DETAILED ACTION
Notices to Applicant
This communication is a final rejection. Claims 1, 3-8, and 10-22, as filed 05/18/2026, are currently pending and have been considered below.
No priority is acknowledged.
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA . 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 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.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1, 3-8, and 10-22 are rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter.
Step 1
The claim(s) recite(s) subject matter within a statutory category as a process, machine, and/or article of manufacture which recite:
1. A system for generating feedback for at least one ultrasound scan, the system comprising:
At least one processor communicatively coupled with at least one memory, the at least one processor also communicatively coupled with at least one user interface; (additional element – applying the abstract idea with a computer)
the at least one ultrasound scan represented by ultrasound scan data generated by an ultrasound scan device; (additional element – insignificant extra-solution activity; mere data gathering)
an image review generator comprising a natural language processing (NLP) module and a large language model (LLM) module, the LLM module trained on a diverse dataset of stored ultrasound scans and associated findings; (abstract idea – mental process and/or mathematical concept)
a set of assessment facts provided by human reviewers and other automated sources related to the stored ultrasound scans and used as the basis of a review of the ultrasound scan data; (additional element – insignificant extra-solution activity; mere data gathering)
a set of instructions for the LLM module detailing a writing format and writing style for the review of the ultrasound scan data created by the LLM module, and providing background medical information for reference; (additional element – insignificant extra-solution activity; mere data gathering) and
the at least one user interface driven by a per-exam configuration allowing a reviewer to select appropriate items based on findings accuracy, scan completeness, and other teaching points, (additional element – applying the abstract idea with a computer)
wherein the LLM module synthesizes the set of assessment facts and the set of instructions to generate a written review of the ultrasound scan data (abstract idea – mental process and/or mathematical concept).
3. The system of claim 1, wherein the LLM module synthesizes the set of assessment facts and the set of instructions to generate the desired written review in multiple languages. (abstract idea – mental process and/or mathematical concept)
4. The system of claim 1, wherein the set of assessment facts comprises binary and categorical facts related to accuracy and completeness of at least one ultrasound scan. (additional element – insignificant extra-solution activity; mere data gathering)
5. The system of claim 1, wherein the set of instructions for the LLM module comprises at least one example of desired output and detailed information on proper ultrasound procedures. (additional element – insignificant extra-solution activity; mere data gathering)
6. The system of claim 1, wherein the user interface allows the reviewer to select appropriate items based on findings accuracy, scan completeness, and other teaching points in multiple languages. (additional element – applying the abstract idea with a computer)
7. The system of claim 1, further comprising a translator module that generates reviews in multiple languages. (abstract idea – mental process)
Claims 1-7 is presented as exemplary claims but the same analysis applies to the other claims.
Step 2A Prong One
The broadest reasonable interpretation (BRI) of these steps includes mental processes and mathematical concepts because the AI model is recited at a high enough level that it encompasses simply models that can be trained mentally. The LLM module is also a mathematical concept because the different techniques used to build and train the model (e.g., “deep learning neural networks, natural language processing (NLP), and machine learning algorithms such as transformer models,” [0057] of the specification) are mathematical concepts because these algorithms represent mathematical relationships between data. As such, the training of the machine learning model represents a mathematical concept that is interpreted to be part of the identified abstract idea.
Additionally, the BRI of the claimed invention further includes certain methods of organizing human activity, namely, steps a human would follow teach another person to interpret ultrasounds, generate findings, and output a report in a desired format.
Dependent claims recite additional subject matter which further narrows or defines the abstract idea embodied in the claims. For example, claims 3 and 7 provide additional details of the model which amount to refinements to the mental processes and mathematical concepts but for recitation of generic computer components.
Step 2A Prong Two
This judicial exception is not integrated into a practical application. In particular, the additional elements do not integrate the abstract idea into a practical application, other than the abstract idea per se, because the additional elements:
amount to mere instructions to apply an exception. For example, claims 1 and 6 amount to invoking computers as a tool to perform the abstract idea, see applicant’s specification [0044], see MPEP 2106.05(f))
add insignificant extra-solution activity to the abstract idea. For example, gathering various scan data, AI model data, assessment facts, and AI model instructions in claim 1 amounts to mere data gathering and selecting a particular data source or type of data to be manipulated, see MPEP 2106.05(g))
generally link the abstract idea to a particular technological environment or field of use such as the various types of ultrasound machines in claims 16 and 17, see MPEP 2106.05(h))
Dependent claims recite additional subject matter which amount to limitations consistent with the additional elements in the independent claims as shown above. Claim recites additional limitations which add insignificant extra-solution activity to the abstract idea which amounts to mere data gathering. Looking at the limitations as an ordered combination adds nothing that is not already present when looking at the elements taken individually. There is no indication that the combination of elements improves the functioning of a computer or improves any other technology. Their collective functions merely provide conventional computer implementation and do not impose a meaningful limit to integrate the abstract idea into a practical application.
Step 2B
The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to discussion of integration of the abstract idea into a practical application, the additional elements amount to no more than mere instructions to apply an exception, add insignificant extra-solution activity to the abstract idea, and generally link the abstract idea to a particular technological environment or field of use. Additionally, the additional limitations, other than the abstract idea per se amount to elements that have been recognized as well-understood, routine, and conventional activity in particular fields, such as receiving or transmitting data over a network, Symantec, MPEP 2106.05(d)(II)(i), performing repetitive calculations, Flook, MPEP 2106.05(d)(II)(ii), electronic recordkeeping, Alice Corp., MPEP 2106.05(d)(II)(iii), and/or storing and retrieving information in memory, Versata Dev. Group, MPEP 2106.05(d)(II)(iv).
Dependent claims recite additional subject matter which, as discussed above with respect to integration of the abstract idea into a practical application, amount to invoking computers as a tool to perform the abstract idea. Dependent claims recite additional subject matter which amount to limitations consistent with the additional elements in the independent claims. Looking at the limitations as an ordered combination adds nothing that is not already present when looking at the elements taken individually. There is no indication that the combination of elements improves the functioning of a computer or improves any other technology. Their collective functions merely provide conventional computer implementation.
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.
The factual inquiries set forth in Graham v. John Deere Co., 383 U.S. 1, 148 USPQ 459 (1966), that are applied 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, 11-15, and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Paik (US20210264212A1) in view of Singhal (Singhal, K., Azizi, S., Tu, T. et al. Large language models encode clinical knowledge. Nature 620, 172–180 (2023). https://doi.org/10.1038/s41586-023-06291-2).
Regarding claim 1, Paik discloses: A system for generating feedback for at least one ultrasound scan, the system comprising:
--at least one processor communicatively coupled with at least one memory, the at least one processor also communicatively coupled with at least one user interface; the at least one ultrasound scan represented by ultrasound scan data generated by an ultrasound scan device (“said medical image is…an ultrasound image” [0010]; “a computer based system or platform, comprising: (a) a processor; (b) a display configured to show a graphical user interface for displaying a medical image,” [0009]; “aid processor is configured to collect analytics on user interaction with said system and provide feedback for improving efficiency or quality,” [0010]);
--an image review generator comprising a natural language processing (NLP) module (“said machine learning software module is trained using at least one medical image and at least one corresponding medical report,” [0016]; FIG. 13; “an NLP-based system which detects incongruence in inconsistency between a user's explicit interaction with the outputs of the AI models (through the described UI components) and the words they dictate into the diagnostic report,” [0149]);
--a set of assessment facts provided by human reviewers and other automated sources related to the stored ultrasound scans and used as the basis of a review of the ultrasound scan data (“Accordingly, described here is system for gathering that ground truth feedback from users, and well as an NLP-based system which detects incongruence in inconsistency between a user's explicit interaction with the outputs of the AI models (through the described UI components) and the words they dictate into the diagnostic report. This system ensures that the highest-fidelity ground truth data is fed back into the algorithms for training…Both decisions provide ground truth data that can be used to further train the algorithm to improve future performance,” [0149]; expert annotation of training data in [0263]-[0264]);
--a set of instructions for the LLM module detailing a writing format and a writing style for the review of the ultrasound scan data created by the LLM module, and providing background medical information for reference (“the natural text representation can be created through a simple set of production rules given the structure of the knowledge graph. Alternatively or in combination, the natural text representation can be created through a simple set of production rules given the structure of the knowledge graph. In the example from the previous paragraph, a query might return “A mild neuroforaminal stenosis is observed at the C2-3 level” from an existing database of parsed findings while a production rule like “<anatomy> has <severity><observation>” might return “C2-C3 foramen has mild stenosis,” [0109]; “the Ontology can contain hierarchical information as well as synonyms, which allows for the aggregation and ordering of findings appropriate for communication to end users,” [0220]; “Accordingly, in some cases, the system comprises an algorithm configured to detect a shorthand or slang from the spoken audio and convert the audio into the long-form or appropriate format (e.g., plain English) for a medical report,” [0103]); and
--the at least one user interface driven by a per-exam configuration allowing a reviewer to select appropriate items (accepts or amending findings in FIG. 7; “an audio detection component configured to detect or record an input that indicates when said user accepts inclusion of said computer-generated finding,” [0010]; “(a) displaying a medical image; (b) generating a medical report including a computer-generated finding related to said medical image when a user accepts inclusion of said computer-generated finding within said report,” [0011]; “said processor is configured to collect analytics on user interaction with said system and provide feedback for improving efficiency or quality,” [0011]) based on findings accuracy, scan completeness, and other teaching points (the examiner notes that this limitation is non-functional, descriptive material that does not limit the claim. The thought process of the user making appropriate selections is not part of the structure of the claim. Because Paik has a user interface that allows the user to select items, Paik discloses this limitation regardless of the reasons justifying the user’s actions. That said, for purposes of compact prosecution, the examiner notes that Paik expressly discloses users making decisions in order to improve efficiency or quality in [0011]).
Paik does not expressly disclose but Singhal teaches:
--the LLM module trained on a diverse dataset of stored ultrasound scans and associated findings (“instruction prompt tuning, a parameter efficient approach for aligning LLMs to new domains using a few exemplars. The resulting model, Med-PaLM, performs encouragingly, but remains inferior to clinicians. We show that comprehension, knowledge recall and reasoning improve with model scale and instruction prompt tuning, suggesting the potential utility of LLMs in medicine,” Abstract; “Our evaluation assesses answers for agreement with the scientific and clinical consensus, the likelihood and possible extent of harm, reading comprehension, recall of relevant clinical knowledge, manipulation of knowledge via valid reasoning, completeness of responses, potential for bias, relevance and helpfulness,” page 173);
--wherein the LLM module synthesizes the set of assessment facts and the set of instructions to generate a written review of the ultrasound scan data (“instruction prompt tuning…using few exemplars,” Abstract; “We refer to this method of prompt tuning as ‘instruction prompt tuning’. Instruction prompt tuning can thus be seen as a lightweight way (data-efficient, parameter-efficient, compute-efficient during both training and inference) of training a model to follow instructions in one or more domains. In our setting, instruction prompt tuning adapted LLMs to better follow the specific type of instructions used in the family of medical datasets that we targeted,” page 13).
One of ordinary skill in the art would have been motivated before the effective filing date to expand the AI medical imaging analysis of Paik with formulaic “production rules” for generating reports to include the generative architecture using LLMs of Singhal because this would improve the naturalness of the report language, the accuracy of the clinical reasoning when synthesizing multiple facts, and allow for input/output in multiple languages.
Additionally, it can be seen that each element is taught by either Paik or Singhal. The LLM teachings workflows of Singhal do not affect the normal functioning of the elements of the claim which are taught by Paik. Because the elements do not affect the normal functioning of each other, the results of their combination would have been predictable. Therefore, before the effective filing date of the claimed invention, it would have been obvious to combine the teachings of Paik with the teachings of Singhal since the result is merely a combination of old elements, and, since the elements do not affect the normal functioning of each other, the results of the combination would have been predictable.
Regarding claim 4, Paik discloses: wherein the set of assessment facts comprises binary and categorical facts related to accuracy and completeness of at least one ultrasound scan (“said identification or evaluation of said pathology comprises at least one of a severity, quantity (e.g., number of lung nodules), measurement (e.g., length, area, and/or volume of a lung nodule), presence,” [0010]; labeling anatomy including presence/absence of pathology in [0075]; [0104]).
Regarding claim 5, Paik discloses: wherein the set of instructions for the LLM module comprises at least one example of desired output and detailed information on proper ultrasound procedures (“Alternatively or in combination, the natural text representation can be created through a simple set of production rules given the structure of the knowledge graph. In the example from the previous paragraph, a query might return “A mild neuroforaminal stenosis is observed at the C2-3 level” from an existing database of parsed findings while a production rule like “<anatomy> has <severity><observation>” might return “C2-C3 foramen has mild stenosis,”” [0109]; “As an example, if the entire spine is not within the field of view of the image series being analyzed, a shift of the labels by one or more vertebrae is more likely, especially in the presence of anatomic variants,” [0080]; “by using the anatomic segmentation, the AI system would know which portion of the anatomy to measure using the segmented anatomy as the boundaries for diametrically opposed ray casting to make define the line segment from which to make the linear measurement. Then, the AI system can automatically construct the sentence to be inserted into the report. One example in the context of oncological imaging would be to apply the maximal height/width measurements of the RECIST (“Response evaluation criteria in solid tumors”) system,” [0104]).
Regarding claims 8 and 11-13, the claims are substantially similar to claims 1, 4, and 5 are rejected with the same reasoning.
Regarding claim 14, Paik further discloses: delivering the written review of the at least one ultrasound scan to the healthcare practitioner (“a medical image may be evaluated using artificial intelligence to detect one or more segmented features in the image, which are analyzed to generate a textual description of a finding that a user may be prompted to accept,” [0009]; “the user or radiologists is presented with a card overlay, on top of any AI findings/diagnoses, with the associated medical text that are suggested for insertion into the diagnostic report,” [0128]) .
Regarding claim 15, the claim is substantially similar to claim 1 and is rejected with the same reasoning.
Regarding claim 18, Paik further discloses: wherein the set of assessment facts includes data related to the accuracy of the at least one ultrasound scan data, the completeness of the at least one ultrasound scan data, and teaching points derived from the at least one ultrasound scan data (“the system automatically flags reports that contain significant discrepancy between user and the AI computer vision system for further inspection and review,” [0178]).
Claims 3, 7, and 10 are rejected under 35 U.S.C. 103 as being unpatentable over Paik in view of Singhal and Gulati (Gulati V et al., Transcending Language Barriers: Can ChatGPT Be the Key to Enhancing Multilingual Accessibility in Health Care? Journal of the American College of Radiology, 2024; 21, 1888-1895; Published online June 14, 2024).
Regarding claim 3, Paik does not expressly disclose but Singhal teaches: wherein the LLM module synthesizes the set of assessment facts and the set of instructions to generate the written review(“We pilot a framework for physician and lay user evaluation to assess multiple axes of LLM performance beyond accuracy on multiple-choice datasets. Our evaluation assesses answers for agreement with the scientific and clinical consensus, the likelihood and possible extent of harm, reading comprehension, recall of relevant clinical knowledge, manipulation of knowledge via valid reasoning, completeness of responses, potential for bias, relevance and helpfulness,” page 173).
The motivation to combine is the same as in claim 1.
Paik and Singhal do not expressly disclose that the output is in multiple languages, but Gulati teaches this (“translating radiology reports into Spanish, Hindi, and Russian languages,” Abstract).
One of ordinary skill in the art would have been motivated before the effective filing date to expand the AI medical imaging analysis of Paik and Singhal to include the multiple languages and translation workflows of Gulati because allowing a wide variety of reviewers such as those with different primary languages to use the system would allow the system of Paik to digest more finding and thus generate better analyses. Further, Gulati states, “Large language Models can be used for translating and simplifying radiology reports, potentially improving access to health care and helping reduce health care costs” on page 1889.
Additionally, it can be seen that each element is taught by either Paik, Singhal, or Gulati. The translation workflows of Gulati do not affect the normal functioning of the elements of the claim which are taught by Paik and Singhal. Because the elements do not affect the normal functioning of each other, the results of their combination would have been predictable. Therefore, before the effective filing date of the claimed invention, it would have been obvious to combine the teachings of Paik and Singhal with the teachings of Gulati since the result is merely a combination of old elements, and, since the elements do not affect the normal functioning of each other, the results of the combination would have been predictable.
Regarding claim 7, Paik does not expressly disclose but Gulati further teaches: a translator module that generates reviews in multiple languages (“translating radiology reports into Spanish, Hindi, and Russian languages,” Abstract).
The motivation to combine is the same as in claim 3.
Regarding claim 10, Paik does not expressly disclose but Gulati further teaches: wherein the large language model synthesizes the set of assessment facts and instructions to generate a written review in multiple languages (“translating radiology reports into Spanish, Hindi, and Russian languages,” Abstract).
The motivation to combine is the same as in claim 3.
Claims 6 and 19 are rejected under 35 U.S.C. 103 as being unpatentable over Paik in view of Singhal and Official Notice.
Regarding claim 6, Paik discloses: wherein the user interface allows the reviewer to select appropriate items based on findings accuracy, scan completeness, and other teaching points as described above in the rejection of claim 1. Singhal further discloses that “ability to respond to queries in multiple languages” was a recognized capability of the prior art (page 178). Paik and Singhal do not expressly disclose that this selection can be in multiple languages. However, the Examiner takes Official Notice that multilingual user interfaces (i.e., software localization in which a language-selection control selection causes the interface’s labels and selectable options to be rendered in the user’s chosen language) were well-known before the effective filing date. It would have been obvious to one of ordinary skill in the art to present the Paik/Singhal selection interface in multiple languages using conventional localization because this would improve access for reviewers and practitioners who prefer other languages.
Regarding claim 19, the claim is substantially similar to claim 6 and is rejected with the same reasoning.
Claims 16, 17, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Paik in view of Singhal and Savitsky (US20170018204A1).
Regarding claim 16, Paik does not expressly disclose but Savitsky teaches: wherein the diverse dataset of stored ultrasound scans and associated findings comprises data from various types of ultrasound machines and different healthcare settings (supporting “lower-end ultrasound machines [0031] and “modern ultrasound machines” [0010]; various types of ultrasounds in [0030]).
One of ordinary skill in the art would have been motivated before the effective filing date to expand the AI medical imaging analysis of Paik and Singhal to include the multiple types of ultrasound machines of Savitsky because this would ensure that the training scans accurately represent the diverse array of ultrasound machines that a practitioner is likely to encounter and thus have context to provide better analyses regardless of the machine type. See Savitsky [0005].
Regarding claim 17, Paik does not expressly disclose but Savitsky teaches: wherein the diverse dataset of stored ultrasound scans further comprises data from stored ultrasound scans selected from the group comprising of 2D, 3D, and Doppler scans (2D and 3D in [0030]).
The motivation to combine is the same as in claim 16.
Regarding claim 20, Paik does not expressly disclose but Savitsky teaches: wherein the user interface comprises options for a reviewer to select teaching points based on the specific type of ultrasound scan being reviewed (“creating customized learning content for ultrasound simulators,” Abstract; “The user proceeds by selecting a collection of volumes that are deemed relevant for a particular teaching purpose, and imports them into a virtual patient, who exemplifies a particular condition or medical scenario,” [0017]; “A collection of customized virtual patients will constitute a case library designed by the user to fulfill a particular curriculum of his/her choosing. The customized case library will be made available to the simulation environment, allowing learners to acquire hands-on experience interacting with the virtual patients in a manner that mimics closely how a medical practitioner would act in the presence of a real patient,” [0023]-[0024]).
One of ordinary skill in the art would have been motivated before the effective filing date to expand the AI medical imaging analysis of Paik and Singhal to include the multiple languages and educational workflows of Savitsky because this would allow tailored curricula for a given student or exam and thus “creat[e] a community of world experts contributing their knowledge and expertise to improve the quantity and quality of medical content available,” [0025].
Claims 21 and 22 are rejected under 35 U.S.C. 103 as being unpatentable over Paik in view of Singhal and Ma (Chong Ma et al., ImpressionGPT: An Iterative Optimizing Framework for Radiology Report Summarization with ChatGPT, arXiv:2304.08448v2 (Year: 2023)).
Regarding claim 21, Paik does not expressly disclose but Ma teaches:
--wherein the set of instructions comprises one or more sample reviews demonstrating a respective writing style (“dynamic prompt approach enables the model to learn contextual knowledge from semantically similar examples form existing data,” Abstract; “Specifically, we use similarity search techniques to construct a dynamic prompt to include semantically- and clinically-similar existing reports. These similar reports are used as examples to help ChatGPT learn the text descriptions and summarizations of similar imaging manifestations in a dynamic context,” page 3), and
--wherein the LLM module self-reviews the written review to ensure that statements in the written review are in alignment with the set of assessment facts and that a style of the written review is consistent with the set of instructions (“An iterative optimization algorithm is designed and implemented via prompt engineering to take advantage of ChatGPT's in-context learning ability while also continuously improving it through interaction,” page 3; “The above results show that ChatGPT learns more relevant prior knowledge through the dynamic context we designed, which significantly improves the quality of the generated results. This demonstrates the feasibility of using a small number of samples to fine-tune the LLMs. With the introduction of iterative optimization, the model can learn how to generate impression correctly with good responses, while learning how to avoid similar writing styles with bad responses. The model completes self-iterative updates in an interactive way, thus further optimizing the generated results, page 17”).
One of ordinary skill in the art would have been motivated before the effective filing date to expand the AI medical imaging analysis of Paik and Singhal to include the dynamic context of Ma because this would improve the quality of the generated results (“The above results show that ChatGPT learns more relevant prior knowledge through the dynamic context we designed, which significantly improves the quality of the generated results,” page 17).
Regarding claim 22, the claim is substantially similar to claim 21 and is rejected with the same reasoning.
Response to arguments
Applicant's arguments filed 05/18/2026 have been fully considered and are discussed below.
Regarding the subject matter ineligibility rejections, the Examiner acknowledges that the claims are now directed to statutory categories in Step 1.
Applicant argues that the claimed invention is not directed to a mental process (Step 2A Prong One) because a human cannot mentally “transformer models, attention mechanisms, or deep learning neural networks.” Remarks pages 7-8. These features are described in the specification only at high levels (i.e., specification [0057]-[0058]) and are not recited at all in the claims which merely require “an LLM module” and “a natural language processing (NLP) module”. Reciting these generic computer tools to perform the mental process does not remove the recited steps from the mental process grouping.
Applicant argues that the claimed invention integrates any abstract idea into a practical application (Step 2A Prong Two) by producing bespoke, individualized reviews with controlled style and multilingual output. Remarks pages 9-10. This is not persuasive because the claims recite desired results rather than the technical mechanisms for achieving them. The asserted benefits of faster, more consistent, and less labor-intensive reviews are improvements to the review workflow itself rather than to the functioning of a computer or LLM.
Applicant argues that the claimed invention amounts to significantly more than any abstract idea (Step 2B) because they include additional elements that go beyond well-understood, routine, and conventional elements such as a “specific combination of NLP and LLM modules to synthesize assessment facts and instructions into written reviews.” Remarks pages 10-11. This is not persuasive because the identified additional elements are either described as conventional in the specification, i.e., “the image review generator 114 uses prompt engineering with existing LLMs” in [0058], or are the types of functions that courts have held to be well-understood, routine, and conventional such as receiving and transmitting data over a network, storing and retrieving information in memory, and receiving user selections.
Regarding the prior art rejections, the anticipation rejections are withdrawn in view of claim amendments and arguments. Regarding the amended independent claims, Applicant argues that the combination of Paik and Singhal fails to render them obvious because “Singhal does not disclose an image review generator comprising both an NLP module and an LLM module.” Remarks page 14. This is not persuasive because the rejections do not rely on Singhal for these elements. The NLP module, ultrasound context, assessment facts, and per-exam configuration are all disclosed by Paik as described above. In response to applicant's arguments against the references individually, one cannot show nonobviousness by attacking references individually where the rejections are based on combinations of references. See In re Keller, 642 F.2d 413, 208 USPQ 871 (CCPA 1981); In re Merck & Co., 800 F.2d 1091, 231 USPQ 375 (Fed. Cir. 1986). Further, characterizing Singhal as being limited to “medical question answering” fails because the answering questions in context is a text-generation technique that is not confined to answering questions rather than composing reviews. This is particularly evident when viewed in combination with Paik.
Conclusion
Applicant’s amendment necessitated the new ground(s) of rejection presented in this Office Action (See MPEP 706.07(a)). Accordingly, 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 JOSHUA BLANCHETTE whose telephone number is (571)272-2299. The examiner can normally be reached on Monday - Thursday 7:30AM - 6:00PM, EST.
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, Shahid Merchant, can be reached on (571) 270-1360. 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.
/JOSHUA B BLANCHETTE/ Primary Examiner, Art Unit 3624
Prosecution Insights