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 .
Restriction Requirement
In response to the Restriction Requirement set forth in the Office Action mailed December 22, 2025, Applicant elected, without traverse, the Group I invention, corresponding to claims 1-5.
Information Disclosure Statement
The information disclosure statements (IDSs) submitted on February 20, 2026 and February 24, 2026 are in compliance with 37 CFR 1.97 and 1.98 and therefore have been considered by the examiner and placed in the file.
Claim Interpretation
The claims in this application are given their broadest reasonable interpretation (BRI) using the plain meaning of the claim language in light of the specification as it would be understood by one of ordinary skill in the art. The BRI of a claim element (also commonly referred to as a claim limitation) is limited by the description in the specification.
In the following, some of the terms in the claims have been given BRIs in light of the specification. These BRIs are used for purposes of searching for prior art and examining the claims, but cannot be incorporated into the claims. Should Applicant believe that different interpretations are appropriate, Applicant should point to the portions of the specification that clearly support a different interpretation.
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.
Claims 1-3 and 5 are rejected under 35 U.S.C. 103 as being unpatentable over International Published Appl. No. WO 2020/219968 A1 to Jia et al. (hereinafter referred to as “Jia”) in view of U.S. Publ. Appl. No. 2021/0378505 A1 to Sobol et al. (hereinafter referred to as “Sobol”).
Regarding claim 1, Jia discloses a medical device for diabetic retinopathy (DR) identification (para. [0037]: “[i]n this disclosure, various deep learning solutions for segmenting avascular areas in OCTA of DR are disclosed”), the medical device comprising:
an optical coherence tomography (OCT) scanner configured to obtain a three-dimensional (3D) image of a retina, the 3D image comprising voxels, an example voxel among the voxels comprising a first value representing an OCT value of an example volume and a second value representing an OCTA value of the example volume (Fig. 1, para. [0073]: “[a]s illustrated in FIG. 1 , at least one clinical device 1 12 may transmit at least one diagnostic OCT image 114 to the prediction system 102. In some examples, the diagnostic OCT image(s) 1 14 can be captured by at least one OCT reflectance imaging device, at least one OCT angiography device, a combination thereof, or the like.” Fig. 13 shows the acquired OCT image 1302 and the acquired OCTA image 1306; the acquired OCT/OCTA images are 3D volumetric images, as disclosed in paras. [0044], [0061] and [0064], which means they comprise voxels);
at least one processor (Fig. 10, processor(s) 1106); and
memory (Fig. 11, memory 1104) storing instructions that, when executed by the at least one processor, cause the at least one processor to perform operations (para. [0110] discusses the operations performed by the processor(s) 1106 by executing instructions stored in memory 1104) comprising:
generating, by a convolutional neural network (CNN) and using the 3D image (Fig. 5A, the neural network 502 is a CNN having sub-CNNs 508-1 – 508-A that process the 3D OCT and OCTA images 506 to generate respective vectors that are then merged by merge block 510 to generate a merge vector. The sub-CNNs 508 have the configuration shown in Fig. 5B and described in paras. [0086]-[0087] and [0090]-[0095]);
generating, by a first model and using the vector, a first likelihood that the retina exhibits a first level of DR (Fig. 5A, sub-CNN 518-A+1 is a first model that receives the merge vector output from merge block 510 and performs segmentation/classification to generate an avascular map 504 with regions labeled to indicate the likelihood that they are vascular or avascular regions, as discussed in para. [0040]: “the neural network may output a probability map including multiple elements that respectively correspond to the likelihood that a given pixel in the particular OCT image depicts a particular area being segmented, such as an area of vascularity, an area of avascularity, or an area of signal reduction. In some cases, each element of the probability map includes three different probabilities, corresponding respectively to the probabilities that the corresponding pixel depicts a vascular area, an avascular area, or a signal reduction area.” The indications of the vascular and/or avascular regions convey the level of DR, as discussed in paras. [0114] and [0132] and shown in Fig. 17. For example, segmentation result D3 in Fig. 17 conveys the likelihood that the image exhibits a mild-to-moderate level of non-proliferative DR (NPDR));
generating, by a second model and using the vector, a second likelihood that the retina exhibits a second level of DR (In Jia, the same model, sub-CNN 518-A+1, is used to generate a second likelihood that the retina exhibits a second level of DR, i.e., the avascular map indicates whether the image exhibits a severe level of DR, as shown in Fig. 17, segmentation result D4); and
determining whether the retina exhibits an absence of DR, the first level of DR, or the second level of DR based on the first likelihood and the second likelihood (the segmentation/classification layers of sub-CNN 518-A+1 determine during segmentation of the vascular and avascular regions and generation of the avascular map whether the retina exhibits an absence of DR (Fig. 17, segmentation results D1 and/or D2), the first level of DR (Fig. 17, segmentation result D3), or the second level of DR (Fig. 17, segmentation result D4 and/or D5) based on the first and second likelihoods); and
a display (Fig. 11, output device(s) 1114 and Fig. 8 computer display of clinical device 800) configured to output an indication of whether the retina exhibits the absence of DR, the first level of DR, or the second level of DR (Fig. 8 shows overlayed image and avascular map indicative of whether the retina exhibits the absence of DR, the first level of DR, or the second level of DR. Fig. 17 shows displayed images with indications of an absence of DR (segmentation results D1 and/or D2 showing healthy), a first level of DR (segmentation result D3 showing mild to moderate DR) and a second level of DR (segmentation results D4 and D5 showing severe DR without and with signal reduction artifacts, respectively), as discussed in para. [0132]).
Jia does not explicitly disclose using first and second models for generating the first and second likelihoods, respectively. Sobol, in the same field of endeavor, discloses using multiple classifier models to classify an OCT image as containing a particular retinal disease by generating respective likelihoods (“confidence values” in Sobol) that the image exhibits the disease, where DR is given as an example of the retinal disease for which the image is being tested (Para. [0023] discusses examples of the retinal diseases including DR and para. [0067] discusses the classification of the disease being determined based on the likelihoods generated by the respective models: “[w]here multiple models are used by the OCT image classifier 154, the results of one model (e.g., a model having a highest confidence) can be used to classify the OCT image. Alternatively, the classification determined by two or more models can be combined (e.g., OCT image classifier 154 implemented as a prediction head with multiple trained machine learning model backbones) to classify the OCT image.”).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the present disclosure, to modify the system of Jia based on the teachings of Sobol to include a second sub-CNN similar or identical to sub-CNN 518-A+1 as a second model that receives the merge vector output from merge block 510 and performs segmentation/classification to generate an avascular map 504 with regions labeled to indicate a second likelihood that they are vascular or avascular regions and therefore exhibit the second level of DR. One of ordinary skill in the art would have been motivated to make the modification to improve accuracy of the system by allowing the segmentation/classification result having the highest confidence value to be used as the final DR level as taught by Sobol. The modification could have been made by one of ordinary skill in the art before the effective filing date of the present disclosure with a reasonable expectation of success because making the modification merely involves combining prior art elements according to known methods to yield predictable results (modifying the software executed by the processor(s) of Jia to implement a second sub-CNN that receives and processes the output of the merge block 510).
Regarding claim 2, Jia discloses processing, by multiple convolution blocks arranged in parallel, the 3D image, and wherein the multiple convolution blocks comprise at least one first convolution block with a stride of 1 and at least one second convolution block with a stride of 2 (Para. [0122] discusses, with reference to Fig. 14A, the configuration that can be used for the CNNs and sub-CNNs having at least one first convolution block with a stride of 1 and at least one second convolution block with a stride of 2).
Regarding claim 3, the BRI for the limitations of this claim, based on Fig. 12 and the corresponding description of the present disclosure, is that the first likelihood is generated by multiplying the vector that is output from the CNN by a weight matrix of an activation layer that introduces non-linearities, such as a Rectified Linear Unit (ReLU) layer, to produce an intermediary matrix that is processed with a first set of parameters to generate a second matrix and then applying a softmax operation on the second matrix. The second likelihood is generated in like manner, but using a second set of parameters.
Para. [0122] of Jia discloses, with reference to Fig. 14A, that for each of the CNNs and sub-CNNs, each convolution layer includes an ReLU activation stage that multiplies an ReLU matrix by the vector from the previous layer to produce an intermediary matrix that is processed with a first set of parameters corresponding to values of one or more filter matrices to produce a second matrix and that the output convolution block applies a softmax activation function that is applied to the second matrix to produce the avascular map corresponding to the first likelihood. Therefore, Jia discloses the limitations of claim 3 for generating the first likelihood.
As indicated above in the rejection of claim 1, Jia does not disclose using a second sub-CNN model for generating the second likelihood. For the reasons stated above in the rejection of claim 1, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the present disclosure, to modify the system of Jia based on the teachings of Sobol to include a second sub-CNN similar or identical to sub-CNN 518-A+1 but having a second set of parameters as a second model that receives the merge vector output from merge block 510 and performs segmentation/classification to generate an avascular map 504 with regions labeled to indicate a second likelihood that they are avascular regions and therefore exhibit the second level of DR. Since Jia discloses that all of the CNNs and sub-CNNs can have the configuration described therein with reference to Fig. 14A, it would have been obvious before the effective date of the present disclosure for the second sub-CNN to implement the ReLU and softmax operations described above to generate the second likelihood. A person of ordinary skill would have been motivated to make the modification for the reasons discussed above in the rejection of claim 1 and because Jia discloses that the CNNs and sub-CNNs can all have the same configuration. The modification could have been made by one of ordinary skill in the art before the effective filing date of the present disclosure with a reasonable expectation of success because making the modification merely involves combining prior art elements according to known methods to yield predictable results (modifying the software executed by the processor(s) of Jia to implement a second sub-CNN that receives and processes the output of the merge block 510 to generate an avascular map).
Regarding claim 5, Jia discloses that the operations performed by the processor(s) including training the CNN based on training data (Fig. 1, Trainer 104 of training neural network 110 using training data 200 shown in Fig. 2).
Claim 4 is rejected under 35 U.S.C. 103 as being unpatentable over Jia in view of Sobol as applied to claims 1-3 and 5 and further in view of U.S. Publ. Appl. No. 2024/0144478 A1 to Choi et al. (hereinafter referred to as “Choi”).
Regarding claim 4, Jia does not explicitly disclose generating, based on the 3D image, a CAM indicating at least one region in the 3D image that is indicative of DR, and wherein the display is further configured to output the CAM. Choi, in the same field of endeavor, discloses obtaining OCT/OCTA images (Paras. [0547]-[0549]), processing the images in a CNN (Para. [0233]) that generates different likelihoods that the image exhibits different levels of DR (Para. [0327]), and generates CAMs (Paras. [0336]-[0337], [0334]) that are displayed to a user (Para. [0437]).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the present disclosure, to modify the system of Jia as modified based on the teachings of Sobol further based on the teachings of Choi such that the sub-CNNs generate CAMs in addition to the avascular maps that the system causes to be displayed to the user. One of ordinary skill in the art would have been motivated to make the modification to provide clinicians with additional diagnostic information to assist them in diagnosing and prescribing treatment. The modification could have been made by one of ordinary skill in the art before the effective filing date of the present disclosure with a reasonable expectation of success because making the modification merely involves combining prior art elements according to known methods to yield predictable results (modifying the software executed by the processor(s) of Jia to implement the sub-CNNs to generate CAMs and cause them to be displayed to the user).
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
U.S. Publ. Appl. No. 2023/0337908 A1 to Nakazawa et al. discloses obtaining OCT and/or OCTA images and processing them in a CNN that generates likelihoods in the form of confidence values that are indicative to different levels of DR detected by the CNN (Para. [0049]).
Any inquiry concerning this communication or earlier communications from the examiner should be directed to DANIEL J SANTOS whose telephone number is (571)272-2867. The examiner can normally be reached M-F 9-5.
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, Matt Bella can be reached at (571)272-7778. 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.
/DANIEL J. SANTOS/Examiner, Art Unit 2667
/MATTHEW C BELLA/Supervisory Patent Examiner, Art Unit 2667
Prosecution Insights