Dental Monitoring SAS v. Align Technology, Inc.

[5] 6 UNITED STATES DISTRICT COURT [7] NORTHERN DISTRICT OF CALIFORNIA [9] 10 DENTAL MONITORING SAS, 11 Plaintiff, No. C 22-07335 WHA 12 v. 13 ALIGN TECHNOLOGY, INC., ORDER RE SUMMARY JUDGMENT 14 Defendant. [15] 16 INTRODUCTION 17 Dueling summary judgment motions contest the validity of two of plaintiff’s patents and 18 their infringement. The asserted claims of both patents recite an abstract idea and include no 19 further inventive concept and are therefore invalid. Defendants’ motion is GRANTED IN PART; 20 all else is DENIED AS MOOT. 21 STATEMENT 22 Dental aligners, like braces, are used to reposition a patient’s teeth. After a dental 23 practitioner scans a patient’s starting dentition and determines a desired final dentition, third 24 parties like defendant Align Technology, Inc. design and manufacture a series of aligners that, 25 when worn in sequence, exert pressure on the patient’s teeth and thereby gradually move them 26 from the starting to final dentition. Typically, a dental practitioner must visually assess a 27 patient’s aligner at regular intervals to evaluate progress and determine if and when a patient 1 Defendant Align Technology, Inc., creator of the “Invisalign” group of aligner products, 2 is described by plaintiff as a “dominant provider[ ] of aligner and dental treatments” (Dkt. No. 3 101 at 1). Align, plaintiff says, was “caught flat-footed” when COVID-19 upended the 4 “outdated approach of in-person visits” then common among dental practitioners employing 5 Align’s products (ibid.). In contrast, plaintiff-patentee Dental Monitoring SAS, founded in 6 2014, describes itself as a “pioneer[ ] and the technological leader in remote dentistry and 7 orthodontics,” with some 200 pending and issued patents in the space (ibid.). Plaintiff now 8 alleges that in defendant’s scramble to re-gain its footing and adopt COVID compatible 9 methods for remote aligner treatment, it infringed upon plaintiff’s patents. 10 The parties have identified two claims for summary judgment argument. First is claim 12 11 (and claim 1, on which it depends) of U.S. Patent No. 10,755,409, titled “Method For 12 Analyzing an Image of a Dental Arch,” which discloses a method for acquiring an image of the 13 dental arch of a patient. The invention recited in the ’409 patent claims a method for guiding a 14 patient to capture images of their dental arches and aligner fit in order to facilitate remote 15 treatment. Claims 1 and 12 are reproduced here: 16 1. A method for acquiring an image of a dental arch of a patient, said method comprising the following steps: [17] a) activation of an image acquisition apparatus so as to acquire an 18 image, called “analysis image,” of said arch; 19 b) analysis of the analysis image by means of a deep learning device trained by means of a learning base; [20] c) determination, for the analysis image, as a function of the result 21 of the analysis in the preceding step, of a value of an image attribute; [22] d) comparison of said image attribute value with a setpoint; [23] e) sending of an information message as a function of said 24 comparison, the information message being related to the quality of the image acquired or to the position of the 25 acquisition apparatus in relation to said arch or to the setting of the acquisition apparatus or to the opening of the mouth or to 26 the wearing of a dental appliance, or to a combination thereof, 27 to check whether the analysis image respects the setpoint and, if it 1 12. The method as claimed in claim 1, in which the information message is sent by the acquisition apparatus. [2] 3 The second claim at issue is claim 14 (and claim 1, on which it depends) of U.S. Patent 4 No. 11,049,248, titled “Method for Analyzing an Image of a Dental Arch,” which discloses a 5 method for assessing the shape and fit of an orthodontic aligner by way of a “deep learning 6 device, trained by means of a learning base.” The claim at issue provides as follows: [7] 1. A method for assessing the shape of an orthodontic aligner, said 8 method comprising the following steps: 9 a) more than 1 week after the start of the treatment with the aligner, acquisition of at least one image at least partially 10 representing the aligner in a service position in which it is worn by a patient, called “analysis image”, the analysis image being 11 a photograph, or an image extracted from a film; 12 b) analysis of the analysis image by means of a deep learning device, trained by means of a learning base, so as to determine 13 a value 14 for at least one tooth attribute of an “analysis tooth zone” representing, at least partially, a tooth on said analysis image, 15 the tooth attribute relating to a separation between the tooth represented by the analysis tooth zone, and the aligner 16 represented on the analysis image, 17 in the step a), a cellphone used to acquire the analysis image. [18] 14. The method of claim 1, in which the step b) comprises the 19 following steps: 20 1) creation of a learning base comprising more than 1000 images of dental arches, or “historical images”, each historical image 21 representing an aligner worn by a “historical” patient and comprising one or more zones each representing a tooth, or 22 “historical tooth zones”, to each of which, for at least one tooth attribute relating to a separation between the tooth represented 23 by the historical tooth zone considered, and the aligner represented, a tooth attribute value is assigned; [24] 2) training of at least one deep learning device, by means of the 25 learning base; [26] 3) Submission of the analysis image to the deep learning device 27 for it to determine at least one probability relating to: analysis tooth zone; and [1] the attribute value of the tooth represented on said analysis 2 tooth zone; 3 4) determination, as a function of said probability, of an amplitude of said separation. [4] 5 The parties have filed dueling summary judgment motions concerning the above claims. 6 Plaintiff argues that defendant has infringed upon the claims as a matter of law. Defendant, 7 meanwhile, argues that the claims are invalid under 35 U.S.C. Section 101 , that they are 8 invalid under 35 U.S.C. Section 112 , and that, if they are valid, defendant’s products do not 9 infringe. This order follows full briefing and oral argument. 10 ANALYSIS 11 A “court shall grant summary judgment if the movant shows that there is no genuine 12 issue as to any material fact and that the movant is entitled to judgment as a matter of law.” 13 Fed. R. Civ. P. 56(a). “Summary judgment is appropriate only if, taking the evidence and all 14 reasonable inferences drawn therefrom in the light most favorable to the non-moving party, 15 there are no genuine issues of material fact.” Furnace v. Sullivan, 705 F.3d 1021, 1026 (9th 16 Cir. 2013) (internal quotation marks omitted). Material facts are those that may affect the 17 outcome of the suit. Anderson v. Liberty Lobby, Inc., 477 U.S. 242 , 247–49 (1986). A genuine 18 dispute about a material fact exists if there is enough evidence for a reasonable jury to find for 19 the non-moving party. Ibid. “In judging evidence at the summary judgment stage, the court 20 does not make credibility determinations or weigh conflicting evidence. Rather, it draws all 21 inferences in the light most favorable to the nonmoving party.” Soremekun v. Thrifty Payless, 22 Inc., 509 F.3d 978, 984 (9th Cir. 2007). 23 Infringement requires a valid patent. This order therefore begins, and, because they have 24 met their burden, ends, with defendant’s Section 101 argument. 25 1. INVALIDITY. 26 Section 101 of the Patent Act defines patent eligible subject matter to include “any new 27 and useful process, machine, manufacture, or composition of matter, or any new and useful 1 nature, natural phenomena, and abstract ideas – “the basic tools of scientific and technological 2 work” – from the otherwise unfettered reach of Section 101. Alice Corp. Pty. v. CLS Bank 3 Int'l, 573 U.S. 208 , 216 (2014). In creating the exception, the Supreme Court reasoned that 4 “monopolization of those tools through the grant of a patent might tend to impede innovation 5 more than it would tend to promote it.” Mayo Collaborative Servs. v. Prometheus Lab'ys, Inc., [6] 566 U.S. 66, 71 (2012). 7 The Supreme Court’s two-step Alice test serves to sift ineligible laws, phenomena, and 8 ideas from patent-eligible subject matter. “First, we determine whether the claims are directed 9 to a patent-ineligible concept, such as an abstract idea. If so, we consider the elements of each 10 claim both individually and as an ordered combination to determine whether the additional 11 elements transform the nature of the claim into a patent-eligible application.” Customedia 12 Techs., LLC v. Dish Network Corp., 951 F.3d 1359, 1362 (Fed. Cir. 2020) (citing Mayo, 566 13 U.S. at 78-79) (cleaned up). If a claim is captured by step one and flunks step two, it is 14 ineligible under Section 101. That is the case here. 15 A. ALICE STEP ONE. 16 At step one, the district court must consider the claims “in their entirety to ascertain 17 whether their character as a whole is directed to excluded subject matter.” Internet Pats. Corp. 18 v. Active Network, Inc., 790 F.3d 1343, 1346 (Fed. Cir. 2015). However, “[a]t some level, all 19 inventions . . . embody, use, reflect, rest upon, or apply laws of nature, natural phenomena, or 20 abstract ideas. Thus, an invention is not rendered ineligible for patent simply because it 21 involves an abstract concept.” Alice, 573 U.S. at 217. The district court must therefore 22 “articulate what the claims are directed to with enough specificity to ensure the step one 23 inquiry is meaningful.” Thales Visionix Inc. v. United States, 850 F.3d 1343, 1347 (Fed. Cir. 24 2017). Where, as here, a particular technological environment is at issue, the step one inquiry 25 often turns on whether the claims “focus on a specific means or method that improves the 26 relevant technology” or are instead “directed to a result that itself is the abstract idea and 27 merely invoke generic processes and machinery.” McRO, Inc. v. Bandai Namco Games Am. (i) Claim 14 of the ’248 Patent. [1] This order begins with claim 14 (and claim 1, on which it depends) of the ’248 [2] patent. Stripped of excess verbiage and techno-jargon, claim 1 is of “a familiar class of claims [3] directed to a patent-ineligible concept,” namely “collecting information, analyzing it, and [4] displaying certain results of the collection and analysis.” Electric Power Group, 830 F.3d 5 1350, 1353 (Fed. Cir. 2016). As the Federal Circuit explained: [6] Information as such is an intangible. Accordingly, we have treated 7 collecting information, including when limited to particular content (which does not change its character as information), as within the 8 realm of abstract ideas. In a similar vein, we have treated analyzing information by steps people go through in their minds, 9 or by mathematical algorithms, without more, as essentially mental processes within the abstract-idea category. And we have 10 recognized that merely presenting the results of abstract processes of collecting and analyzing information, without more (such as 11 identifying a particular tool for presentation), is abstract as an ancillary part of such collection and analysis. [12] 13 Electric Power Group, 830 F.3d 1350, 1353-54 (citations omitted) (emphasis added). Here, 14 too, the purported advance lies in a “process of gathering and analyzing information of a 15 specified content,” here dental aligner shape and fitment, “and displaying the results, and not 16 any particular assuredly inventive technology for performing those functions.” Id. at 17 1354. Moreover, the patent’s specification concedes that this analysis was conventionally 18 performed by an orthodontist, via in-person visual inspection or through review of patient19 provided images (’248 patent cols. 1:11-15; 27:55-59). Claims directed to “analyzing 20 information by steps people go through in their minds, or by mathematical algorithms, without 21 more,” are directed to “essentially mental processes within the abstract-idea 22 category.” Electric Power Group, 830 F.3d at 1354 ; see also Synopsys, Inc. v. Mentor 23 Graphics Corp., 839 F.3d 1138 (Fed. Cir. 2016). Claim 14 details the method of training by 24 which a deep learning device is applied to the task above. True, the claim describes the 25 training process with a great deal of particularity, but the focus of the claim itself remains the 26 abstract idea, albeit accompanied by a detailed set of instructions regarding its implementation 27 in a particular technological context and field of use. 1 First, plaintiff asserts that claim 14 is directed to “a specific technological solution: a 2 first-of-its-kind, remote aligner assessment technology powered by a particular trained deep 3 learning device” (Dkt. No. 112 at 3). However, the patent specification concedes that neural 4 networks, the patent’s preferred species of a deep learning device, are “set[s] of algorithms 5 well known to a person skilled in the art,” and simply directs practitioners to pick from any one 6 of over a dozen off-the-shelf neural networks “specializing in the classification of images” or 7 “in the location and detection of objects in an image” (’248 patent col. 16:5-36). Neither claim 8 nor specification details a novel (or any) algorithm or otherwise discloses an improvement to 9 extant deep learning devices. It is not enough to be the first to apply an abstract idea to a 10 specific technological environment, such as a deep learning device. Electric Power Group, 11 830 F.3d at 1354 (“Most obviously, limiting the claims to [a] particular technological 12 environment . . . is, without more, insufficient to transform them into patent-eligible 13 applications of the abstract idea at their core.”). Nor is it enough to be the first to introduce 14 generic neural networks or other deep learning devices to a particular field, such as dental 15 aligners, without more. Ibid. 16 Relatedly, plaintiff notes that the claim “provides numerous concrete benefits,” including 17 greater speed, ease, and the ability to “make a quantitative measurement of the separation 18 between an aligner and teeth,” as opposed to the “qualitative[ ] assess[ment]” performed by 19 dental practitioners’ visual or manual inspections (Dkt. No. 112 at 2). But “[c]laiming the 20 improved speed or efficiency inherent with applying the abstract idea on a computer is 21 insufficient to render the claims patent eligible.” Enco Systems Inc. v. DaVincia, 845 22 Fed.Appx. 953, 955 (2021) (cleaned up); see also Recentive Analytics, Inc. v. Fox Corp., No. 23 CV 22-1545-GBW, 2023 WL 6122495 (D. Del. Sept. 19, 2023) (Judge Gregory Williams) 24 (rejecting patentee’s attempt to distinguish “qualitative” human processing from “quantitative” 25 machine learning analysis). The assertion that a computer can perceive more granular aligner 26 separations than the naked eye is likewise an improvement inherent to applying the abstract 27 idea on a machine learning device. This is evidenced by the claim and specification’s 1 But plaintiff argues more: this is a “specific deep learning device trained by the specific 2 learning base of claim 14,” the first trained to perform “aligner assessment . . . for aligner 3 treatment” (Dkt. No. 112 at 3) (emphasis added). As plaintiff’s counsel stated during oral 4 argument, the generic neural networks listed in the specification “are architectures for products 5 that can be trained, but they don’t do anything. They’re not useful until they have been 6 trained” (Dkt. No. 143 at 33). The argument, then, is that the training of the generic device is 7 transformative: it improves the deep learning device by allowing it to operate in a particular 8 field of use, dental aligner assessment. 9 While a method of training and application may conceivably constitute a “specific 10 asserted improvement in [deep learning device] capabilities,” that is not the case here. Finjan, 11 Inc. v. Blue Coat Sys., Inc., 879 F.3d 1299, 1303 (Fed. Cir. 2018) (internal quotation marks 12 omitted). Here, claim 14’s method of training merely describes how a generic deep learning 13 device is applied (or limited) to a particular field of use. Neural networks specialized in the 14 classification of images, as well as the location and detection of objects in an image, were well 15 known in the prior art (’248 patent col. 16:5-30). These networks, by their very nature, must 16 be trained before they are put to a use. (Dkt. No. 143 at 33). The claim limitations identified 17 by plaintiff as “improved computer techniques” – that the images in the learning base be 18 “images of dental arches,” that they represent “an aligner work by a ‘historical patient” and 19 “compris[e] one or more zones each representing a tooth,” that a “tooth attribute value is 20 assigned,” and so forth – are not improvements to the training or operation of deep learning 21 devices, they are field-of-use limitations describing the configuration of a deep learning device 22 to perform the abstract idea at issue in the field at issue. 23 Alice is instructive. There, a claim drawn to the abstract idea of intermediated settlement 24 required the following steps: 25 (1) “creating” shadow records for each counterparty to a transaction; (2) “obtaining” start-of-day balances based on the 26 parties’ real-world accounts at exchange institutions; (3) “adjusting” the shadow records as transactions are entered, 27 allowing only those transactions for which the parties have permitted transactions. [1] Alice, 573 U.S. at 224. True, the claims at issue in Alice “simply instruct[ed] the practitioner to [2] implement the abstract idea of intermediated settlement on a generic computer.” Ibid. But [3] implicit in the command to “do it on the computer” is a great deal of field-of-use specific [4] configuration: a computer yanked off the shelf will not create shadow records, obtain start-of5 day balances from accounts, and so on. That is true for most, if not all, “do it on a computer” [6] claims held invalid under Section 101. The work necessary to turn a generic computer to the [7] field of use at issue does not improve the computer, just as the limitation of an abstract idea to [8] the same field of use does not render the idea non-abstract. [9] Also true, claim 14 provides a great deal of particularity when disclosing how one may go [10] about applying the abstract idea (collecting, analyzing, and displaying information) in a [11] specific technological environment (a generic deep learning devices) adapted to a particular [12] field of use (dental aligners). But these field of use particulars are not improvements. “Do it [13] on a computer” is “do it on a computer,” even when a draftsman takes the time to describe the [14] doing in great detail. Alice, 573 U.S. 208 at 224 (“This Court has long warned against [15] interpreting § 101 in ways that make patent eligibility depend simply on the draftsman’s art.”) [16] (cleaned up). To be clear, this order does not hold that the process of re-tooling an extant [17] device to a field of use can never constitute patentable subject matter, only that in this instance, [18] the claims asserted do little more than describe the process of “doing it on a computer.” [19] Second, plaintiff argues that claim 14 is directed to a “specific, new treatment method of [20] aligner assessment,” citing to Vanda Pharms. Inc. v. W.-Ward Pharms. Int'l Ltd. 887 F.3d 21 1117 (Fed. Cir. 2018). Not so. In Vanda, the claim at issue recited a method for treating a [22] schizophrenic patient with iloperidone, which required, inter alia, “(1) determining the [23] patient's CYP2D6 metabolizer genotype by (a) obtaining a biological sample and (b) [24] performing a genotyping assay; and (2) administering specific dose ranges of iloperidone [25] depending on the patient's CYP2D6 genotype.” Id. at 1134 . Vanda held that this constituted a [26] “novel method of treating a disease” that required the administration of specific dosages of [27] 1 highlight[ed] the significance of the specific dosages by explaining how certain ranges of 2 administered iloperidone correlate with the risk of QTc prolongation,” and how the method 3 patented resulted in treatment “safer for patients because it reduce[d] the risk of QTc 4 prolongation.” Id. at 1135 . 5 Claim 14 gets only halfway there: it discloses a method of evaluating aligner fit, but no 6 more. The claim does not disclose any novel, specific course of treatment to be carried out on 7 the basis of that information. Nor does the specification. What if the aligner separation is 8 0.5mm? 0.7mm? When should a patient revert to a previous aligner, or progress to the next, 9 and so forth? The claim in Vanda disclosed novel answers to these questions. The claim at 10 issue here does not get beyond collect, analyze, and display. 11 (ii) Claim 12 of the ’409 Patent. 12 Claim 12 of the ’409 patent, dependent on claim 1, does not fare any better. Claim 1 is 13 directed to an abstract idea: the acquisition of an image, the analysis of the contents of that 14 image, and the presentation of information regarding the quality and composition of the image 15 to the camera operator, intended to guide them in the acquisition of a new, better image. Claim 16 12 adds only that the information message be sent by the same device taking the picture (e.g., a 17 cellphone). Neither claim nor specification provides any specific algorithm or other 18 improvement to the deep learning device to be used, beyond that it may be selected from any 19 of a dozen or more existing options, and that it must be “trained by means of a learning base” 20 (’409 patent cols. 16:15-45; 32:17-22). 21 Again, this claim merely recites a common practice long performed by dental 22 practitioners. The specification concedes that orthodontists “conventionally performed” 23 assessments of aligner treatment via images transmitted to them by patients (’409 patent col. 24 1:10-15; Dkt. No. 112 at 14). Defendant further concedes that, because “the quality and 25 accuracy of photos taken by medically untrained patients were not consistent or guaranteed,” a 26 dental practitioner would “review [the photos], and “[i]f a patient sent poor-quality or incorrect 27 photos,” the treatment provider would “ask[ ] the patient to send new ones” (Dkt. No. 112 at 1 message” relating to the quality or composition of the subpar image, and guidance as to the 2 acquisition of a better one. Moving beyond the field of dental aligners, it is beyond cavil that 3 “capture, analysis, critique, recapture” is a dialogue as old as photography itself. 4 Again, attempts to reframe the focus of the claim fall short. For example, plaintiff argues 5 that the claimed method allows “even medically untrained patients to remotely acquire aligner 6 and dental images that are suitable for clinical assessment” (Dkt. No. 112 at 16). But neither 7 remote acquisition nor suitability for clinical assessment are limitations connected to the 8 language of the claim. Moreover, medically untrained patients have long captured and relayed 9 images of their aligners to practitioners for assessment (’409 patent col. 1:10-15). Plaintiff also 10 notes that absent a deep learning device, the acquisition and analysis of patient images was a 11 slower and less reliable process, but again, “[c]laiming the improved speed or efficiency 12 inherent with applying the abstract idea on a computer is insufficient to render the claims 13 patent eligible.” Enco Systems, 845 Fed.Appx. at 955 (cleaned up). 14 Finally, plaintiff asserts that the claim cannot involve “only generic computer elements” 15 because, as conceded by defendant’s expert, “using a deep learning device to carry out image 16 analysis was not part of the standard of care in aligner treatment in 2017” (Dkt. No. 112 at 16). 17 This argument again presumes that the limitation of a generic computer component to a new 18 field of use constitutes an improvement of the computer. As discussed above, that is not the 19 case. 20 B. ALICE STEP TWO. 21 At step two we ask what more, if anything, a claim has to offer. The claim elements, 22 individually or as an ordered combination, must contain an “inventive concept” or “additional 23 features” sufficient to ensure that the claim, in practice, amounts to “significantly more” than a 24 patent upon the ineligible abstract idea. Ultramercial, Inc. v. Hulu, LLC, 772 F.3d 709 , 715 25 (Fed. Cir. 2014) (emphasis added); Alice, 573 U.S. at 221. The object of the inquiry is the 26 claim language: a district court may not rely on “technological details set forth in the patent’s 27 specification and not set forth in the claims to find an inventive concept.” Intell. Ventures I 1 Comcast Cable Commc’ns, LLC, 874 F.3d 1329, 1338 (Fed. Cir. 2017) (“To save a patent at 2 step two, an inventive concept must be evident in the claims.”). The Federal Circuit has 3 explained that “an inventive concept can be found in the non-conventional and non-generic 4 arrangement of known, conventional pieces.” BASCOM Global Internet Servs. v. AT & T 5 Mob., 827 F.3d 1341, 1350 (Fed. Cir. 2016). But, “the components must involve more than 6 performance of ‘well-understood, routine, conventional activit[ies]’ previously known to the 7 industry.” In re TLI Commc'ns LLC Pat. Litig., 823 F.3d 607, 613 (Fed. Cir. 2016) (quoting 8 Alice, 573 U.S. at 225). 9 (i) Claim 14 of the ’248 Patent. 10 Taken alone, the elements of claim 1 invoke no more than the performance of an abstract 11 idea in a particular field of use (dental aligner assessment) via generic and functional hardware 12 (a deep learning device). A “deep learning device” is “preferably a neural network,” which, in 13 turn, “is a set of algorithms well known to a person skilled in the art,” including the dozen or 14 so listed by the patent. The “training” of that device is not described with any specificity – the 15 limitation merely commands one to “train[ ] at least one deep learning device, by means of the 16 learning base” containing those 1000 plus images (’248 patent col. 34:32-33). The use of a 17 “learning base,” including one “comprising more than 1000 images,” was well known at the 18 time (Dkt. No. 104-4 at ¶111 – 118). The device’s method of analysis is equally opaque: 19 handed an “analysis image,” the device simply “determine[s] at least one probability relating 20 to: the presence in a location of said analysis image, of an analysis tooth zone; and the 21 attribute value of the tooth represented on said analysis tooth zone,” after which the deep 22 learning device again “determine[s], as a function of said probability, [ ] an amplitude of said 23 separation.” The remaining particulars do not move the needle – they describe the process of 24 aligner fit evaluation, long practiced by dental practitioners, adapted to a deep learning device. 25 Plaintiff’s arguments to the contrary cut no figure. First, plaintiff argues that the 26 individual elements of the claim were not well understood, routine, or conventional. Plaintiff’s 27 arguments here boil down to the fact that, while the neural network architectures themselves 1 specific learning base in the claim] to perform the particular steps of the claimed aligner 2 method” (Dkt. No. 112 at 10). Elsewhere, plaintiff contends that defendant’s cited references 3 are “deficient” because they do not “disclose a deep learning device,” or the use of “a specific 4 deep learning device trained by means of a specific learning base” (id. at 13). This is the “first 5 to the field” argument in new guise. As explained at length above, generic computer parts and 6 functions do not become patentable every time they enter a new field. Nor is applying a 7 generic computer to a new field an inventive concept. Moreover, the method of training by 8 which claim 14 tools a generic deep learning device for operation in the relevant field was well 9 known and understood (Dkt. No. 104-4 at ¶ 111-118). 10 Second, plaintiff argues that “the record is replete with evidence that the specific 11 combination of claim elements provided numerous concrete benefits that simply were not 12 available in July 2017” (Dkt. No. 112 at 11). Even if true, “simply appending generic 13 computer functionality to lend speed or efficiency to the performance of an otherwise abstract 14 concept does not meaningfully limit claim scope for purposes of patent eligibility.” CLS Bank, 15 Int'l v. Alice Corp., 717 F.3d 1269, 1286 (Fed.Cir.2013) (en banc) aff'd, 573 U.S. 208 (2014); 16 see also Intell. Ventures, 792 F.3d at 1367; Bancorp Servs., LLC v., Sun Life Assurance Co. of 17 Can., 687 F.3d 1266, 1278 (Fed. Cir. 2012). 18 Finally, plaintiff argues that Exergen Corp. v. Kaz USA, Inc. compels a finding of patent 19 eligibility because the combination of the claim’s elements was not well understood, routine, 20 and conventional. 725 F. App’x 959 (Fed. Cir. 2018). Not so. In Exergen, the court reasoned 21 that “[1] the inventor determined for the first time the coefficient representing the relationship 22 between temporal-arterial temperature and core body temperature and [2] incorporated that 23 discovery into an unconventional method of temperature measurement. As a result, the 24 method is patent eligible[.]” Id. at 966 . Nothing in the claim at issue discloses a discovery 25 related to aligner fit assessment. A long-practiced process has simply been transposed from 26 the mind to a machine. [27] (ii) Claim 12 of the ’409 Patent. [1] Claim 12 also lacks an inventive concept, whether its limitations are considered in [2] isolation or as an ordered combination. An “image acquisition apparatus” is just about [3] anything with a camera attached to it, including a cellphone. An “analysis image” is “an image [4] . . . of a dental arch of a patient.” That image is “analyz[ed]” by means of a “deep learning [5] device trained by means of a learning base.” A “deep learning device” includes, but is not [6] limited to, more than a dozen existing neural networks. The “learning base” is the same as [7] above. The “analysis,” meanwhile, produces a “determination, for the analysis image, as a [8] function of the results of the analysis in the preceding step, of a value for an image attribute.” [9] The image attribute may, but need not, relate to “the orientation of the acquisition apparatus [10] upon acquisition of the image,” “the quality of the image,” “the dental situation of the patient,” [11] or “the presence or absence of a . . . orthodontic, appliance, or to the state of the opening of the [12] mouth.” It is, fairly stated, just an attribute related to the image. An “information message” is [13] again fairly summed up as a message containing any information related to the quality or [14] content of the image. Finally, the claim discloses that, if necessary, the message should “guide [15] the operator . . . to acquire a new analysis image.” That “guidance” is undefined, but [16] presumably includes anything relating to the content or quality of the analysis image. Finally, [17] the dependent claim adds that the cellphone or other “acquisition apparatus” that took the [18] “analysis image” may be used to send the “information message” (i.e., the cellphone you took [19] the picture with displays a message telling you how to take a better one). None of the above [20] constitutes an inventive concept. Nor do the various limitations that serve to limit the claim to [21] the field of dental aligners. Likewise, when considered in an ordered combination, these claim [22] limitations do not amount to significantly more than the execution of an abstract idea by means [23] of generic computer components. [24] Plaintiff’s contrary arguments re-tread unproductive ground. It is not enough that a deep [25] learning device had not previously been turned to the field of dental aligners. And as before, [26] the training of the device for use in that field does not constitute a sufficient inventive concept. [27] 1 CONCLUSION 2 Claim 14 and underlying independent claim | of U.S. Patent No. 11,049,248 B2 recite an 3 abstract idea without an inventive concept and thus, under Alice, rate as invalid under 35 4 U.S.C. Section 101 . The came is true for claim 12 and independent claim 1 of U.S. Patent No. 5 10,755,409 B2. To the extent above, defendant’s motion is GRANTED. 6 Because the claims at issue are invalid, this order need not reach the question of written 7 description or infringement. The remainder of the parties’ motions are DENIED AS MOOT. 8 As to next steps, the Court has reviewed defendant’s pending motion to stay, and 9 understands that plaintiff intends to file an opposition, but has not yet done so (Dkt. No. 150). 10 The Court will come to the issue when it has been fully briefed. [11] 12 IT IS SO ORDERED. || Dated: May 16, 2024 3 15 Pee 16 A_ILLIAM ALSUP 17 UNITED STATES DISTRICT JUDGE