Prosecution Insights
Last updated: November 04, 2026
Application No. 19/199,165

MEDICAL SPECTROSCOPY AND IMAGING ANALYSIS

Final Rejection §103
Filed
May 05, 2025
Priority
Feb 14, 2022 — provisional 63/310,014 +4 more
Examiner
LIMA, FABIO S
Art Unit
2486
Tech Center
2400 — Computer Networks
Assignee
Pathscience Inc.
OA Round
2 (Final)
77%
Grant Probability
Favorable
3-4
OA Rounds
9m
Est. Remaining
92%
With Interview

Examiner Intelligence

Grants 77% — above average
77%
Career Allowance Rate
345 granted / 446 resolved
+19.4% vs TC avg
Moderate +15% lift
Without
With
+14.6%
Interview Lift
resolved cases with interview
Typical timeline
2y 3m
Avg Prosecution
33 currently pending
Career history
490
Total Applications
across all art units

Statute-Specific Performance

§101
2.9%
-37.1% vs TC avg
§103
49.5%
+9.5% vs TC avg
§102
17.2%
-22.8% vs TC avg
§112
20.5%
-19.5% vs TC avg
Black line = Tech Center average estimate • Based on career data from 446 resolved cases

Office Action

§103
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 Arguments Applicant's arguments filed on July 20, 2026, have been fully considered but they are not persuasive. Applicant argues that the combination of Aidt, Bayramoglu, Ozcan and Zhang fail to teach, disclose or suggest all limitations of claim2. In particular, Applicant asserts that theses references fail to disclose “wherein the generative adversarial network is trained using a first co-registered image pair comprising an unstained image and a ground truth stained image of a same tissue, wherein a first training of the generative adversarial network generates a first digitally stained image, wherein a second co-registration is performed between the first digitally stained image and the ground truth stained image to generate a second co- registered image pair, and wherein a second training of the generative adversarial network is performed using the second co-registered image pair.” However, the Examiner submits that Zhang, previously applied against claims 5 and 21, discloses the newly recited limitations. Zhang discloses wherein the generative adversarial network is trained using a first co-registered image pair comprising an unstained image and a ground truth stained image of a same tissue (See pp. 9 and 10 disclosing co-registration between the fluorescence and brightfield images stained images began with global registration. Following this registration process, the neural network was trained using patches from these coarsely matched images. Using this network, the autofluorescence images were then virtually stained) wherein a first training of the generative adversarial network generates a first digitally stained image, wherein a second co-registration is performed between the first digitally stained image and the ground truth stained image to generate a second co- registered image pair (See p.10 disclosing elastic pyramidal co-registration was then applied to match the histochemically stained images with the initial virtually stained images, resulting in a matched image pair), and wherein a second training of the generative adversarial network is performed using the second co-registered image pair (See p. 10 disclosing after the refined co-registration, the accurately aligned FOVs are partitioned into patches and the patches were then used to train the GAN. During GAN training, the generator learns to perform a statistical transformation to generate a virtually stained image, while the discriminator attempts to distinguish between histochemically stained images and their virtually stained counterparts) Accordingly, the claims remain rejected for the reasons set forth above. Claim Rejections - 35 USC § 103 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. Claims 2-21 are rejected under 35 U.S.C. 103 as being unpatentable over Aidt et al. (US20240029409A1), hereinafter referred to as Aidt, in view of Bayramoglu et al. ("Towards Virtual H&E Staining of Hyperspectral Lung Histology Images Using Conditional Generative Adversarial Networks," 2017 IEEE International Conference on Computer Vision Workshops (ICCVW), Venice, Italy, 2017, pp. 64-71, doi: 10.1109/ICCVW.2017.15.), hereinafter referred to as Bayramoglu further, in view of Ozcan (US 20210043331 A1), hereinafter referred to as Ozcan, and Zhang et. (Digital synthesis of histological stains using micro-structured and multiplexed virtual staining of label-free tissue. Retrieved from https://par.nsf.gov/biblio/10179788. Light: Science & Applications 9.1 Web. doi:10.1038/s41377-020-0315-y.( Published: 06 May 2020)), hereinafter referred to as Zhang. Regarding claim 2, Aidt discloses computer-implemented method for analyzing and staining tissue (See ¶[0015] disclosing a computer implemented method ), the computer-implemented method comprising: receiving, by a computer processor, an unstained electronic digital image of a tissue sample captured by an electronic image capture device (See ¶¶[0242], [0256], [0258] and [0261] disclosing that a first image of a sample tissue may not be stained, wherein the image is obtained from an image sensor or from a scanner that captures images and sent to analysis after capture); dividing, by the computer processor, the compressed image representation into a plurality of subfields (See ¶[0259], [0742] and [0820] disclosing dividing the tissue image into a plurality of patches) processing, by the computer processor, each subfield of the plurality of subfields of the compressed image representation through an artificial intelligence model (See ¶[0709] and FIGs 4a-4c disclosing that the training process takes this whole multitude of patches-sets and trains an (AI) model with them) to identify one or more features in the unstained digital image, wherein the one or more features comprise at least one of a composition, a shape, a lipid, a nucleus, or a carbohydrate, and wherein the artificial intelligence model is trained to determine that features observed in the unstained digital image are indicative of one or more medical conditions (See [0242], [0244], [0254]-[0255] and [0256] disclosing identifying features in the image (e.g. nucleus, shape and composition) and using these features to determine information indicative of one or more medical conditions); analyzing, by the computer processor, the identified one or more features to determine one or more regions with identified features that correspond to an abnormality in the tissue sample (See ¶¶[0348], [0408]-[0409], and [0418 disclosing analyzing identified features to determine regions in the mage corresponding to abnormality in the tissue sample ); generating, by the computer processor, a feature map of the one or more identified features and the one or more regions indicating abnormalities, wherein the feature map comprises one or more indicators that indicate regions of interest in the unstained digital image, and wherein the feature map indicates the one or more medical conditions associated with the identified abnormalities (See ¶¶[0244], [0348], [0383] and [0409] disclosing generating a featured map comprising indicators of regions of interest and corresponding to medical condition information); generating, by the computer processor, a digitally stained image using a generative adversarial network (See ¶¶[0033], [0274],[0323] and [0420] disclosing a virtual stained machine learning model that generates a synthetic and/or virtual stainer image from an input image, wherein the virtual stainer may be implemented using GAN framework and that the generated image is trained using a loss function comparing the predicted virtually stained image patch to a corresponding target patch) outputting an indication of the one or more regions with identified features that correspond to the abnormality in the tissue sample (See ¶¶[0244],[0348], [0385] and [0409] disclosing that identified features such as abnormal cell and tumors are detected within the image and that corresponding outputs are provided in form of segmentation, markers, overlays, mapped locations, and labels tied to specific regions of the image and associating such outputs with diagnostic applications); and displaying, by the computer processor, on a display device the digitally stained image and the feature map (See ¶¶[0295] and [0316] disclosing presenting the input and processed images, including virtual image (digitally stained), on a display device; See also [0348]). Aidt does not explicitly disclose generating, by the computer processor using raw spectral data from the unstained electronic digital image, a compressed image representation, wherein the unstained electronic image comprises spatial information, and wherein the spatial information is preserved in the compressed image representation; wherein at least one subfield of the plurality of subfields overlaps with another subfield of the plurality of subfields; wherein a generator model of the generative adversarial network learns mapping from the compressed image data to perform digital staining; and wherein a discriminator network of the generative adversarial network learns to discriminate between the digitally stained image and a ground truth image; and wherein the generative adversarial network is trained using a first co-registered image pair comprising an unstained image and a ground truth stained image of a same tissue, wherein a first training of the generative adversarial network generates a first digitally stained image, wherein a second co-registration is performed between the first digitally stained image and the ground truth stained image to generate a second co- registered image pair, and wherein a second training of the generative adversarial network is performed using the second co-registered image pair. However, Bayramoglu from the same or similar endeavor of image processing discloses generating, by the computer processor using raw spectral data from the unstained electronic digital image, a compressed image representation, wherein the unstained electronic image comprises spatial information, and wherein the spatial information is preserved in the compressed image representation (See section 4 disclosing that the hyperspectral image cubes are composed of 133 bands. We reduce the dimension of our hypercubes to 3 via PCA transform); wherein at least one subfield of the plurality of subfields overlaps with another subfield of the plurality of subfields (See section 4 disclosing that the registered HSI and H&E image pairs are divided into overlapping patches of 64 × 64 pixels with a stride of 16 pixels) wherein a generator model of the generative adversarial network learns mapping from the compressed image data to perform digital staining (See section 3.3 disclosing a mapping from reduced HSI to H&E stained image using conditional generative adversarial network(cGAN) which is an extension of generative adversarial networks (GANs)). It would have been obvious to the person of ordinary skill in the art before the effective filing date of the claimed invention to modify the teachings disclosed by Aidt to add the teachings of Bayramoglu as above, in order to automate some of the tasks in the diagnostic pathology workflow which could be used to speed up the sample processing time, reduce costs, prevent adverse effects of chemical stains on tissue specimens, reduce observer variability, and increase objectivity in disease diagnosis (Bayramoglu, Abstract). Furthermore, Ozcan from the same or similar endeavor of image processing discloses wherein a discriminator network of the generative adversarial network learns to discriminate between the digitally stained image and a ground truth image (See ¶[0068] and claims 5 and 36 disclosing the GAN training procedure involves training of two different networks: a deep neural network and a discriminator network which is configured to discriminate between a ground truth chemically stained image of the sample and the outputted digitally stained microscopic image of the sample) It would have been obvious to the person of ordinary skill in the art before the effective filing date of the claimed invention to modify the teachings disclosed by Aidt and Bayramoglu to add the teachings of Ozcan as above, in order to minimize the output loss, while maximizing the probability of correctly classifying the real label (Ozcan, [0070]). Moreover, Zhang from the same or similar endeavor of image processing discloses wherein the generative adversarial network is trained using a first co-registered image pair comprising an unstained image and a ground truth stained image of a same tissue (See pp. 9 and 10 disclosing co-registration between the fluorescence and brightfield images stained images began with global registration. Following this registration process, the neural network was trained using patches from these coarsely matched images. Using this network, the autofluorescence images were then virtually stained) wherein a first training of the generative adversarial network generates a first digitally stained image, wherein a second co-registration is performed between the first digitally stained image and the ground truth stained image to generate a second co- registered image pair (See p.10 disclosing elastic pyramidal co-registration was then applied to match the histochemically stained images with the initial virtually stained images, resulting in a matched image pair), and wherein a second training of the generative adversarial network is performed using the second co-registered image pair (See p. 10 disclosing after the refined co-registration, the accurately aligned FOVs are partitioned into patches and the patches were then used to train the GAN. During GAN training, the generator learns to perform a statistical transformation to generate a virtually stained image, while the discriminator attempts to distinguish between histochemically stained images and their virtually stained counterparts) It would have been obvious to the person of ordinary skill in the art before the effective filing date of the claimed invention to modify the teachings disclosed by Aidt, Bayramoglu and Ozcan to add the teachings of Zhang as above, in order to improve the accuracy and consistency of tissue-based diagnoses and allow pathologists to obtain more relevant information from tissue than is otherwise possible (Zhang, p. 7). Regarding claim 3, Aidt, Bayramoglu, Ozcan and Zhang disclose all the limitations of claim 2, and is analyzed as previously discussed with respect to that claim. Furthermore, Aidt discloses the computer-implemented method of claim 2, further comprising: receiving, by the computer processor, an indication of a condition of interest; and determining, by the computer processor based on the condition of interest, at least one transformation to apply to the unstained electronic digital image (See ¶¶[0254]-[0255], [0409] and [0420] disclosing determine features indicative of disease ( condition of interest) and applying a transformation ( virtual staining) based on the input image and learned model). Regarding claim 4, Aidt, Bayramoglu, Ozcan and Zhang disclose all the limitations of claim 2, and is analyzed as previously discussed with respect to that claim. Furthermore, Aidt discloses the computer-implemented method of claim 2, further comprising: determining, by the computer processor based on the identified one or more features, one or more additional transformations to apply to the unstained electronic digital image; and recommending, by the computer processor, the one or more additional transformations to a user (See ¶[0244] disclosing that “records of the synthetic training dataset may be further labeled with a ground truth indication of a diagnosis;” and ¶[0385] disclosing “a diagnosis for the sample may be obtained.” Therefore, Aidt discloses generating diagnostic output based on identified futures. A person of ordinary skill in the art would understand that such diagnostic outputs inherently guide further analysis or transformations, making it obvious to generate a recommendation for additional transformations based on identified features). Regarding claim 5, Aidt, Bayramoglu, Ozcan and Zhang disclose all the limitations of claim 2, and is analyzed as previously discussed with respect to that claim. Aidt does not explicitly disclose the computer-implemented method of claim 2, further comprising: receiving a selection of a plurality of regions of interest in the unstained electronic digital image ;receiving a stain selection for each region of interest of the plurality of regions of interest ;applying a digital stain to each region of interest based on the stain selection for the region of interest; and combining the digitally stained regions of interest to form a composite image. However, Zhang from the same or similar endeavor of image processing discloses the computer-implemented method of claim 2, further comprising: receiving a selection of a plurality of regions of interest in the unstained electronic digital image (See p. 12 disclosing receiving and defining multiple ROIs in the tissue image, including user defined selection) ;receiving a stain selection for each region of interest of the plurality of regions of interest (See pp.6 and 12 disclosing that each ROIs is assigned a stain via the digital staining matrix. (i.e. a per region stain selection);applying a digital stain to each region of interest based on the stain selection for the region of interest (See pp. 2 and 7 (FIG. 3) disclosing applying selected stain to specific regions based on the user defined map); and combining the digitally stained regions of interest to form a composite image (See Abstract and p.2 disclosing merging and blending multiple stains into one output image, i.e., a composite image with region specific stains) It would have been obvious to the person of ordinary skill in the art before the effective filing date of the claimed invention to modify the teachings disclosed by Aidt and Bayramoglu to add the teachings of Zhang as above, in order to improve the accuracy and consistency of tissue-based diagnoses and allow pathologists to obtain more relevant information from tissue than is otherwise possible (Zhang, p. 7). Regarding claim 6, Aidt, Bayramoglu, Ozcan and Zhang disclose all the limitations of claim 2, and is analyzed as previously discussed with respect to that claim. Aidt does not explicitly disclose the computer-implemented method of claim 2, wherein the processing is performed in substantially real-time as the unstained electronic digital image is being captured However, Ozcan from the same or similar endeavor of image processing discloses the computer-implemented method of claim 2, wherein the processing is performed in substantially real-time as the unstained electronic digital image is being captured (See ¶¶ [0007][, [0054] and claim 23 disclosing generating the digitally stained image substantially in real time during the capture of the unstained image ). The motivation for combining Aidt, Bayramoglu, Ozcan and Zhang has been discussed in connection with claim 2, above. Regarding claim 7, Aidt, Bayramoglu, Ozcan and Zhang disclose all the limitations of claim 2, and is analyzed as previously discussed with respect to that claim. Aidt does not explicitly disclose the computer-implemented method of claim 2, wherein the unstained electronic digital image comprises a hyperspectral image. However, Bayramoglu from the same or similar endeavor of image processing discloses the computer-implemented method of claim 2, wherein the unstained electronic digital image comprises a hyperspectral image. (See section 1 disclosing performing digital H&E staining on unstained hyperspectral microscopy image data, on a proof of concept level; and section 4 disclosing hyperspectral image cubes are composed of 133 bands.). The motivation for combining Aidt, Bayramoglu, Ozcan and Zhang has been discussed in connection with claim 2, above. Regarding claims 8-14 and 15-21, these claims are rejected based on the same art and evidentiary limitations applied to claims 2-7, since they claim analogous subject matter for performing the same or equivalent functionality. 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 extension fee 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 FABIO S LIMA whose telephone number is (571)270-0625. The examiner can normally be reached on Monday - Friday 8 am - 4 pm. 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, Jamie Atala can be reached on (571) 272-7384. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300. Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000. /FABIO S LIMA/Primary Examiner, Art Unit 2486
Read full office action

Prosecution Timeline

May 05, 2025
Application Filed
Apr 03, 2026
Response after Non-Final Action
Apr 20, 2026
Non-Final Rejection mailed — §103
Jul 20, 2026
Response Filed
Sep 16, 2026
Final Rejection mailed — §103 (current)

Precedent Cases

Applications granted by this same examiner with similar technology

Patent 12786474
ADAPTIVE PICTURE MODIFICATIONS FOR VIDEO CODING
2y 1m to grant Granted Nov 03, 2026
Patent 12786475
AI TEXT FOR VIDEO STREAMS
2y 1m to grant Granted Nov 03, 2026
Patent 12777190
THREE-DIMENSIONAL DATA ENCODING METHOD, THREE-DIMENSIONAL DATA DECODING METHOD, THREE-DIMENSIONAL DATA ENCODING DEVICE, AND THREE-DIMENSIONAL DATA DECODING DEVICE
3y 3m to grant Granted Oct 27, 2026
Patent 12764310
CROSS-COMPONENT QUANTIZATION IN VIDEO CODING
2y 1m to grant Granted Oct 13, 2026
Patent 12756864
POINT CLOUD DATA TRANSMISSION DEVICE, POINT CLOUD DATA TRANSMISSION METHOD, POINT CLOUD DATA RECEPTION DEVICE AND POINT CLOUD DATA RECEPTION METHOD
3y 1m to grant Granted Oct 06, 2026
Study what changed to get past this examiner. Based on 5 most recent grants.

Strategy Recommendation AI-generated — please review before filing

Get a prosecution strategy drawn from examiner precedents, rejection analysis, and claim mapping.
Typically takes 5-10 seconds — AI-generated, attorney review required before filing

Prosecution Projections

3-4
Expected OA Rounds
77%
Grant Probability
92%
With Interview (+14.6%)
2y 3m (~9m remaining)
Median Time to Grant
Moderate
PTA Risk
Based on 446 resolved cases by this examiner. Grant probability derived from career allowance rate.

Sign in with your work email

Enter your email to receive a magic link. No password needed.

Personal email addresses (Gmail, Yahoo, etc.) are not accepted.

Free tier: 3 strategy analyses per month