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 Amendments/Arguments
Response papers dated 3/11/26 are being entered.
In light of the amendments, the examiner removes the 101 rejection.
In light of amendments the examiner is using additional references rendering the arguments moot. The examiner maintains the art rejection.
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 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 1-3, 10-12, 17-19 are rejected under 35 U.S.C. 103 as being unpatentable over Strader (US 20190121532 A1). in further view of Kulshreshtha (US 20250005063 A1).
With respect to claims 1, 10, 17 Strader teaches (Claim 1) A method comprising: (Claim 10) A non-transitory computer-readable medium storing instructions that, when executed by at least one processor, cause a computer device to ([0044] In FIG. 3, there is shown one possible format of the display of the transcript on the interface of the workstation 210. The page shown includes a transcript tab 308 which has been selected, as well as a note tab 306, a chart tab 304, and patient instructions tab 302): (Claim 17) A system comprising: at least one processor; and at least one non-transitory computer-readable storage medium storing instructions that, when executed by the at least one processor, cause the system to ([0044] In FIG. 3, there is shown one possible format of the display of the transcript on the interface of the workstation 210. The page shown includes a transcript tab 308 which has been selected, as well as a note tab 306, a chart tab 304, and patient instructions tab 302 ):
receiving, from a client device of a first user, a communication stream containing contents of a communication between the first user and a second user (Strader ¶ [0042] The readers' attention will now be directed to FIGS. 3-14. We will now describe a method for generating a note summarizing a conversation between a patient and a healthcare provider. As will be explained in greater detail below in conjunction with FIG. 3, in essence we provide on a workstation (210 of FIG. 2) a tool (e.g., icon, tab, link, or other action on the interface) for rendering, i.e., playing on speaker associated with the workstation, an audio recording of the conversation and generating a transcript of the audio recording using a speech-to-text engine (part of the model 110 of FIG. 1). We then display on a first region 312 (FIG. 3) of a display of the workstation the transcript of the recording and simultaneously on a second region 314 of the display a note summarizing the conversation);
generating, from the communication stream, a transcript having a textual representation of the contents of the communication (Strader ¶ [0042] The readers' attention will now be directed to FIGS. 3-14. We will now describe a method for generating a note summarizing a conversation between a patient and a healthcare provider. As will be explained in greater detail below in conjunction with FIG. 3, in essence we provide on a workstation (210 of FIG. 2) a tool (e.g., icon, tab, link, or other action on the interface) for rendering, i.e., playing on speaker associated with the workstation, an audio recording of the conversation and generating a transcript of the audio recording using a speech-to-text engine (part of the model 110 of FIG. 1). We then display on a first region 312 (FIG. 3) of a display of the workstation the transcript of the recording and simultaneously on a second region 314 of the display a note summarizing the conversation);
generating, using the transcript, a communication summary that describes the contents of the communication (Strader ¶ [0042] The readers' attention will now be directed to FIGS. 3-14. We will now describe a method for generating a note summarizing a conversation between a patient and a healthcare provider. As will be explained in greater detail below in conjunction with FIG. 3, in essence we provide on a workstation (210 of FIG. 2) a tool (e.g., icon, tab, link, or other action on the interface) for rendering, i.e., playing on speaker associated with the workstation, an audio recording of the conversation and generating a transcript of the audio recording using a speech-to-text engine); and
providing the communication summary for display within a graphical user interface of the client device of the first user (Strader ¶ [0004] This disclosure relates to an interface (e.g., a display of a workstation used by a provider) for displaying a transcript of a patient-healthcare provider conversation and automated generation of a note or summary of the conversation using machine learning. The transcript and note can be generated in substantial real time during the office visit when the conversation is occurring, or later after the visit is over, ¶[0042] We then display on a first region 312 (FIG. 3) of a display of the workstation the transcript of the recording and simultaneously on a second region 314 of the display a note summarizing the conversation).
Strader does not explicitly disclose however Kulshreshtha teaches generating a large language model prompt to facilitate use of the transcript in creating a communication summary that describes the contents of the communication ([0024] An individual record of the plurality of records may comprise (a) an instruction prompt requesting a creation of an annotated summary of a transcript of a dialog between a given set of dialog participants (such as a doctor and a patient), (b) the transcript of the dialog, and (c) an annotated summary of the transcript. The terms “dialog” and “conversation” may be used synonymously herein to refer to natural language exchanges between two or more communicating entities, with the different entities/participants involved taking respective turns to communicate with one another. A given annotated summary of the data set may comprise a plurality of text sequences such as sentences or turns (where each turn may comprise a portion of one or more sentences), and a respective annotation corresponding to individual text sequences of the plurality of text sequences. An annotation corresponding to a particular text sequence may for example indicate (e.g., via a numeric identifier in a format such as “[1]”, “[2]” etc.) a portion of the transcript which comprises evidence for the particular text sequence. Such annotations may also be referred to as evidence mappings herein, as each such annotation may indicate a specific relationship or mapping between a text sequence in the summary and the corresponding evidence in the source transcript);
generating, using a large language model, a communication summary from the transcript in accordance with the large language model, the large language model including a set of model parameters that have been updated during an additional training phase from initial values determined via an initial training phase ([0037] The preparation of the models may include one or more training phases, including (in the case of LLMs) a pre-training phase and a fine-tuning phase, some or all of which may be orchestrated by training coordinators 139 implemented using software and hardware of one or more computing devices in various embodiments. For example, one or more LLMs may be pre-trained using the pre-training data 141, and one or more of the pre-trained LLMs may then be selected for instruction fine-tuning (using the pre-training records which contain instruction prompts of the kind mentioned above) during which the pre-trained LLMs are further trained to produce annotated summaries of conversations); Examiner Note: Fine-tuning by definition adjusts parameters from initial to final values; and
It would have been obvious to one of ordinary skill in the art prior to the effective filing date of the invention to modify summary of Strader to include LLM of Kulshreshtha in order to create automates summaries ([0025], Kulshreshtha);
With respect to claims 2, 11, and 18 Strader teaches generating the communication summary to include at least one content entry within the one or more entry fields based on the textual representation of the transcript. (Strader¶[0046] FIG. 4 shows an example of how the highlighted phrases “leg hurts”, 402, and “feeling feverish” 403 are extracted from the transcript and placed into a note. Furthermore, attributes of the symptom [entry field] “feeling feverish”, and specifically, onset, alleviating, and tempo (404) are also extracted from the transcript and placed into the note. The area 406 is a field for the user to indicate the chief complaint (CC), and whereas originally this field is blank the user is able to click and drag the text “leg hurts” 402 into the CC field 406 as indicated on the right side of FIG. 4. Thus, the grouping of text elements in the note is editable, ¶[0058] The workstation includes a display (FIG. 3) having a first transcript region (312, FIG. 3) for display of the transcript of the recording and simultaneously a second note region (314, FIG. 3, FIG. 4, FIG. 13, etc) [generating templates] for display of a note summarizing the conversation. A trained machine learning model (FIG. 1, 104) extracts words or phrases in the transcript related to medical topics relating to the patient, and in the system the extracted words or phrases are displayed in the note region 314 of the display (FIG. 4, FIG. 17, etc.) ). Examiner Note: Fig. 4 shows the template and the entry field
PNG
media_image1.png
331
688
media_image1.png
Greyscale
Kulshreshtha further teaches generating, using the large language model, a summary template having one or more entry fields and natural language in accordance with the large language model prompt ([0026] In some embodiments, the instruction prompts included in the different records of the fine-tuning data set may be identical. In other embodiments, the instruction prompts may vary at least slightly from one record to another; this may be done so as to make the LLM more capable of responding to prompts that provide similar guidance but need not be identical. In one embodiment, one or more keywords (such as names of medicines or names of treatment options in the case of medical dialogs) may be extracted automatically from a dialog transcript, and the instruction prompt may specify that the generated summary should include at least some of the keywords.)
With respect to claims 3, 12, 19 Strader teaches wherein: generating the summary template having the one or more entry fields comprises generating, within the summary template, an entry field corresponding to at least one of a call reason, an action item, or a topic discussed during the communication ; and (¶[0046] FIG. 4 shows an example of how the highlighted phrases “leg hurts”, 402, and “feeling feverish” 403 are extracted from the transcript and placed into a note. Furthermore, attributes of the symptom [entry field] “feeling feverish” [topic discussed] , and specifically, onset, alleviating, and tempo (404) are also extracted from the transcript and placed into the note. The area 406 is a field for the user to indicate the chief complaint (CC), and whereas originally this field is blank the user is able to click and drag the text “leg hurts” 402 into the CC field 406 as indicated on the right side of FIG. 4. Thus, the grouping of text elements in the note is editable, ¶[0058] The workstation includes a display (FIG. 3) having a first transcript region (312, FIG. 3) for display of the transcript of the recording and simultaneously a second note region (314, FIG. 3, FIG. 4, FIG. 13, etc) [generating templates] for display of a note summarizing the conversation. A trained machine learning model (FIG. 1, 104) extracts words or phrases in the transcript related to medical topics relating to the patient, and in the system the extracted words or phrases are displayed in the note region 314 of the display (FIG. 4, FIG. 17, etc.) ).
generating the communication summary to include the at least one content entry within the one or more entry fields based on the textual representation of the transcript comprises using a categorization model to generate a content entry by generating a call reason entry, an action item entry, or a topic entry based on the textual representation of the transcript (¶[0005]in a second note region a note summarizing the conversation, the note including automatically extracted words or phrases in the transcript related to medical topics relating to the patient, the extraction of the words or phrase performed with the aid of a trained machine learning model [categorization model], ¶ [006] The words or phrase can be placed into appropriate categories or classifications in the note such as under headings for symptoms, medications,).
Claims 4, 13 are rejected under 35 U.S.C. 103 as being unpatentable over Strader, Kulshreshtha in further view of Ghosh (US 20160132494 A1) and Chizi (US-20190166102-A1).
With respect to claims 4 and 13, Strader and Kulshreshtha do not explicitly disclose however Ghosh teaches further comprising: providing, for display within a graphical user interface of an additional client device, one or more interactive options for generating or modifying a pre-configured rule for generating content entries for communication summaries (Ghosh ¶[0060] In one embodiment, 406 includes accessing one or more rules for generating the summary. In this embodiment, various rules can be used to generate the summary data associated with the user's reading history. In one embodiment, the rules are configurable and definable by the user. For example, in one embodiment, a user interface for defining the one or more rules for generating the summary is provided.);
It would have been obvious to one of ordinary skill in the art prior to the effective filing date of the invention to modify summary of Strader in view of LLM of Kulshreshtha to include display of Ghosh in order to automate summary generation ([0042], Ghosh);
None of Strader, Kulshreshtha and Ghosh explicitly disclose however Chizi teaches generating or modifying the pre-configured rule in response to one or more user interactions with the one or more interactive options, wherein using the categorization model to generate the content entry comprises using the categorization model to generate the content entry in accordance with the pre-configured rule (Chizi ¶[0159] Moreover, we create a rule “if a store name contains ‘PIZZERIA’ then assign category ‘RESTAURANT’” and store it for future usage (2.2). To improve accuracy, these rules can be further validated, altered and maintained by domain experts (2.3). The above-mentioned rules are stored and managed in a rule management tool (2.4). Similarly to assigning categories, the system also allows for tagging transactions across categories. The above-described method can be also used to incorporate external data sources that map store names to categories (e.g. Gouden Gids, Resto.be). When a sufficient number of stores is categorized, a machine learning model can be used that predicts categories based on store characteristics (described as summary data in the 2nd paragraph) and external data.).
It would have been obvious to one of ordinary skill in the art prior to the effective filing date of the invention to modify summary of Strader in view of LLM of Kulshreshtha in view of display of Ghosh to include modifications of rules of Chiz in order to automate summary generation.
Claims 5, 14 are rejected under 35 U.S.C. 103 as being unpatentable over Strader, Kulshreshtha in further view of Vyas (US 20250252384 A1) and Leidner (US 20210043211 A1).
With respect to claims 5 and 14, Strader, Kulshreshtha do not explicitly disclose however Vyas teaches further comprising generating a pre-configured rule for generating content entries for communication summaries using the large language model (Vyas ¶[0007] In accordance with some embodiments of the present disclosure, the computerized-method may include: (i) operating a prompt-generator module to yield a prompt-text. The prompt-generator module may include a. retrieving a rule of configuration of the text-summary of the multiple-sections text-document, from a summary-configuration database; b. fetching data related to the multiple-sections text-document from a database of the application. The data related to the multiple-sections text-document may include one or more sections. Each section of the one or more sections may include one or more questions and each question of the one or more questions has a corresponding answer; and c. generating the prompt-text based on the rule of configuration and the data related to the multiple-sections text-document and the multiple-sections text-document; (ii) generating the text-summary by operating a Generative Artificial Intelligence (GenAI) with Large Language Models (LLM) s service of a cloud GenAI with LLM service-provider to execute the prompt-text; and (iii) storing the text-summary in a summary-database to be used to operate one or more actions for the tenant),
It would have been obvious to one of ordinary skill in the art prior to the effective filing date of the invention to modify summary of Strader in view of LLM of Kulshreshtha to include LLM of Vyas in order to improve the summarization ([0005] Vyas).
None of Strader, Kulshreshtha and Vyas explicitly disclose however Leidner teaches wherein using the categorization model to generate the content entry comprises using the categorization model to generate the content entry in accordance with the pre-configured rule (Leidner ¶[0047] “Classify” may refer to generating a classification that indicates whether a given segment of a transcript 104 is relevant to the transcript 104. Such classification may be binary (yes/no) or continuous (range of values). A segment of a transcript 104 is relevant to the transcript 104 if the segment is to be included in the summary 161 of the transcript 104 or otherwise content for the summary 161 is based on the segment. In a rule-based context, such classification may be made based on the evaluation of rules. In a machine-learning context, such classification may be made based on supervised learning using labeled transcripts annotated by subject matter experts.).
It would have been obvious to one of ordinary skill in the art prior to the effective filing date of the invention to modify summary of Strader in view of LLM of Kulshreshtha in view of LLM of Vyas to include categorization of Leidner in order to improve performance ([0091] Leidner).
Claims 6, 15 are rejected under 35 U.S.C. 103 as being unpatentable over Strader and Kulshreshtha in further view of Asl (US 20240346232 A1).
With respect to claims 6 and 15 Strader and Kulshreshtha do not explicitly disclose however Asl teaches generating, using the large language model, the communication summary using a large language model based on one or more triggering utterances represented within the transcript in accordance with the large language model prompt (Asl ¶ [0004] Example solutions for performing dynamic construction of large language model prompts include: generating a summary of a transcript using a summarization model; extracting topic-specific data from stored data using a scoring model; dynamically generating a language model prompt using the topic-specific data and the summary; and generating an output text using a language model and the language model prompt, ¶[0026] Summarization model 200 has been trained to intake large transcripts, covering conversations lasting several tens of minutes (e.g., 30 minutes or more), and produce a relevant summary. Thus, summary 114 includes at least one action item 204 and at least one summary highlight 206. Example action items and highlights are shown in FIGS. 5A-6B. Highlights are short phrases summarizing a single-topic aspect of transcript 112, which may each cover several minutes of a conversation, and action items are similar to highlights, but giving an expectation of some future activity by one or more persons)
It would have been obvious to one of ordinary skill in the art prior to the effective filing date of the invention to modify summary of Strader in view of LLM of Kulshreshtha to include triggering of ASL in order to improve the summarization.
Claims 7 and 16 are rejected under 35 U.S.C. 103 as being unpatentable over Strader in further view of Armady (US 20160241546 A1).
With respect to claims 6 and 15 Kulshreshtha teaches large language models with prompt.
None of Strader, Kulshreshtha explicitly disclose however Armady teaches wherein generating a redacted transcript by redacting personally identifiable information associated with the first user or the second user from the transcript, wherein generating, using the large language model, the communication summary from the transcript comprises generating, using the large language model, the communication summary from the redacted transcript. (Armady ¶[0092] FIG. 9 is a flowchart illustrating a method 900 to summarize an electronic document without using sensitive information according to an embodiment).
It would have been obvious to one of ordinary skill in the art prior to the effective filing date of the invention to modify summary of Strader in view of LLM of Kulshreshtha to include redaction of Armady in order to increase efficiency without compromising security ([0018] Armady).
Claims 8 are rejected under 35 U.S.C. 103 as being unpatentable over Strader and Kulshreshtha in further view of Narayanaswamy (US 20190268379 A1).
With respect to claim 8 Strader and Kulshreshtha do not explicitly disclose however Narayanaswamy teaches further comprising: determining personally identifiable information associated with the first user or the second user from metadata related to the communication (Narayanaswamy¶[0016] Enterprise organizations have a business need to store sensitive data, such as financial or patient information, ¶[0127] determining that the retrieved sensitivity metadata identifies the document as sensitive and blocking the data egress request, ¶and revising or copying of the document can include sensitive data because the document source is the HR application. ); and
generating the communication summary to include the personally identifiable information ¶[0127] determining that the retrieved sensitivity metadata identifies the document as sensitive and blocking the data egress request, ¶and revising or copying of the document can include sensitive data because the document source is the HR application.
It would have been obvious to one of ordinary skill in the art prior to the effective filing date of the invention to modify summary of Strader in view of LLM of Kulshreshtha to include sensitive metadata information of Narayanaswamy in order to increase computational efficiency without the need for content-based analysis ([0095] Narayanaswamy).
Claims 9 and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Strader and Kulshreshtha in further view of Ramalingam (US 20240320684 A1).
With respect to claims 9 and 20 , Strader and Kulshreshtha do not explicitly disclose however Ramalingam teaches receiving, via the graphical user interface of the client device of the first user, one or more user interactions with respect to the communication summary (Ramalingam¶[0031] To provide services, the service team (including chat agents and service representatives) interfaces with both customers and the CSB service engine 120 to carry out service related functions. The present teaching discloses the aspects of such services that are facilitated by chat summaries created automatically based on chat transcripts in accordance with models learned via training and adapted dynamically over the course of providing services. The CSB service engine 120 comprises an automated chat summary generator 240, a chat summary modification unit 260, a chat summary indexing/archive unit 270, and a customer service module 280. In this illustrated embodiment, the automated chat summary generator 240 is provided to automatically generate a chat summary based on a transcript from an online chat between a customer and a chat agent. The chat summary modification unit 260 is provided as an optional unit which is to be used to interface with a chat agent to modify a chat summary, which may correspond to an automatically generated chat summary or a previously modified chat summary (e.g., retrieved from the summary database 130);
modifying the communication summary in response to the one or more user interactions (Ramalingam ¶[0031] The chat summary modification unit 260 is provided as an optional unit which is to be used to interface with a chat agent to modify a chat summary, which may correspond to an automatically generated chat summary or a previously modified chat summary (e.g., retrieved from the summary database 130.)
It would have been obvious to one of ordinary skill in the art prior to the effective filing date of the invention to modify summary of Strader in view of LLM of Kulshreshtha to include user interactions of Ramalingam in order to improve summarization models ([0018] Ramalingam).
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). 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 date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ATHAR N PASHA whose telephone number is (408)918-7675. The examiner can normally be reached Monday-Thursday Alternate Fridays, 7:30-4:30 PT.
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, Daniel Washburn can be reached on (571)272-5551. 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.
/ATHAR N PASHA/ Primary Examiner, Art Unit 2657
Prosecution Insights