 11:54-55.
`
`Registration does not require the things being registered to no longer remain
`
`separate representations, of course, as a POSITA would understand. In other
`
`words, therefore, Sachdeva explicitly defines the concept of “superposition” to
`
`include superimposing representations such that they remain separate
`
`representations after be aligned.
`
`25. This definition of “superimpose” as set forth in Sachdeva, in which
`
`superimposed representations may remain separate, is also consistent with the use
`
`of the term “superimpose” in the ’336 patent itself. See Ex.1001 at 18:1-6 (“In
`
`general it is an advantage of the method and embodiments that it/they enable(s)
`
`
`
`
`
`9
`
`exocad Ex. 1023
`exocad v. 3Shape, IPR2018-00788
`
`Page 12 of 37
`
`
`
`dental laboratories (labs) to superimpose a patient’s actual face and smile images
`
`in the design process ….”).
`
`26.
`
`I also note that the word “combined” in Sacheva also imposes no
`
`requirement that the morphable model and tooth model no longer remain separate
`
`representations after being aligned. A POSITA would not understand the word
`
`“combine” to have such a meaning, either. Furthermore, the word “combine” is
`
`used in the ’336 patent in a manner that the representations being combined may
`
`remain separate. See Ex. 1001 at 9:49-55 (“combined view”); 11:15-22
`
`(“combined in this way”).
`
`27.
`
` Taken together, this means that a POSITA would understand that the
`
`statement in Sachdeva that “the images can be “combined or superimposed to
`
`create a virtual patient model” (Ex. 1005 at 25:28-45) means that the
`
`representations may remain separate after being combined or superimposed in the
`
`virtual patent model arrangement.
`
`28.
`
`I understand that Patent Owner also argues that Figure 6 of Sachdeva
`
`does not show the morphable model 102 and tooth model 104 aligned and able to
`
`be moved separately. 3Shape Resp. at 37-38. This misses the point of my citation
`
`to Figure 6. Ex. 1002 ¶337. Figure 6 shows the separateness of the representations
`
`pre-alignment, and Sachdeva describes that they are then moved into alignment,
`
`remaining separate. Ex. 1005 at 14:28-32 (stating that various icons “allow the
`
`
`
`
`
`10
`
`exocad Ex. 1023
`exocad v. 3Shape, IPR2018-00788
`
`Page 13 of 37
`
`
`
`user to position the tooth model 104 relative to the morphable model 102 in order
`
`to combine the two in a common coordinate system and construct a composite
`
`model.”); 14:32-45 (further disclosing that the “Align References” icon allows the
`
`user to perform the alignment). Again, the fact that the alignment may be
`
`performed, e.g., by human interaction, which includes the user dragging the tooth
`
`model over to the morphable model to make them superimposed, is part of the
`
`description here that shows they are separate representations, at least at that
`
`moment of alignment.
`
`29.
`
`I understand that Patent Owner also cites to a separate disclosure of
`
`the 2D-to-3D alignment in Figure 28 (28:65-29:17) and the ability of a user to see
`
`different perspective views of the teeth and occlusal plane 508 in Figure 43
`
`(29:18-26), as alleged evidence that the 2D and 3D representations do not remain
`
`separate in Figure 28. 3Shape Resp. at 40. 3Shape is misreading Sachdeva, in my
`
`opinion. Figure 28 only further supports that a 2D face image and 3D tooth model
`
`are superpositioned and remain separate representations in Sachdeva. Figure 28
`
`shows alignment of a 2D face image, 3D tooth model and a 2D representation of
`
`the occlusal plane (the straight line). Sachdeva states that the occlusal plane is a
`
`straight line, and that it also can be converted into a 3D representation if the user
`
`wants to examine it with the 3D tooth model. Ex. 1005 at 28:65-29:17. That result
`
`is then shown in a different figure – Figure 43. Id. at 29:18-26. The ability of the
`
`
`
`
`
`11
`
`exocad Ex. 1023
`exocad v. 3Shape, IPR2018-00788
`
`Page 14 of 37
`
`
`
`user to visualize the 2D image and 3D tooth model in Figure 28, as superpositioned
`
`and separate representations, is not changed, however. Moreover, if one 2D
`
`feature – the occlusal plane – can be converted to 3D in another figure, that only
`
`further supports that the 2D face image and 3D tooth model remain separate
`
`representations the whole time in Figure 28.
`
`30. The fact that the representations are able to be pulled up separately
`
`later further shows that they are separate (e.g., Fig. 67 and 45:26-34), whether
`
`pulled up later in an aligned or non-aligned manner, because it shows that the
`
`representations remain in a separate format after the alignment occurs.
`
`31. The ability to “toggle back and forth” between the views (9:40-48),
`
`further demonstrates that the representations remain separate. I disagree with
`
`Patent Owner’s contention that such toggling does not include toggling back and
`
`forth between the morphable model 102 and the tooth model 104. See 3Shape
`
`Resp. at 40. The toggling refers broadly to a “variety of image information,” and
`
`thus a POSITA would not read this broad language to be limiting and excluding
`
`the morphable model and tooth model.
`
`
`
`
`
`12
`
`exocad Ex. 1023
`exocad v. 3Shape, IPR2018-00788
`
`Page 15 of 37
`
`
`
`C.
`
`Sachdeva Discloses “Either Virtually Cut[ting] at Least a Part of
`Teeth Out of the at Least One 2D Image or Render[ing] a Part of
`the at Least One 2D Image that Includes Teeth Partly or Wholly
`Transparent”
`32. As I previously explained in my first declaration, Sachdeva discloses
`
`this limitation both Figure 6 and also with various references to the ability to hide
`
`and make data transparent or partially transparent. Ex. 1002 ¶¶313-16.
`
`33. Again, Figure 6 clearly shows the teeth cut from the image:
`
`
`
`
`
`34.
`
`I understand that despite what is shown in this Figure, Patent Owner
`
`argues that the teeth are not cut or transparent in the Figure because the
`
`corresponding description mentions teeth. See 3Shape Resp. at 42-43. In my
`
`opinion, however, a POSITA would not read Sachdeva this way. In fact, a
`
`POSITA would have looked at the difference in how face images are depicted in
`
`Sachdeva and have understood that the inventors knew how to draw an image that
`
`
`
`
`
`13
`
`exocad Ex. 1023
`exocad v. 3Shape, IPR2018-00788
`
`Page 16 of 37
`
`
`
`included teeth that had not been cut or made transparent, yet purposefully
`
`portrayed Figure 6 differently than, for example, Figures 22 or 67, which show
`
`teeth:
`
`
`
`
`
`35.
`
`I also note that the statement in Sachdeva, cited by 3Shape, about
`
`scaling the morphable model to the tooth model such that the spatial dimensions of
`
`the teeth are substantially the same (Ex. 1005 at 14:46-52) would not be read by a
`
`POSITA to exclude the possibility (as shown) that the scaled teeth in the
`
`morphable model are subsequently cut out or made transparent after such scaling is
`
`performed.
`
`36.
`
`In addition to Figure 6, as I previously stated in my first declaration,
`
`there are other portions of Sachdeva that also disclose the hiding or transparency of
`
`portions of the images and models. I understand that Patent Owner argues that
`
`Sachdeva is only disclosing hiding portions of the virtual patent model, and not the
`
`morphable model 102, nor teeth. 3Shape Resp. at 43. I disagree.
`
`
`
`
`
`14
`
`exocad Ex. 1023
`exocad v. 3Shape, IPR2018-00788
`
`Page 17 of 37
`
`
`
`37. For Sachdeva’s reference at 15:25-27, Sachdeva broadly discloses the
`
`hiding of “image data,” which is not a term that would be understood by a POSITA
`
`to be limited to the virtual patient model. The reference to the virtual model
`
`having been created does not limit the subsequent actions to being performed only
`
`on that model, and there is also no statement that the image data would not include
`
`teeth.1
`
`38. For Sachdeva’s reference at 30:8-12, the statement is that tasks
`
`include “hiding and displaying various aspects of the virtual patient model, soft
`
`tissue, occlusional planes, and other features …” By referring broadly to “other
`
`features,” a POSITA would understand that the task is not limited to hiding
`
`features on the virtual patient model and not excluding those features from being
`
`teeth.
`
`39.
`
` There are additional references to hiding/transparency in Sachdeva,
`
`also. For example, Sachdeva provides for “[t]he ability to see transparently
`
`through one or more objects …” (44:35-38), which is in a section (43:45-44:49)
`
`that lists various tools that can be used concurrently (44:39), including the
`
`1 In my opinion, the claim limitation also does not require that teeth, as opposed to
`
`another part of the 2D image, be rendered partly or wholly transparent. The
`
`limitation refers back to the “2D image that includes teeth,” not that the “part of”
`
`this image that is rendered transparent must be limited to the teeth.
`
`
`
`
`
`15
`
`exocad Ex. 1023
`exocad v. 3Shape, IPR2018-00788
`
`Page 18 of 37
`
`
`
`simultaneous tool of superimposing x-rays and other images on the proposed set-
`
`up and compare the set-up with images (43:52-58). Together, this means that
`
`transparency may be used on the morphable model in the superposition. Another
`
`reference to semi-transparency in Sachdeva is that statement that “[a]nother feature
`
`of the software is that it allows the teeth in either or both arches to be displayed as
`
`semi-transparent objects ….” Ex. 1005 at 38:61-67.
`
`40. Finally, while somewhat unnecessary to discuss in view of all the
`
`other disclosures in Sachdeva of transparency and cutting, I note that my citations
`
`in my first declaration to the transparent 2D X-ray image in Figure 9 and the
`
`transparency shown in Figure 8 also meet the limitation. I understand that 3Shape
`
`is arguing that lips are not shown, but x-rays may include lips and teeth, as the
`
`’336 patent itself provides with dependent claim 25, which allows for the 2D
`
`image (from claim 1) that includes lips and teeth to be an X-ray. See Ex. 1001,
`
`claim 25. I also note that the teeth in Figure 9 are shown as included in the label
`
`for X-ray 116 and are shown as transparent. Id., Fig. 9. I also note that a POSITA
`
`would not understand the word “render” to require that the portion that is
`
`transparent to be modified to be transparent, as 3Shape argues incorrectly, in my
`
`opinion. For example, the dictionary definition of render includes much broader
`
`meanings. See Ex.1025 (defining “render” in various ways, including to simply
`
`mean “provide”).
`
`
`
`
`
`16
`
`exocad Ex. 1023
`exocad v. 3Shape, IPR2018-00788
`
`Page 19 of 37
`
`
`
`IV. Sachdeva Anticipates the Dependent Claims
`A. Dependent Claims 6-8
`41.
`I understand that Patent Owner argues that Sacheva does not disclose
`
`these limitations because Sachdeva mentions teeth without stating that they are
`
`prepared or unprepared. 3Shape Resp. at 46-48. I disagree. As a POSITA would
`
`understand, the functions of Sachdeva apply to any type of teeth for the 3D tooth
`
`models, including both prepared or unprepared. The superpositioning functions
`
`are the same, regardless of what type of teeth are in the 3D tooth model. And,
`
`without stating any limit on what type of teeth the tooth model may consist of, a
`
`POSITA would have understood that the 3D tooth model in Sachdeva can include
`
`any type of teeth, prepared and unprepared.
`
`B. Dependent Claim 9
`42.
`I note that all of the disclosures I cited in my first declaration and
`
`above concerning the limitation in the independent claim about “either virtually
`
`cut[ting] at least a part of teeth out of the at least one 2D image or rendering a part
`
`of the at least one 2D image that includes teeth partly or wholly transparent” also
`
`meet this claim limitation in dependent claim 9 because the hiding and rendering
`
`fully transparent portions of images are examples of virtual cutting. This is
`
`supported, for example, by the use of the term “cutting” in ’336 patent. Ex. 1001
`
`at 8:54-56 (“cutting” means that a portion is “removed, deleted, made invisible,
`
`etc.”).
`
`
`
`
`
`17
`
`exocad Ex. 1023
`exocad v. 3Shape, IPR2018-00788
`
`Page 20 of 37
`
`
`
`V. Alternatively, Claims 1-14, 16-20 and 22-30 Would Have Been Obvious
`Based on Sachdeva Combined with Kopelman
`A. A POSITA Would Have Replaced the Morphable Model in
`Sachdeva with a 2D Image in Kopelman
`43. For this part of the analysis, I am assuming that Sachdeva only
`
`discloses a 3D, and not a 2D, morphable model, even though in my opinion
`
`Sachdeva does disclose both a 2D or 3D morphable model. As previously
`
`explained in my first declaration, it is my opinion, in that case, that the above-listed
`
`claims would have been obvious based on Sachdeva combined with Kopelman. I
`
`understand that Patent Owner argues that a POSITA would not have replaced the
`
`(alleged) 3D morphable model in Sachdeva with the 2D image in Kopelman
`
`because “Sachdeva already discloses providing a 3D morphable model” and
`
`replacing it with a 2D image “would have defeated the purpose of converting the
`
`2D color pictures into a 3D morphable model in the first place.” 3Shape Resp. at
`
`48-49. I disagree with 3Shape’s position.
`
`44. As an initial matter, Sachdeva is not limited to the face data coming
`
`from 2D color pictures that are converted into a 3D morphable model. For
`
`example, Sachdeva provides that the 3D face model could come from a single face
`
`scanner. Ex. 1005 at 10:11-40. In my opinion, a POSITA could have thus been
`
`replacing such a 3D face model created by a scanner with a 2D face image of
`
`Kopelman.
`
`
`
`
`
`18
`
`exocad Ex. 1023
`exocad v. 3Shape, IPR2018-00788
`
`Page 21 of 37
`
`
`
`45. Additionally, even when considering just Sachdeva’s embodiment in
`
`which the morphable model is created from multiple 2D color pictures, a POSITA
`
`would nonetheless have recognized that all the reasons to simplify that process by
`
`using a single 2D image instead, which I have previously mentioned, would still
`
`apply. In my opinion, a POSITA would have been interested in simplification of
`
`the steps, and Kopelman discloses such a simplification with the use of one 2D
`
`image, instead of multiple 2D images combined into a 3D morphable model, for all
`
`the reasons I have previously expressed.
`
`46.
`
`I understand that Patent Owner also takes issue with my contention
`
`that one reason to combine is that a POSITA would have recognized the fact that
`
`sometimes only a 2D image is available to the dentist or dental clinician. 3Shape
`
`Resp. at 49. In my opinion, 3Shape is misconstruing the reason to combine. While
`
`Sachdeva may contemplate one scenario in which there may be multiple 2D
`
`images that are available to the dentist or dental clinician, this does not change the
`
`fact that in the real world, there may be times when only one 2D image is
`
`available to the dentist or dental clinician. In my opinion, a POSITA would have
`
`recognized this real world fact, and that would have been a reason to combine
`
`Sachdeva with Kopelman, making the system in Sachdeva useful even when only a
`
`2D image of the face is available to the dentist or dental clinician.
`
`
`
`
`
`19
`
`exocad Ex. 1023
`exocad v. 3Shape, IPR2018-00788
`
`Page 22 of 37
`
`
`
`47.
`
`I understand that Patent Owner also cites to a typographical error in
`
`my first declaration and attempts to misconstrue it. See 3Shape Resp. at 49-50.
`
`The sentence in my first declaration should have read as follows: “changing
`
`Sachdeva such that the morphable model could also be 2D would merely be
`
`substituting one technique (3D-to-3D) for another (2D-to-[3]D) to obtain a
`
`predictable result.” Ex. 1002 ¶577 (i.e., change the emphasized word from “2D” to
`
`“3D”). This is obviously how the sentence was meant to read, which is consistent
`
`with the sentences and paragraphs that precede it and footnote 28 (and especially
`
`since I never suggest the 3D tooth model is changed to 2D).
`
`B. Kopelman Discloses that the 2D Image and 3D Model “Remain
`Separate Representations After Being Arranged”
`I understand that Patent Owner argues that the Kopelman does not
`
`48.
`
`disclose that the 2D image and 3D model “remain separate representations after
`
`being arranged.” 3Shape Resp. at 50. This is similar to 3Shape’s argument about
`
`Sachdeva allegedly not disclosing this limitation. While I note that the analysis is
`
`not required if Sachdeva discloses that these representations remain separate after
`
`being arranged, it is also my opinion that the representations in Kopelman remain
`
`separate after being arranged.
`
`49. First, I note that there is no statement anywhere in Kopelman that the
`
`representations do not remain separate. Again, a POSITA would not read a
`
`reference as disclosing a requirement that the representations are no longer
`
`
`
`
`
`20
`
`exocad Ex. 1023
`exocad v. 3Shape, IPR2018-00788
`
`Page 23 of 37
`
`
`
`separate after being arranged unless there were some explicit statement in the
`
`reference stating this. Since Kopelman does not include such a statement, a
`
`POSITA would not have read Kopelman in the manner suggested by 3Shape. A
`
`POSITA r