DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Amendment
The Amendment filed on Mar 10th, 2026 has been entered. Claims 1, 4, 6–9, 12 and 14–16 are currently pending. Claims 2–3, 5, 10–11 and 13 have been canceled.
Response to Arguments
Applicant's arguments filed 03/10/2026 have been fully considered. Some arguments are persuasive with respect to the 35 U.S.C. §101 rejection, while others regarding the 35 U.S.C. §112(a) and §103 rejections are not persuasive, as explained below.
Applicant’s arguments filed 3/10/2026 have been fully considered and are persuasive with respect to the prior rejection under 35 U.S.C. §101, in light of Applicant’s arguments, see pages 8–12 of the Remarks over the amended claims 1 and 9. Upon reconsideration of the amended claim scope, the Examiner finds Applicant’s arguments persuasive at least to the extent that the present record does not support maintaining the prior §101 rejection. Therefore, the previous 35 U.S.C. § 101 rejection has been withdrawn following Applicant's amendment to the claims.
Applicant’s arguments filed 3/10/2026 have been fully considered but are not persuasive with respect to the rejection under 35 U.S.C. §112(a). Applicant’s arguments, see pages 7–8 of the Remarks, state that amended independent claims 1 and 9 are now limited to a specific stepwise processing flow incorporating pixel comparison, RGB three-dimensional coordinate values, distance calculation, Gaussian filtering, and display/visualization of a candidate biopsy site, and therefore are sufficiently supported by the specification. The Examiner respectfully disagrees.
The amendments add additional downstream comparison and display limitations, but they do not cure the written-description deficiency regarding the claimed generation model and the claimed removal of melanoma features. The amendment provides no flowchart, pseudocode, training procedure, or architectural diagram showing how the generation model performs melanoma feature removal. A person of ordinary skill in the art reading the specification would not know: (i) Which GAN training methodology to use (vanilla GAN, conditional GAN, CycleGAN, StyleGAN2-ADA, etc.); (2) what training constraints, losses, supervision enforce removal of melanoma features while preserving corresponding non-lesion skin structure; or (3) how the generation model identifies melanoma features, how such features are represented or disentangled from non-melanoma image content.
Accordingly, the amended claims remain directed to computer-implemented functional subject matter for which the specification fails to describe the computer and algorithm in sufficient detail such that one of ordinary skill in the art could reasonably conclude that the inventors possessed the claimed subject matter. Therefore, the rejection under 35 U.S.C. §112(a) is maintained.
Applicant’s arguments filed 3/10/2026 have been fully considered but are not persuasive with respect to the rejection under 35 U.S.C. §103. Applicant’s arguments, see pages 13–17 of the Remarks, state that Lapiere does not disclose or suggest applying a generation model to generate a hypothetical normal image in which lesion features are removed, and that Lee does not disclose first classifying a lesion as melanoma or nevus and then applying a generation model only when the lesion is classified as melanoma. Applicant also argues that Lee does not disclose converting corresponding RGB pixel values of the generated image and original image into a three-dimensional coordinate space to derive candidate biopsy sites. The Examiner respectfully disagrees.
The rejection is based on the combined teachings of Lapiere, Lee, and Teixeira, not on any single reference alone. Lapiere is relied upon for teaching an AI-based skin-lesion/ melanoma analysis system including classification of skin images/ lesions and RGB image-analysis operations. Lee is relied upon for teaching use of a generation model GAN to generate a fake or normal medical image in which a lesion area is removed and for identifying a lesion region by comparing the input medical image with the generated image. Teixeira is relied upon for teaching automated difference-based detection of candidate abnormal regions by thresholding or selecting pixels/ regions based on pixel differences. The prior Office action already identified Lapiere for melanoma/ nevus-type classification, Lee for generating a lesion-removed image and comparing it with the input image, Lapiere for RGB three-dimensional coordinate values, and Teixeira for predefined threshold/ connected-pixel candidate-region detection.
Applicant’s argument that Lapiere does not teach a generation model is not persuasive because Lee is relied upon for that limitation. Applicant’s argument that Lee does not teach first classifying the lesion as melanoma or nevus is not persuasive because Lapiere is relied upon for the classification limitation. Applicant’s argument that Lee does not teach RGB three-dimensional coordinate values is not persuasive because Lapiere is relied upon for the RGB coordinate-value representation, while Lee supplies the generated image and original/generated image comparison.
Applicant argues that Lee's purpose is "visually highlighting lesion regions to assist diagnosis" whereas the present invention "provides an algorithm for recommending candidate biopsy sites for biopsy, which is a different technical objective". This argument is not persuasive because it conflates intended use with structure and function. See Hewlett-Packard Co. v. Bausch & Lomb Inc., 909 F.2d 1464, 1469 (Fed. Cir. 1990) ("apparatus claims cover what a device is, not what a device does"). Recommending biopsy at the site of maximum lesion difference is an obvious application of Lee's visualization. It requires no additional structure, algorithm, or technological insight beyond recognizing that biopsies should target diseased tissue. In addition, when Lee's technique is applied to Lapiere’s melanoma-classified skin image, the identified lesion/ difference region corresponds to the claimed candidate biopsy site.
Applicant argues the combination fails to teach: "Determining pixel values as three-dimensional coordinate values having R, G, and B as axes", "Comparing pixel values of corresponding pixels", "Identifying the at least one candidate biopsy site by comparing the pixel values". These arguments ignore express disclosures in the cited prior arts: Lapiere teaches RGB three-dimensional pixel-value analysis, Lee teaches comparing the original/input image with the generated fake/ normal image to identify the lesion/difference region, and Teixeira teaches threshold-based/ connected-pixel selection of candidate anomaly regions based on image differences.
The combination would have produced predictable results with a reasonable expectation of success because the references use known computer-implemented medical-image analysis techniques for their established purposes: Lapiere’s skin-lesion classification and RGB pixel analysis, Lee’s generated lesion-removed comparison image, and Teixeira’s threshold-based pixel/region selection. Therefore, the rejection under 35 U.S.C. §103 is maintained.
Applicant’s arguments concerning the canceled claims have been fully considered. Claims 2–3, 5, 10–11, and 13 have been canceled. Accordingly, the prior rejections directed specifically to those canceled claims are moot. However, the limitations from the canceled claims have been incorporated into amended independent claims 1 and 9, those limitations have been considered in the present rejection of the amended claims.
Based on these facts, this action is made FINAL.
Claim Rejections - 35 USC § 112(a)
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
The following is a quotation of the first paragraph of pre-AIA 35 U.S.C. 112:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.
Claims 1, 4, 6–9, 12, and 14–16 are rejected under 35 U.S.C. § 112(a) as failing to comply with the written description requirement.
Independent claim 1 recites an electronic device comprising a processor configured to perform, among other functions, "classify a skin image as melanoma or nevus by inputting the skin image to a classification model, when the skin image is classified as melanoma, identify melanoma features in the skin image by using a generation model to generate an image from which the melanoma features are removed, determine pixel values according to color of pixels of the skin image and the generated image, the pixel values identified as three-dimensional coordinate values having R (Red), G (Green), and B (Blue) as axes, respectfully, according to RGB values of the color of each pixel, identify at least one candidate biopsy site by comparing pixel values of corresponding pixels of the skin image and the generated image, and display the at least one candidate biopsy site on the skin image".
For computer-implemented functional claim limitations, the specification must disclose the computer and the algorithm, such as the necessary steps, formulas, rules, flowcharts, or other operative details, that perform the claimed function in sufficient detail such that one of ordinary skill in the art can reasonably conclude that the inventor possessed the claimed subject matter. See MPEP § 2161.01(I).
Here, the specification does not describe the claimed generation-model function in sufficient detail to demonstrate possession of the claimed invention. In particular, the claims require a generation model that identifies melanoma features in a skin image and generates an image from which the melanoma features are removed. However, the specification describes this feature at a high level of result-oriented functionality, without providing sufficient operative detail regarding how the generation model performs the claimed melanoma-feature removal.
For example, the specification states that the generation model may learn melanoma features and nevus features from skin-image learning data and may generate an image similar to an input skin image with melanoma features removed or with morphological characteristics of a nevus. The specification also states that the generation model may be learned using a GAN, such as StyleGAN2. However, merely naming GAN or StyleGAN2, and stating the intended result of producing an image from which melanoma features are removed, does not describe the specific algorithm by which melanoma features are identified, represented, separated from non-melanoma skin content, and removed while preserving the corresponding non-lesion or nevus-like image structure.
The specification does not provide sufficient disclosure of, for example, the particular training inputs/ outputs, labeling strategy, ground-truth targets, loss functions, constraints, segmentation masks, feature-disentanglement technique, conditioning mechanism, model architecture modifications, inference rules, or other algorithmic steps that cause the generation model to remove melanoma features while preserving corresponding non-melanoma image features. Without such disclosure, the claim language covers any generation model that achieves the desired result of generating a melanoma-feature-removed image, rather than an invention that the specification shows the inventors actually possessed.
Applicant’s amendment adds further limitations concerning determining pixel values according to color, representing RGB pixel values as three-dimensional coordinate values, comparing corresponding pixel values of the skin image and the generated image, calculating distances, filtering with a Gaussian filter, and displaying the candidate biopsy site on the skin image. These limitations may further describe downstream comparison and display operations after the generated image has already been produced. However, these additional limitations do not cure the lack of written description for the earlier and essential claimed function of using a generation model to generate an image from which melanoma features are removed. That is, the added pixel-value comparison and RGB-coordinate limitations explain, at most, how the skin image and generated image may be compared once both images exist. They do not explain how the generated image is produced in the first instance, how melanoma features are identified by the generation model, or how melanoma features are removed from the generated image. Therefore, the amended claims remain broader than the disclosure reasonably supports.
Furthermore, dependent claims 4, 6–8, 12, and 14–16 do not cure this deficiency. Claims 4 and 12 add distance calculation and identification of a predefined number of pixels as the candidate biopsy site. Claims 6 and 14 add Gaussian filtering. Claims 7 and 15 add learning of the classification model using melanoma and nevus images. Claims 8 and 16 further recite that the generation model is learned to identify melanoma features and nevus features and to generate an image in which melanoma features are removed from melanoma images. These dependent limitations do not provide, or point to, sufficient algorithmic disclosure in the specification showing how the claimed generation model performs melanoma-feature identification and removal.
Accordingly, because the specification does not describe the generation model and its algorithm in sufficient detail to show that the inventors possessed the claimed function of identifying melanoma features and generating an image from which melanoma features are removed, claims 1, 4, 6–9, 12, and 14–16 are rejected under 35 U.S.C. § 112(a) for lack of written description.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claim(s) 1, 7–9, and 15–16 are rejected under 35 U.S.C. 103 as being unpatentable over Lapiere (Lapiere et al, US 2019/0340762 A1, 2019) in view of Lee (Lee et al. KR 2021-0098381, 2021).
Regarding claim 1, with deficiencies of Lapiere noted in square brackets [ ], Lapiere discloses an electronic device for AI-based recommendation of a melanoma biopsy site ( [0039], [0070], [Fig. 2]: an electronic/ computer-based skin abnormality monitoring system for analyzing skin images and classifying skin lesions, including melanoma, using machine-learning/ image-processing techniques. ), comprising:
a processor configured to:
( [0041]: processor 104 includes a memory 118 configured to. )
classify a skin image as melanoma or nevus by inputting the skin image to a classification model,
( [0039], [0067]-[0068], [0070], [Fig 2, Step 204, Step 216]: Lapiere teaches a machine-learning dermoscopy/ skin-lesion analysis pipeline for classifying skin abnormalities or lesions, including melanoma and mole/ nevus-type lesions. )
when the skin image is classified as melanoma, identify melanoma features in the skin image [ by using a generation model to generate an image from which the melanoma features are removed, ]
( [0094], [Fig 2, Step 224, box 202f ]: Lapiere teaches performing image data matching/comparison after classification of a skin abnormality/lesion, including a melanoma-type lesion, and using the image-analysis result to identify and evaluate the skin abnormality. )
determine pixel values according to color of pixels of the skin image and [ the generated image ], the pixel values identified as three-dimensional coordinate values having R (Red), G (Green), and B (Blue) as axes, respectfully, according to RGB values of the color of each pixel,
( [0005], [0081–0082]: Lapiere teaches pixel-value computations for skin-image analysis and teaches placing pixels in a three-dimensional space, wherein the three axes represent the R-G-B color channels, and further teaches computing three-dimensional Euclidean coordinate values/metrics in that RGB space. )
[ identify at least one candidate biopsy site by comparing pixel values of corresponding pixels of the skin image and the generated image, and ]
display the skin-lesion analysis with [ at least one candidate biopsy site on ] the skin image.
( [0034–0035], [Fig. 1 & 22]: Lapiere teaches displaying/ presenting skin-image and skin-abnormality analysis information to a user through an electronic/ computer-based system, including skin images and associated skin-lesion/ abnormality analysis results. )
As noted above, Lapiere does NOT expressly discloses where Lee teaches:
identify melanoma features in the skin image by using a generation model to generate an image from which the melanoma features are removed,
( [0008], [0045–0054]: Lee teaches using a generation model GAN to generate a fake/ normal medical image from an input medical image, wherein a lesion area is removed from the input medical image. Lee further teaches training, using the generation model to generate the fake/ normal image. )
the generated image
( [0008], [0045–0054]: Lee teaches the generated image as the fake/ normal medical image generated by the generation model from the input medical image, wherein the lesion area is removed.
Accordingly, with respect to the limitation “determine pixel values according to color of pixels of the skin image and the generated image”, Lee supplies the claimed generated image, i.e., the fake/ normal medical image generated from the input medical image with the lesion area removed. Therefore, Lapiere’s RGB pixel-value determination is applied to the pixels of both Lapiere’s skin image and Lee’s generated fake/ normal image. )
identify at least one candidate biopsy site by comparing pixel values of corresponding pixels of the skin image and the generated image, and
( [0008], [0072]: Lee teaches comparing the input/original medical image with the generated fake/ normal medical image to identify the lesion region, based on pixel change/ difference between the disease/ original image and the fake/ normal generated image. )
display the at least one candidate biopsy site on the skin image.
( [0072]: Lee teaches visualization processing for highlighting and displaying the difference part between the input medical image and the fake medical image as the lesion area. )
It would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to modify Lapiere’s AI-based skin-lesion analysis system with Lee’s generation-model technique to generate a fake/ normal lesion-removed image and compare it with the original skin image. Lapiere already classifies skin lesions, including melanoma, performs post-classification image analysis, and represents pixel color values in RGB three-dimensional coordinate space. Lee provides the known technique of generating a lesion-removed image and comparing the original image with the generated image to identify and display the lesion/ difference region. A person of ordinary skill in the art would have been motivated to combine these teachings to more accurately localize the melanoma/ lesion region after Lapiere’s classification and present that localized region as a candidate biopsy site. Applying Lapiere’s RGB pixel-value analysis to Lee’s original/ generated image comparison would have been a predictable use of known image-analysis techniques for their established purposes, with a reasonable expectation of success.
Regarding claim 7, Lapiere [as modified by Lee] teaches the electronic device of claim 1, wherein the classification model is learned to classify whether a skin image being input is melanoma or nevus by using, as learning data, a plurality of skin images including a plurality of melanoma images and a plurality of nevus images and answer information on whether each skin image is melanoma or nevus.
( [0061], [0068], [0072]: Lapiere teaches training/ learning a machine-learning classification model using skin-lesion image data, including melanoma images and mole/ nevus-type images, with corresponding classification information, so that the model classifies an input skin image/ lesion as melanoma or nevus/ mole-type lesion. )
Regarding claim 8, Lapiere [as modified by Lee] teaches the electronic device of claim 1, wherein the generation model is learned to identify melanoma features and nevus features from a plurality of skin images including a plurality of melanoma images and a plurality of nevus images
( [0061], [0068], [0072]: Lapiere teaches using a plurality of skin images including melanoma images and mole/ nevus-type images as learning data for training model on dataset labeled melanoma (skin anomalies) vs nevus (moles). )
and to generate an image in which the melanoma features are removed from the melanoma images.
( [0008], [0045–0054]: Lee teaches using/ training the generation model GAN to generate a fake/ normal medical image from an input medical image, wherein a lesion area is removed from the input medical image. )
Regarding claims 9 and 15–16. The rationale provided for claims 1 and 7–8 is incorporated herein. In addition, the electronic device of claims 1 and 7–8 correspond to the method of claims 9 and 15–16. Therefore, the claims are all rejected.
Claim(s) 4, 6 and 12, 14 are rejected under 35 U.S.C. 103 as being unpatentable over Lapiere as modified by Lee, and further in view of Teixeira (Teixeira et al, US 10,849,585 B1, 2020).
Regarding claim 4, with deficiencies of Lapiere [as modified by Lee] noted in square brackets [ ], Lapiere [as modified by Lee] teaches the electronic device of claim 1, wherein the processor is configured to calculate a distance between pixels for each pixel by using coordinate values of corresponding pixels of the skin image and the generated image,
( [0008], [0036], [0072]: Lee teaches computing distance between pixels using coordinate values (a degree of change in pixels in each region, or pixel difference) between the input medical image and the fake medical image. )
and identify [ a predefined number of pixels as the ] at least one candidate biopsy site based on the distance between pixels.
( [0008], [0072]: Lee teaches identifying the lesion/ difference region by comparing the input/ original medical image with the generated fake/ normal medical image based on pixel change/ difference. )
As noted above, Lapiere [as modified by Lee] does NOT expressly discloses where Teixeira teaches: identify a predefined number of pixels based on the distance between pixels.
( [Page 9, column 10, line 60-65]: Teixeira teaches setting a predefined or user-selected threshold value, detecting groups of connected pixels above the threshold value, and detecting a region of connected pixels having a predefined difference between a generated image and an acquired image as an anomaly/candidate region. )
It would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to further modify Lapiere [as modified by Lee] with Teixeira’s predefined threshold / connected-pixel selection technique in order to automatically select pixels or pixel regions having sufficiently large differences between the original skin image and the generated lesion-removed image. Lapiere [as modified by Lee] already teaches comparing the original image and generated fake/ normal image based on pixel differences to identify a lesion/ difference region. Teixeira teaches a known way to make such difference-based detection more reliable by using a predefined or user-selected threshold and detecting connected pixels above the threshold as an anomaly/ candidate region. A person of ordinary skill in the art would have been motivated to apply Teixeira’s technique to reduce noise from insignificant pixel differences, isolate meaningful high-difference regions, and identify the candidate biopsy region more consistently. The modification is a predictable use of known medical-image difference-analysis techniques according to their established functions, with a reasonable expectation of success.
Regarding claim 6, Lapiere [as modified by Lee and Teixeira] teaches the electronic device of claim 4, wherein the processor is configured to use the skin image and the generated image by filtering with a Gaussian filter.
( [0060]: Lapiere teaches applying a Gaussian filter to smooth the image and reduce noise during skin-image processing.
In addition, Lee provides the generated fake/ normal image. Accordingly, Lapiere [as modified by Lee and Teixeira]’s Gaussian filtering is applied both to the skin image and the generated image in order to reduce noise and improve reliability of the candidate-region/ biopsy-site identification. )
Regarding claims 12 and 14, the rationale in the rejection of claims 4 and 6 is incorporated herein. In addition, the electronic device of claims 4 and 6 correspond to the method of claim 12 and 14. Therefore, the claims are all rejected.
Conclusion
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to KEN KUDO whose telephone number is (571)272-4498. The examiner can normally be reached M-F 8am - 5pm.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Vincent Rudolph can be reached at 571-272-8243. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
KEN KUDO
Examiner
Art Unit 2671
/KEN KUDO/Examiner, Art Unit 2671
/VINCENT RUDOLPH/Supervisory Patent Examiner, Art Unit 2671
Prosecution Insights