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 .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 7/9/2026 has been entered.
Amendments
Applicant’s Amendment filed on 7/9/2026 has been entered and made of record.
Currently Pending claims: 1, 2, and 4-17
Independent claims: 1
Amended claims: 1
Response to Arguments
This office action is responsive to Applicant’s Arguments/Remarks Made in an Amendment received on 7/9/2026.
Applicant’s arguments, see pages 5-7, filed 7/9/2026, with respect to the rejection of claims 1, 5, 8-10, and 12-17 under 35 U.S.C. § 102 have been fully considered but are moot because a new ground of rejection, as necessitated by amendment, is made of Cruz et al. and Alarcon et al.
Applicant’s arguments, see pages 5-7, filed 7/9/2026, with respect to the rejection of claims 2, 4, 6, 7, and 11 under 35 U.S.C. § 103 have been fully considered but are moot as they depend on arguments related to the rejection of claim 1 under 35 U.S.C. § 102.
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.
Claims 1, 4, 5, 8-11, and 12-17 are rejected under 35 U.S.C. 103 as being unpatentable over Cruz et al. (Garcia Santa Cruz, Beatriz, et al. "Generalising from conventional pipelines using deep learning in high-throughput screening workflows." Scientific Reports 12.1 (2022): 11465) (hereafter, "Cruz") in view of Alarcon et al. (Alarcon, Mauro Lemus, et al. "Remote Instrumentation Science Environment for Intelligent Image Analytics." 2022 IEEE 18th International Conference on e-Science (e-Science), IEEE, 2022) (hereafter, “Alarcon”).
Regarding claim 1, Cruz discloses a method for training a machine-learning model for use in a scientific or surgical imaging system, the method comprising: obtaining a plurality of images of a scientific or surgical imaging system, for use as training input images (Page 3 section titled Methods, subsection on page 4 titled hiPSC generation and imaging acquisition, lines 4-6, Confocal images were obtained with an Opera QEHS spinning disk microscope (Perkin Elmer) under a 60x water immersion objective); obtaining a plurality of training outputs that are generated by processing the training input images by an image processing workflow of the scientific or surgical imaging system (Page 3 Figure 1(a), Steps 1.1 Weakly labeled masks; See page 6 section titled Results, lines 1-2, the generalisation capabilities of a DL network trained on noisy label data generated by a CIP pipeline), the image processing workflow comprising a plurality of image processing steps (Page 3 Figure 1(a), Steps 1.1-1.2, captions [1.1] First a weakly labelled dataset is created using conventional imaging processing (CIP). [1.2] After that, a U-net like architecture is trained); evaluating an output of the machine-learning model (Page 3 Figure 1(a), steps 1.3, caption [1.3] the accuracy of the [predicted mask] evaluated; See page 3 section titled Methods, subsection on page 3-4 titled Pipeline setup, lines 8-9, Using the GUI tool, the experts can also correct the potential inaccuracies of the images); training the machine-learning model using the plurality of training input images and the plurality of training outputs (Page 3 Figure 1(a), Steps 1.1-1.2, captions [1.1] First a weakly labelled dataset is created using conventional imaging processing (CIP). [1.2] After that, a U-net like architecture is trained); evaluating an output of the machine-learning model (Page 3 Figure 1(a), steps 1.3, caption [1.3] the accuracy of the [predicted mask] evaluated; See page 3 section titled Methods, subsection on page 3-4 titled Pipeline setup, lines 8-9, Using the GUI tool, the experts can also correct the potential inaccuracies of the images) according to a quality criterion (Page 3 section titled Methods, page 4 subsection titled Part A: Measuring the DL generalization…, page 5 sub-subsection titled Evaluation, lines 6-8, (1) a qualitative analysis using blind expert ranking. (2) a quantitative analysis using Bounding Boxes (BB) as a surrogate metric. (3) a qualitative analysis using the overlapping segmentation using dice-coefficient employing manually corrected samples); and [providing a feedback signal] for adapting one or more parameters of the image processing workflow (Page 4, §CIP Pipeline, The CIP pipeline was implemented in MATLAB and was specifically designed and fine-tuned for the case study of Autophagy. Examiner is providing this citation to show that Cruz discloses adjustment of the CIP pipeline. The limitation as a whole is disclosed by the combination of Cruz and Alarcon) [based on the evaluation of the output of the machine-learning model such that the image processing workflow regenerates the plurality of] training outputs (Page 3 Figure 1(a), Steps 1.1-1.2, captions [1.1] First a weakly labelled dataset is created using conventional imaging processing (CIP)) [based on the adapted parameters].
However, Cruz fails to explicitly disclose providing a feedback signal for adapting one or more parameters of the image processing workflow based on the evaluation of the output of the machine-learning model such that the image processing workflow regenerates the plurality of training outputs based on the adapted parameters.
Alarcon teaches providing a feedback signal for adapting one or more parameters of the image processing workflow based on the evaluation of the output of the machine-learning model such that the image processing workflow regenerates the plurality of training outputs based on the adapted parameters (Fig. 3; Fig. 5; Page 87, §Workflow Automation Requirement, control experimental settings; Page 88, left column, second paragraph, a request for new data is submitted indicating the adjustments needed to improve the characteristics of the images; Page 88, right column, first paragraph, use the feedback command to adjust the microscope parameters; Page 90, left column, third paragraph, determine the proper settings that need to be sent to the SEM to adjust for getting the expected results in the upcoming iterations of the experiment; Page 90, §Risk Management…, The AI recommendations on how to adjust the instrument for the next iteration. Figs. 3 and 5 illustrate processes where a machine learning model is used to iteratively adjust parameters for a scientific process. Each iteration generates new images and thereby training outputs for the machine learning system. Alarcon shows that the parameters can be changed at different locations of the process and Cruz is relied upon for changing parameters in the image processing workflow).
Both Cruz and Alarcon are analogous to the claimed invention because they are directed at replacing manual steps of the scientific workflow with machine learning models. It would have been obvious to a person of ordinary skill before the effective filing date of the claimed invention to incorporate the iterative parameter updates of Alarcon into the image processing pipeline of Cruz. The suggestion/motivation for doing so would have been to improve the base device of Cruz (image analysis pipeline replacement with machine learning models) with the iterative method of Alarcon (iteratively changing parameters with newly generated images). One of ordinary skill in the art would have been capable of applying the improvement of Alarcon with predictable results, as it would only involve changing the parameters of a different step of the image workflow process. A person of ordinary skill would be further motivated to incorporate Alarcon into Cruz for the purpose of improving efficiency, as suggested by Alarcon at Page 86, right column, first paragraph, saving precious time of researchers/instruments, and maximizing workflow outputs.
This method of improving Cruz was within the ordinary ability of one of ordinary skill in the art based on the teachings of Alarcon.
Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date, to modify Cruz with the teachings of Alarcon to obtain the invention as specified in claim 1.
Regarding claim 4, in which claim 1 is incorporated, Cruz discloses [wherein the machine-learning model is trained to generate the feedback signal for adapting one or more parameters of the] image processing (Page 4, §CIP Pipeline, The CIP pipeline) [workflow].
However, Cruz fails to explicitly disclose wherein the machine-learning model is trained to generate the feedback signal for adapting one or more parameters of the image processing workflow.
Alarcon teaches wherein the machine-learning model is trained to generate the feedback signal for adapting one or more parameters of the workflow (Page 88, right column, first paragraph, generate a feedback command, 7) use the feedback command to adjust the microscope parameters).
Both Cruz and Alarcon are analogous to the claimed invention because they are directed at replacing manual steps of the scientific workflow with machine learning models. It would have been obvious to a person of ordinary skill before the effective filing date of the claimed invention to incorporate the iterative parameter updates of Alarcon into the image processing pipeline of Cruz. The suggestion/motivation for doing so would have been to improve the base device of Cruz (image analysis pipeline replacement with machine learning models) with the iterative method of Alarcon (iteratively changing parameters with newly generated images). One of ordinary skill in the art would have been capable of applying the improvement of Alarcon with predictable results, as it would only involve changing the parameters of a different step of the image workflow process. A person of ordinary skill would be further motivated to incorporate Alarcon into Cruz for the purpose of improving efficiency, as suggested by Alarcon at Page 86, right column, first paragraph, saving precious time of researchers/instruments, and maximizing workflow outputs.
This method of improving Cruz was within the ordinary ability of one of ordinary skill in the art based on the teachings of Alarcon.
Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date, to modify Cruz with the teachings of Alarcon to obtain the invention as specified in claim 4.
Regarding claim 5, in which claim 1 is incorporated, Cruz discloses the method according to claim 1, wherein the machine-learning model is trained, using supervised learning (Page 3, Figure 1(a) captions, [1.2] After that, a U-net like architecture is trained. Here U-net is trained via supervised learning using the input images and the weakly labeled masks), to transform an image of the scientific or surgical imaging system into an output, by applying the plurality of training input images at an input of the machine-learning model (See page 3 Figure 1(a), Steps 1.1-1.2, captions [1.1] First a weakly labelled dataset is created using conventional imaging processing (CIP)) and using the plurality of training outputs as desired output during training of the machine-learning model (Page 3, §Pipeline setup. Labels are automatically generated using CIP techniques; (2) These masks are employed to conduct a supervised train using CNN. Examiner considers the labels and masks as “training outputs”).
Regarding claim 8, in which claim 1 is incorporated, Cruz discloses a method for training a machine-learning model (See page 4, subsection training, The network was trained in MATLAB) for use in a scientific or surgical imaging system (Page 3, first paragraph, This dataset includes imaging samples of the autophagy pathway using the Rosella pH-sensitive biosensor using human iPS cells), the method comprising: generating a plurality of images based on imaging sensor data of an optical imaging sensor of the scientific or surgical imaging system (See page 3 section titled Methods, subsection on page 4 titled hiPSC generation and imaging acquisition, lines 4-6, Confocal images were obtained with an Opera QEHS spinning disk microscope (Perkin Elmer) under a 60x water immersion objective); generating, using an image processing workflow of the scientific or surgical imaging system, a plurality of outputs based on the plurality of images (See page 3 Figure 1(a), Steps 1.1 Weakly labeled masks; See page 6 section titled Results, lines 1-2, the generalisation capabilities of a DL network trained on noisy label data generated by a CIP pipeline), the image processing workflow comprising a plurality of image processing steps (See page 3 section titled Methods, subsection on page 4 titled CIP pipeline, lines 4-6, Examples of techniques used as building blocks to create such pipelines are: image deconvolution, thresholding Gaussian filtering, Top-hat filtering, watershed transformation, difference of Gaussians, Butterworth filter or high pass filter); and providing the plurality of images as training input images and the plurality of outputs as training outputs for training a machine-learning model according (See page 3 Figure 1(a), Steps 1.1-1.2, captions [1.1] First a weakly labelled dataset is created using conventional imaging processing (CIP). [1.2] After that, a U-net like architecture is trained) to the method of claim 1.
Regarding claim 9, in which claim 1 is incorporated, Cruz discloses the method according to claim 8, further comprising obtaining the trained machine-learning model and replacing the image processing workflow with the machine-learning model (See page 3 section titled Methods, page 4 subsection titled Part A: Measuring…, page 4-5 sub-subsection titled User-friendly GUI, lines 1-5, Next, the CNN was integrated into a user-friendly tool using the MATLAB Image Label App. This integration allows easy handling for the potential users, biological researchers that often do not have experience with programming. By integrating the CNN as a segmentation algorithm into the image label app the tool does not only allow for an easy prediction of the mask using the CNN solution but also an intuitive way to correct the errors of the mask using manual segmentation tools) that is trained according to the method of claim 1.
Regarding claim 10, in which claim 1 is incorporated, Cruz discloses the method according to claim 8, further comprising training the machine-learning model using the method of claim 1 (See page 3 Figure 1(a), Step 1.2, caption [1.2] After that, a U-net like architecture is trained).
Regarding claim 11, in which claim 8 is incorporated, Cruz discloses [further comprising obtaining a feedback signal, the feedback signal being based on the training of the machine-learning model or based on an output of the trained machine-learning model when the machine-learning model is used by the surgical or scientific imaging system, using the feedback signal as input to the] image processing (Page 4, §CIP Pipeline, The CIP pipeline) [workflow or to the trained machine-learning model].
However, Cruz fails to explicitly disclose further comprising obtaining a feedback signal, the feedback signal being based on the training of the machine-learning model or based on an output of the trained machine-learning model when the machine-learning model is used by the surgical or scientific imaging system, using the feedback signal as input to the image processing workflow or to the trained machine-learning model.
Alarcon teaches further comprising obtaining a feedback signal (Page 88, right column, first paragraph, feedback command), the feedback signal being based on the training of the machine-learning model or based on an output of the trained machine-learning model when the machine-learning model is used by the surgical or scientific imaging system (Page 88, right column, first paragraph, use a second AI-based model to process these outcome measures to generate a feedback command. Since the limitation element is recited in the alternative, Examiner considers this citation to fully disclose the limitation element), using the feedback signal as input to the workflow (Page 88, right column, first paragraph, use the feedback command to adjust the microscope parameters. Since the limitation element is recited in the alternative, Examiner considers this citation to fully disclose the limitation element) or to the trained machine-learning model.
Both Cruz and Alarcon are analogous to the claimed invention because they are directed at replacing manual steps of the scientific workflow with machine learning models. It would have been obvious to a person of ordinary skill before the effective filing date of the claimed invention to incorporate the iterative parameter updates of Alarcon into the image processing pipeline of Cruz. The suggestion/motivation for doing so would have been to improve the base device of Cruz (image analysis pipeline replacement with machine learning models) with the iterative method of Alarcon (iteratively changing parameters with newly generated images). One of ordinary skill in the art would have been capable of applying the improvement of Alarcon with predictable results, as it would only involve changing the parameters of a different step of the image workflow process. A person of ordinary skill would be further motivated to incorporate Alarcon into Cruz for the purpose of improving efficiency, as suggested by Alarcon at Page 86, right column, first paragraph, saving precious time of researchers/instruments, and maximizing workflow outputs.
This method of improving Cruz was within the ordinary ability of one of ordinary skill in the art based on the teachings of Alarcon.
Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date, to modify Cruz with the teachings of Alarcon to obtain the invention as specified in claim 11.
Regarding claim 12, in which claim 1 is incorporated, Cruz discloses the method according to claim 8, wherein the image processing workflow comprises at least one of one or more deterministic image processing steps (See page 3 section titled Methods, subsection on page 4 titled CIP pipeline, lines 4-6, Examples of techniques used as building blocks to create such pipelines are: image deconvolution, thresholding Gaussian filtering, Top-hat filtering, watershed transformation, difference of Gaussians, Butterworth filter or high pass filter. The claim only requires “at least one of”), one or more image processing steps with an iterative optimization component, or one or more machine-Learning-based image processing steps.
Regarding claim 13, in which claim 1 is incorporated, Cruz discloses a system comprising one or more processors and one or more storage devices, wherein the system is configured to perform the method of claim 1 (See page 3 section titled Methods, page 4 subsection titled Part A: Measuring…, page 4 sub-subsection titled Training, line 1, The network was trained in MATLAB. As the model was trained on a computer program running on a general-purpose computer, it must have had a processor and storage).
Regarding claim 14, in which claim 1 is incorporated, Cruz discloses a system comprising one or more processors and one or more storage devices, wherein the system is configured to perform the method of claim 8 (See page 3 section titled Methods, page 4 subsection titled Part A: Measuring…, page 4 sub-subsection titled Training, line 1, The network was trained in MATLAB. As the model was trained on a computer program running on a general-purpose computer, it must have had a processor and storage).
Regarding claim 15, in which claim 1 is incorporated, Cruz discloses a system comprising one or more processors and one or more storage devices (See page 3 section titled Methods, page 4 subsection titled Part A: Measuring…, page 4 sub-subsection titled Training, line 1, The network was trained in MATLAB. As the model was trained on a computer program running on a general-purpose computer, it must have had a processor and storage), wherein the system is configured to obtain an image based on imaging sensor data of an optical imaging sensor of the scientific or surgical imaging system (See page 3 section titled Methods, subsection on page 4 titled hiPSC generation and imaging acquisition, lines 4-6, Confocal images were obtained with an Opera QEHS spinning disk microscope (Perkin Elmer) under a 60x water immersion objective); process the image using a machine-learning model that is trained according to the method of claim 1 (See page 3 section titled Methods, page 4 subsection titled Part A: Measuring…, page 5 sub-subsection titled Manual correction, lines 1-2, Instead, the images were firstly predicted with our trained network) to generate an output of the machine-learning model (See page 3 section Methods, page 4 subsection Part A: Measuring…, page 5 sub-subsection Manual correction, the images were firstly predicted with our trained network and then manually corrected using the tools available in the GUI. Examiner considers using outputs in the GUI to imply generating an output).
Regarding claim 16, in which claim 1 is incorporated, Cruz discloses a non-transitory computer-readable storage medium including a program code configured to perform (page 12 section titled Data and code availability, line 1, Code and data will be available thought the R3 platform of the University of Luxembourg. As the data and code are available for access, there must be a non-transitory CRM), when executed by a processor, the method according to claim 1.
Regarding claim 17, in which claim 1 is incorporated, Cruz discloses a non-transitory computer-readable storage medium including a program code configured to perform, when executed by a processor (page 12 section titled Data and code availability, line 1, Code and data will be available thought the R3 platform of the University of Luxembourg. As the data and code are available for access, there must be a non-transitory CRM), the method according to claim 8.
Claim 2 is rejected under 35 U.S.C. 103 as being unpatentable over Cruz et al. (Garcia Santa Cruz, Beatriz, et al. "Generalising from conventional pipelines using deep learning in high-throughput screening workflows." Scientific Reports 12.1 (2022): 11465) (hereafter, "Cruz") in view of Alarcon et al. (Alarcon, Mauro Lemus, et al. "Remote Instrumentation Science Environment for Intelligent Image Analytics." 2022 IEEE 18th International Conference on e-Science (e-Science), IEEE, 2022) (hereafter, “Alarcon”), as applied to claim 1 above, and further in view of Sharma (US 2020/0104994).
Regarding claim 2, Cruz in view of Alarcon discloses the method according to claim 1.
However, neither Cruz nor Alarcon, whether considered individually or in combination, explicitly disclose further comprising obtaining the one or more input parameters of the image processing workflow as further training input and training the machine-learning model using the one or more input parameters as further training input.
Sharma teaches further comprising obtaining the one or more input parameters of the image processing workflow as further training input and training the machine-learning model using the one or more input parameters as further training input (Fig 2, #206; ¶0038, lines 1-5, At step 206, the machine learning model is trained based on the training images, input parameters, output interpretations determined for the training images, and ground truth interpretations associated with the training images).
Cruz, Alarcon, and Sharma are analogous to the claimed invention because Cruz and Alarcon are directed at replacing manual steps of the scientific workflow with machine learning models and Sharma is in the field of medical image processing. It would have been obvious to a person of ordinary skill before the effective filing date of the claimed invention to incorporate the use of input parameters in training the machine learning model from Sharma into the iterative parameter updating of Alarcon and the machine learning based image processing workflow of Cruz. The suggestion/motivation for doing so would have been to optimize the relationship between pre-processing parameters and machine learning models (¶0038, lines 33-35, To learn which pre-processing algorithms and settings are best for various possible downstream AI image-analysis algorithms).
This method of improving Cruz was within the ordinary ability of one of ordinary skill in the art based on the teachings of Alarcon and Sharma.
Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date, to modify Cruz with the teachings of Alarcon and Sharma to obtain the invention as specified in claim 2.
Claim 6 is rejected under 35 U.S.C. 103 as being unpatentable over Cruz et al. (Garcia Santa Cruz, Beatriz, et al. "Generalising from conventional pipelines using deep learning in high-throughput screening workflows." Scientific Reports 12.1 (2022): 11465) (hereafter, "Cruz") in view of Alarcon et al. (Alarcon, Mauro Lemus, et al. "Remote Instrumentation Science Environment for Intelligent Image Analytics." 2022 IEEE 18th International Conference on e-Science (e-Science), IEEE, 2022) (hereafter, “Alarcon”), as applied to claim 1 above, and further in view of Park et al. (US 2022/0183645) (hereafter, "Park").
Regarding claim 6, Cruz in view of Alarcon discloses the method according to claim 1.
However, neither Cruz nor Alarcon, whether considered individually or in combination, explicitly disclose wherein the machine-learning model is trained, using reinforcement learning, to transform an image of the scientific or surgical imaging system into an output, wherein a difference between the output of the machine-learning model during training and a training output of the plurality of training outputs is used to determine a reward during the reinforcement learning-based training.
Park teaches wherein the machine-learning model is trained, using reinforcement learning, to transform an image of the scientific or surgical imaging system into an output (¶0061, lines 1-4, The agent unit 310 may train an agent to have an appropriate color conversion by allowing a color conversion image generated based on a result of the reinforcement learning), wherein a difference between the output of the machine-learning model during training and a training output of the plurality of training outputs is used to determine a reward during the reinforcement learning-based training (¶0064, lines 15-19, a difference between the reconstructed image and a result passing through image preprocessing such as a morphology operation, and the like, is defined to be an error, and a reciprocal of the error is used as a reward of a deep Q-network).
Cruz, Alarcon, and Park are analogous to the claimed invention because Cruz and Alarcon are directed at replacing manual steps of the scientific workflow with machine learning models and Park is in the field of medical image processing. It would have been obvious to a person of ordinary skill before the effective filing date of the claimed invention to incorporate the reinforcement learning of Park into the iterative parameter updating of Alarcon and the machine learning image processing workflow of Cruz. The suggestion/motivation for doing so would have been the substitution of a reinforcement learning model in place of the deep-learning model of Cruz would have been obvious to a person of ordinary skill in the art (¶0040, lines 12-13, perform training of the image data through deep learning of artificial intelligence, and the like. Park suggests the applicability of deep learning as well as similar algorithms).
This method of improving Cruz was within the ordinary ability of one of ordinary skill in the art based on the teachings of Alarcon and Park.
Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date, to modify Cruz with the teachings of Alarcon and Park to obtain the invention as specified in claim 6.
Claim 7 is rejected under 35 U.S.C. 103 as being unpatentable over Cruz et al. (Garcia Santa Cruz, Beatriz, et al. "Generalising from conventional pipelines using deep learning in high-throughput screening workflows." Scientific Reports 12.1 (2022): 11465) (hereafter, "Cruz") in view of Alarcon et al. (Alarcon, Mauro Lemus, et al. "Remote Instrumentation Science Environment for Intelligent Image Analytics." 2022 IEEE 18th International Conference on e-Science (e-Science), IEEE, 2022) (hereafter, “Alarcon”), as applied to claim 1 above, and further in view of Hillen (US 2019/0313963).
Regarding claim 7, Cruz in view of Alarcon discloses the method according to claim 1.
However, neither Cruz nor Alarcon, whether considered individually or in combination, explicitly disclose wherein the machine-learning model is trained, as a generator model of a pair of generative adversarial networks to transform an image of the scientific or surgical imaging system into an output, with a discriminator model of the pair of generative adversarial networks being trained based on the plurality of training outputs.
Hillen teaches wherein the machine-learning model is trained, as a generator model of a pair of generative adversarial networks to transform an image of the scientific or surgical imaging system into an output (¶0033, lines 71-79, For example, the machine learning inference 412 and machine learning trainer 408 can use neural networks such as a generative adversarial networks (GANs) in its machine learning architecture (e.g., an unsupervised machine learning architecture). In general, a GAN includes a generator neural network, a different specific x implementation kind of the Image Augmentation 506, that generates data (e.g., different versions of the same image by flips, inversions, mirroring etc.), with a discriminator model of the pair of generative adversarial networks being trained based on the plurality of training outputs (¶0033, lines 79-81, that is evaluated by a discriminator neural network, a specific type of the Neural Network Model 508, for authenticity (e.g., to identify the dental images).
Cruz, Alarcon, and Hillen are analogous to the claimed invention because Cruz and Alarcon are directed at replacing manual steps of the scientific workflow with machine learning models and Hillen is in the field of medical image processing. It would have been obvious to a person of ordinary skill before the effective filing date of the claimed invention to incorporate the generative adversarial network of Hillen into the iterative parameter updating of Alarcon and the machine learning based image processing workflow of Cruz. The suggestion/motivation for doing so would have been the substitution of a generative adversarial network model in place of the deep-learning model of Cruz would have been obvious to a person of ordinary skill in the art (¶0033, lines 4-5, supervised learning techniques may be implemented; ¶0033, lines 30-31, reinforcement learning techniques may also be used; ¶0033, lines 58-59, stochastic gradient descent. Hillen describes multiple different machine learning training options).
This method of improving Cruz was within the ordinary ability of one of ordinary skill in the art based on the teachings of Hillen.
Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date, to modify Cruz with the teachings of Hillen to obtain the invention as specified in claim 7.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Alam et al. (Alam, Shahinur, et al. "An end-to-end pipeline for fully automatic morphological quantification of mouse brain structures from MRI imagery." Frontiers in bioinformatics 2 (2022): 865443) discloses replacing an image processing pipeline with a machine learning algorithm (Page 3, DeepBrainIPP facilitates high throughput brain structure segmentation).
Knull et al. (Krull, Alexander, et al. "Artificial-intelligence-driven scanning probe microscopy." Communications Physics 3.1 (2020): 54) disclose using a machine learning model to adjust parameters of a microscopy system (Fig. 1 and Fig. 1 caption, DeepSPM determines the control signals (measurement parameters) and acquires an STM image. After acquisition, DeepSPM assesses the image quality. If the image is deemed “good”, DeepSPM processes it and stores it, and performs the next measurement. If “bad”, DeepSPM detects and addresses possible issues).
Larson et al. (US 2023/0108313) discloses updating a feature set based on a criterion and updating the image set (¶0062-0063, a machine-learning model to generate one or more identified features … the image selection criteria are based on one or more metrics associated with an identified feature).
Any inquiry concerning this communication or earlier communications from the examiner should be directed to XIAOMAO DING whose telephone number is (571)272-7237. The examiner can normally be reached Mon-Fri 9:00-5:00.
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, Henok Shiferaw can be reached at (571) 272-4637. 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.
/XIAOMAO DING/Examiner, Art Unit 2676
/Henok Shiferaw/Supervisory Patent Examiner, Art Unit 2676
Prosecution Insights