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 .
Claims 1-20 are pending and have been examined.
Priority
Receipt is acknowledged of certified copies of papers required by 37 CFR 1.55.
Should applicant desire to obtain the benefit of foreign priority under 35 U.S.C. 119(a)-(d) prior to declaration of an interference, a certified English translation of the foreign application must be submitted in reply to this action. 37 CFR 41.154(b) and 41.202(e).
Failure to provide a certified translation may result in no benefit being accorded for the non-English application.
Drawings
New corrected drawings in compliance with 37 CFR 1.121(d) are required in this application because the drawings cannot be read (resolution too coarse, font too small, contrast not enough). In particular, Figures 9-10C. Applicant is advised to employ the services of a competent patent draftsperson outside the Office, as the U.S. Patent and Trademark Office no longer prepares new drawings. The corrected drawings are required in reply to the Office action to avoid abandonment of the application. The requirement for corrected drawings will not be held in abeyance.
Claim Interpretation
The following is a quotation of 35 U.S.C. 112(f):
(f) Element in Claim for a Combination. – An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The following is a quotation of pre-AIA 35 U.S.C. 112, sixth paragraph:
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The claims in this application are given their broadest reasonable interpretation using the plain meaning of the claim language in light of the specification as it would be understood by one of ordinary skill in the art. The broadest reasonable interpretation of a claim element (also commonly referred to as a claim limitation) is limited by the description in the specification when 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is invoked.
As explained in MPEP § 2181, subsection I, claim limitations that meet the following three-prong test will be interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph:
(A) the claim limitation uses the term “means” or “step” or a term used as a substitute for “means” that is a generic placeholder (also called a nonce term or a non-structural term having no specific structural meaning) for performing the claimed function;
(B) the term “means” or “step” or the generic placeholder is modified by functional language, typically, but not always linked by the transition word “for” (e.g., “means for”) or another linking word or phrase, such as “configured to” or “so that”; and
(C) the term “means” or “step” or the generic placeholder is not modified by sufficient structure, material, or acts for performing the claimed function.
Use of the word “means” (or “step”) in a claim with functional language creates a rebuttable presumption that the claim limitation is to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites sufficient structure, material, or acts to entirely perform the recited function.
Absence of the word “means” (or “step”) in a claim creates a rebuttable presumption that the claim limitation is not to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is not interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites function without reciting sufficient structure, material or acts to entirely perform the recited function.
Claim limitations in this application that use the word “means” (or “step”) are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action. Conversely, claim limitations in this application that do not use the word “means” (or “step”) are not being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action.
This application includes one or more claim limitations that do not use the word “means,” but are nonetheless being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, because the claim limitation(s) uses a generic placeholder that is coupled with functional language without reciting sufficient structure to perform the recited function and the generic placeholder is not preceded by a structural modifier. Such claim limitations are (underlined below, in claims 13, 15, and 18):
13. A system for classification of pre-anesthetic physical status of patients, the system comprising:
an electronic medical record database;
an extracting unit configured to extract medical records from an electronic medical record database, the medical records comprising at least one or more of surgical information, hospitalization initial diagnosis, nursing initial diagnosis, hospitalization progress, vital signs, test results, and clinical observation records of the patients;
a summary generation unit configured to generate medical summary information of patients from the extracted medical record using an artificial intelligence-based natural language processing system;
a medical classification model unit configured to classify the pre-anesthetic physical status of the patients using the generated medical summary information as an input of a medical classification model; and
a visualization unit configured to visualize and provide a predictive basis of the medical classification model.
15. The system for classification of pre-anesthetic physical status of patients of claim 13, wherein the summary generation unit is configured to perform:
a function of analyzing the extracted medical records to automatically identify the primary diagnosis and clinical condition of the patients;
a function of matching the relevant drug use history and the previous surgical history based on the diagnosis content of the patients;
a function of comparing and contrasting a plurality of test results with each other, and classifying the test results by diseases; and
a function of integrating the analysis, matching and classification results into a standardized form of medical summary information.
18. The system for classification of pre-anesthetic physical status of patients of claim 13, wherein
the pre-anesthesia physical status of patients comprises an American Society of Anesthesiologists Physical Status (ASA-PS) class, and
further comprising an anesthesia fee calculation unit configured to determine whether to apply additional anesthesia fees for patients classified as ASA-PS class III or higher.
Because these claim limitations are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, they are being interpreted to cover the corresponding structure described in the specification as performing the claimed function, and equivalents thereof.
If applicant does not intend to have these limitations interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, applicant may: (1) amend the claim limitations to avoid them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph (e.g., by reciting sufficient structure to perform the claimed function); or (2) present a sufficient showing that the claim limitations recite sufficient structure to perform the claimed function so as to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph.
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 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Claims 1-20 are directed to a method, product, or system, which are statutory categories of invention. (Step 1: YES).
Claim 12 is directed to a signal, which is not one of the statutory categories. (Step 1: NO). See rejection below. For compact prosecution, the claim is interpreted as a product claim.
The Examiner has identified method Claim 1 as the claim that represents the claimed invention for analysis and is similar to product Claim 12 and system Claim 13.
Claim 1 recites the limitations of:
A method for classification of pre-anesthetic physical status of patients, the method comprising:
extracting medical records from an electronic medical record database, the medical records comprising at least one or more of surgical information, hospitalization initial diagnosis, nursing initial diagnosis, hospitalization progress, vital signs, test results, and clinical observation records of the patients;
generating medical summary information of patients from the extracted medical record using an artificial intelligence-based natural language processing system;
classifying the pre-anesthetic physical status of the patients using the generated medical summary information as an input of a medical classification model; and
visualizing and providing a predictive basis of the medical classification model.
These above limitations, under their broadest reasonable interpretation, cover performance of the limitation as certain methods of organizing human activity. The claim recites elements, highlighted in bold above, which covers performance of the limitation as managing personal behavior and interactions between people. Extracting medical records of patients (following rules/instructions), generating a medical summary information of patients (following rules/instructions and teaching), classifying pre-anesthetic physical status of patients (following rules/instructions and teaching), and visualizing and providing a predictive basis of the classification model (teaching) are abstract steps directed to managing personal behavior and interactions between people. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation as managing interactions between people, then it falls within the “Certain Methods of Organizing Human Activity” grouping of abstract ideas. Accordingly, the claim recites an abstract idea. Claims 12 and 13 are also abstract for similar reasons. (Step 2A-Prong 1: YES. The claims are abstract)
In as much as a person can extract (copy with pen and paper) medical records, generate (write with pen and paper) a medical summary from the extracted records, classify (in their mind and with pen and paper) pre-physical status of patients using the medical summary, and visualize and provide (write down with pen and paper) a predictive basis, the claims are abstract under Mental Processes grouping of abstract ideas. See also MPEP 2106.04(a)(2) III C where using a computer was not enough to make abstract claims statutory.
This judicial exception is not integrated into a practical application. In particular, the claims only recite: natural language processing system (Claim 1); computer-readable recording medium, computer (Claim 12); electronic medical database (Claim 13). The computer hardware is recited at a high-level of generality (i.e., as a generic processor performing a generic computer function) such that it amounts no more than mere instructions to apply the exception using a generic computer component. The artificial intelligence based natural language processing system is recited at a high level of generality. Accordingly, these additional elements, when considered separately and as an ordered combination, do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea. Therefore claims 1, 12, and 13 are directed to an abstract idea without a practical application. (Step 2A-Prong 2: NO. The additional claimed elements are not integrated into a practical application)
The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception because, when considered separately and as an ordered combination, they do not add significantly more (also known as an “inventive concept”) to the exception. As discussed above with respect to integration of the abstract idea into a practical application, the additional element of using a computer hardware amounts to no more than mere instructions to apply the exception using a generic computer component. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. See Applicant’s specification pg. 24-25 about implantation using various computer components and MPEP 2106.05(f) where applying a computer as a tool is not indicative of significantly more. Accordingly, these additional elements, when considered separately and as an ordered combination, do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea. Steps such as providing (transmitting) are steps that are considered insignificant extra solution activity and mere instructions to apply the exception using general computer components (see MPEP 2106.05(d), II). Thus claims 1, 12, and 13 are not patent eligible. (Step 2B: NO. The claims do not provide significantly more)
Dependent claims 2-12 and 14-20 further define the abstract idea that is present in their respective independent claims 1 and 13 and thus correspond to Certain Methods of Organizing Human Activity and Mental Processes and hence are abstract for the reasons presented above. The dependent claims do not include any additional elements that integrate the abstract idea into a practical application or are sufficient to amount to significantly more than the judicial exception when considered both individually and as an ordered combination. Claims 4 and 16 recite large language processing system at a high level of generality. Claims 7 and 18 are further abstract under Certain Methods of Organizing Human Activity as commercial interaction by applying fees. Therefore, the claims 2-12 and 14-20 are directed to an abstract idea. Thus, the claims 1-20 are not patent-eligible.
Regarding claim 12 and signals
Claim 12 is rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter.
Claim 12 recites computer-readable recording medium having a program. Giving the claim its broadest reasonable interpretation, the claim may be directed at a signal, which is non-statutory.
Transitory signals are non-statutory (MPEP 2106.03 §I):
Non-limiting examples of claims that are not directed to one of the statutory categories:
Transitory forms of signal transmission (often referred to as "signals per se"), such as a propagating electrical or electromagnetic signal or carrier wave
For example, machine readable media can encompass non-statutory transitory forms of signal transmission, such as, a propagating electrical or electromagnetic signal per se. See In re Nuijten, 500 F.3d 1346, 84 USPQ2d 1495 (Fed. Cir. 2007). When giving the claim the broadest reasonable interpretation of machine readable media in light of the specification as it would be interpreted by one of ordinary skill in the art encompasses transitory forms of signal transmission, a rejection under 35 U.S.C. 101 as failing to claim statutory subject matter would be appropriate. Thus, a claim to a computer readable medium that can be a compact disc or a carrier wave covers a non-statutory embodiment and therefore should be rejected under 35 U.S.C. 101 as being directed to non-statutory subject matter.
The claim may be amended to include non-transitory computer readable medium.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claim 19 is rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
The term “long medical text” in claim 19 is a relative term which renders the claim indefinite. The term “long” is not defined by the claim, the specification does not provide a standard for ascertaining the requisite degree, and one of ordinary skill in the art would not be reasonably apprised of the scope of the invention. For examination purposes, long could be any length.
Examiner Request
The Applicant is requested to indicate where in the specification there is support for amendments to claims should Applicant amend. The purpose of this is to reduce potential 35 U.S.C. §112(a) or §112 1st paragraph issues that can arise when claims are amended without support in the specification. The Examiner thanks the Applicant in advance.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claims 1, 4-6, 8-13, 16, 17, 19, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Chung et al. – Ref. U (Chung et al, “Prediction of American Society of Anesthesiologists Physical Status Classification from preoperative clinical text narratives using natural language processing,” 2023, BMC Anesthesiology, pp. 1-16) in view of Pub. No. US 2020/0226321 to Burns and in view of Li et al. – Ref. V (Li et al., “A scoping review of using Large Language Models (LLMs) to investigate Electronic Health Records (EHRs), May 5, 2024, University of Michigan, 1-45)
Regarding claims 1, 12, and 13
(claim 1) A method for classification of pre-anesthetic physical status of patients, the method comprising:
extracting medical records from an electronic medical record database, the medical records comprising at least one or more of surgical information, hospitalization initial diagnosis, nursing initial diagnosis, hospitalization progress, vital signs, test results, and clinical observation records of the patients;
Chung et al. teaches:
Directly loaded (extracting) electronic health records with clinical picture of the patient…
Ref. U-1: “Machine learning and natural language processing (NLP) techniques, coupled with adoption of electronic health records (EHR), and widespread availability of high-performance computational resources offer new avenues for perioperative risk stratification whereby free-form text sources, such as medical notes, may be directly loaded into prediction models without the need to define, input or abstract predetermined data elements (e.g. diagnoses, medications, etc.). This offers the opportunity to use these techniques for preoperative assessment triage, flagging of critical/pertinent data in a voluminous electronic medical record, and a variety of other use cases based on clinician notes, which often contain narratives that richly and concisely describe a nuanced clinical picture of the patient while simultaneously prioritizing the clinician’s pertinent concerns. Unlike historical keywordbased approaches, modern NLP techniques using large pretrained language models are able to account for interword dependencies across the entire text sequence and have been shown to achieve state of the art performance on a variety of NLP tasks [1–4] including text classification [5, 6]. However, it is unknown whether these techniques can be successfully applied to perioperative risk stratification.” (pg. 2, col. 1, para. 1)
Example of surgery information…
Ref. U-3: “We compare the model's prediction against the ASA-PS assigned by the anesthesiologist on the day of surgery and assess catastrophic errors made by one of these models.” (pg. 2, col. 1, para. 2)
See Database below.
generating medical summary information of patients from the extracted medical record using an artificial intelligence-based natural language processing system;
Use (generating) values to visualize sections (summary information) of note using NLP models…
Ref. U-2: “… Finally, we use Shapley values to visualize which sections of note text were associated with the model’s predictions to explain these catastrophic errors. This approach shows that it is possible for clinicians to understand how complex NLP models are making their predictions, which is an important criteria for clinical adoption.” (pg. 2, col. 1, para. 2 – col. 2, para. 1)
Example of neural network (artificial language)…
Ref U-12: “Modern NLP techniques have overcome many of these challenges with vector space representation of words [12, 13, 34-36] and subword components [13, 19, 20, 37] as seen in the fastText model, attention mechanism [38, 39], and pretrained deep autoregressive neural networks [40-42] such as transformer neural networks [43]. This has resulted in successful large language models such as BERT [21, ,t4] and the domain-specific BioClinicalBERT [22]. Perhaps the most widely known large language model is ChatGPT {OpenAI, San Francisco, CA), a general purpose chatbot based on the GPT-3 model which contains 175 billion parameters [45]. In contrast, BioClinicalBERT used in this feasibility study contains roughly 1500 times fewer parameters, but has been trained specifically on clinical notes which makes it well suited for the ASA-PS prediction task [46].” (pg. 9, col. 1, par 3 – col. 2, para 1)
See Summary below.
classifying the pre-anesthetic physical status of the patients using the generated medical summary information as an input of a medical classification model; and
Using notes for preanesthesia…
Ref. U-2: “… These preoperative evaluation notes are a pertinent summary of the patient's medical and surgical history and describe why the patient is having surgery, all of which reflect the patient's preanesthesia medical comorbidities that the ASA-PS aims to represent…” (pg. 2, col. 1, para. 2)
Classifier using patient’s age (physical status) for predicting ASA-PS outcome (therefore pre-anesthetic)…
Ref. U-8: “Two baseline models were created for comparison: a random classifier model and an age & medications classifier model. The random classifier model generates a random prediction without using any features, thus serving as a negative control baseline. The age & medications classifier model serves as a simple clinical baseline model. It uses the patient’s age, medication list, and total medication count as input features to a multiclass logistic regression model with cross-entropy loss and L2 penalty for predicting the modified ASA-PS outcome variable. Defaults were used for all other model parameters. Both baselines were implemented using Scikit-learn.” (pg. 4, col. 1, para. 2 – col. 2, para. 1)
Use sections (summary information)…
Ref. U-3: “…We compare the model's prediction against the ASA-PS assigned by the anesthesiologist on the day of surgery and assess catastrophic errors made by one of these models. Finally, we use Shapley values to visualize which sections of note text were associated with the model's predictions to explain these catastrophic errors…”
See Summary below.
visualizing and providing a predictive basis of the medical classification model.
[No Patentable Weight is given to non-functional descriptive claim language of visualizing a predictive basis as this is interpreted as just displaying information.]
Fig. 3 teaches example of visualizing and providing a predictive basis.
PNG
media_image1.png
326
318
media_image1.png
Greyscale
Ref. U-3: “… Finally, we use Shapley values to visualize which sections of note text were associated with the model’s predictions to explain these catastrophic errors. This approach shows that it is possible for clinicians to understand how complex NLP models are making their predictions, which is an important criteria for clinical adoption.” (pg. 2, col. 1, para. 2 – col. 2, para. 1)
Database
Chung et al. teaches records. They do not literally teach database.
Burns et al. also in the business of records teaches:
Database…
“Institution assigned primary anesthesiology CPTs were used as the gold-standard labels developing the models. To assess for potential error with the gold-standard, a physician hand validation was conducted over 501 cases from the Train/Test dataset. In this process, 25 of 501 (5.0%) cases were found to be misclassified by primary anesthesia CPT when analyzed by physician hand auditing was compared to institution input CPT within the database. Nine of these cases were correctly identified by the SVM model.” [0065]
Fig. 1, ref. 11 and medical records…
PNG
media_image2.png
172
346
media_image2.png
Greyscale
Example of surgical procedures (information)…
“Within the practice of Anesthesiology, professional billing staff are responsible for selecting Current Procedure Terminology (CPT) codes to describe anesthesia care provided within the case. These CPT codes are based on surgical procedures performed. The process of assigning CPT codes is complicated and labor-intensive requiring various resources including specialized trained coding personnel for Electronic Medical Record (EMR) extraction, transcription, translation, coding assignment, validation, and auditing. Despite this, error rates in medical coding can be high. Studies have shown high rates of error for standard CPT coding in anesthesia with specialized teams, with error rates as high as 38%. As an alternative, when physicians independently code CPTs for their procedures error rates can be even higher, 54% in one study of interventional radiologists. Furthermore, modest gains in efficiency of billing process can have large effects on revenue—one study showed that a decrease of 10.1 days in accounts receivable or a charge lag decrease of 7.3 days resulted in a revenue gain equivalent to 3.0% of total annual receipts in a single academic anesthesiology practice.” [0005]
Receiving (extracting) medical procedure…
“In another aspect, the method for assigning billing codes includes: receiving a listing of possible billing codes, each billing code in the listing of possible billing codes includes a text description of a medical procedure associated with the billing code; receiving an input record describing a medical procedure, where the input record includes an input text description for the medical procedure; constructing a feature vector by extracting one or more features from the input record, where the input text description serves as a feature in the feature vector; for each billing code in the listing of possible billing codes, computing a classifier score for the feature vector using machine learning; for each billing code in the listing of possible billing codes, computing a term frequency-inverse document frequency (Tf-IDF) score for the input text description in relation to the text description for a given billing code in the listing of possible billing codes, where the Tf-IDF score for the given billing code is a summation of each score for each string in the input text description; for each billing code in the listing of possible billing codes, combining the Tf-IDF score with the classifier score to form a composite score; and assigning a billing code to the input record from the listing of possible billing codes based on the composite scores for each of the billing codes in the listing of possible billing codes.” [0009]
It would have been obvious to one of ordinary skill in the art before the effective filing date to include in the method and system of Chung et al. a database as taught by Burns et al. since the claimed invention is merely a combination of old elements and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable. Further motivation is provided by the need to store and retrieve medical records and a database provides a repository for accessing and storing such information.
Summary
The combined references teach text. They do not teach details of summary.
Li et al. also in the business of text teaches:
Text summarization…
Ref. V-1: “Natural Language Processing (NLP) techniques enable efficient processing of clinical notes and narratives and have been extensively used to process EHR data20,21. One popular area aims to enhance clinical documentation by extracting useful information from both structured data (e.g., lab test results and vital signs) and unstructured data (e.g., clinical notes) in EHRs. NLP techniques, such as named entity recognition22,23 and text summarization24, have been employed to extract entities relevant to patient health status and summarize treatment outcomes. The other area focuses on downstream applications of EHR data, such as the examination of disease progression25 and adverse drug reactions26. These studies can help enhance the understanding of medical conditions and therapeutic interventions.” (pg. 2, para. 3)
Extraction, text summarization, text classification…
Ref. V-2: “We annotated each paper to indicate the NLP tasks it included. This involved a pair-coding process, wherein two authors collaboratively engaged in the reading and annotation of each paper. We employed a pair-coding approach due to its efficacy in facilitating clarification of any ambiguities and expediting the determination of NLP tasks. Upon the conclusion of the initial round of pair coding, a preliminary list of NLP tasks was proposed to the entire team. Later, we held a group discussion to finalize the list of NLP tasks. As a result, we summarized seven major tasks (as illustrated in Figure 1(c)), including (1) named entity recognition, (2) information extraction, (3) text summarization, (4) text similarity, (5) text classification, (6) dialogue system, (7) diagnosis and prediction, and (8) others 13 (e.g., translation). Then, a second round of pair coding was conducted to confirm the NLP task for each paper. This iterative process ensured the accuracy and consistency of our paper annotation.” (pg. 12, para. last, - pg. 13, para. top)
It would have been obvious to one of ordinary skill in the art before the effective filing date to include in the method and system of the combined references the ability to summarize text as taught by Li et al. since the claimed invention is merely a combination of old elements and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable. Further motivation is provided by Li et al. and the benefits of summarizing text for analysis purposes.
Regarding claims 4 and 16
(claim 4) The method for classification of pre-anesthetic physical status of patients of claim 1, wherein the artificial intelligence-based natural language processing system comprises at least one of a large language model or a multi-agent collaboration network
Chung et al. teaches:
Large language models…
Ref. U-1: “… Unlike historical keyword based approaches, modern NLP techniques using large pretrained language models are able to account for interword dependencies across the entire text sequence and have been shown to achieve state of the art performance on a variety of NLP tasks [1–4] including text classification [5, 6]. However, it is unknown whether these techniques can be successfully applied to perioperative risk stratification.” (pg. 2, col. 1, para. 1)
Regarding claims 5 and 17
(claim 5) The method for classification of pre-anesthetic physical status of patients of claim 1, wherein visualizing a predictive basis of the medical classification model comprises visualizing an impact of each portion of the input text on the prediction results using model explainability methods,
Chung et al. teaches:
Visualize text with errors (impact) on model predictions…
Ref. U-3: “… Finally, we use Shapley values to visualize which sections of note text were associated with the model's predictions to explain these catastrophic errors. This approach shows that it is possible for clinicians to understand how complex NLP models are making their predictions, which is an important criteria for clinical adoption.” (pg. 2, col. 1, para. 2 – col. 2, para. 1)
wherein the contribution of key medical terms related to at least one or more of the patient's major disease, surgical history, current condition is highlight, and
Highlighting text…
Ref. U-11: “Shapley values in Fig. 7 provide clinically plausible explanations for model explanations, highlighting the directional probability of how specific input text contributes to predicting a specific ASA-PS. These feature attributions often provide clinically plausible explanations for why a model is making a wrong prediction and allows the clinician to evaluate the evidence the model is considering. Additional examples shown in Supplemental Figs. 2, :3, 4 and 5.” (pg. 8, col. 1 para. 2 – col. 2, para. 1)
separately displaying the contribution of each input text to the prediction result by different visual elements,
Fig. 7, teach separately displaying each input text using different visual elements…
Ref. U-16:
PNG
media_image3.png
120
386
media_image3.png
Greyscale
(pg. 12, Fig. 7, col. 1)
wherein a direction and a magnitude of the contribution is distinguishably visualized.
Direction and magnitude…
Ref. U-15:
PNG
media_image4.png
90
386
media_image4.png
Greyscale
(Fig. 7, pg. 12, col. 1)
Magnitude and direction…
“Fig. 7 Attribution of input text features to predicting modified ASA-PS for the BioClinicalBERT model on Note512 task, Shapley values for each text token is shown to compare feature attributions to ASA I (top) and feature attributions to ASA IV-V (bottom), Red tokens positively support predicting the target ASA-PS whereas blue tokens do not support predicting the target ASA-PS, The magnitude and direction of support is overlaid on a force plot above the text The baseline probability of predicting each class in the test set is shown as the "base value" on the force plot The base value+ sum of Shapley values from each token corresponds to the probability of predicting the ASA-PS and is shown as the bolded number, For simplicity, feature attributions to ASA II and Ill are omitted in this figure, but a full-visualization with all outcome ASA-PS for this text snippet is available in Supplemental Fig, i, Text examples are de-identified by replacing ages, dates, names, locations, and entities with pseudonyms to achieve data obfuscation while preserving structural similarity to the original passage” (pg. 12, para. 2)
Regarding claim 6
The method for classification of pre-anesthetic physical status of patients of claim 1, wherein the pre-anesthesia physical status of patients comprises an American Society of Anesthesiologists Physical Status (ASA-PS) class.
Chung et al. teaches:
ASA-PS…
Ref. U-2: “In this feasibility study, we hypothesize that NLP models can be applied to unstructured anesthesia preoperative evaluation notes written by clinicians to predict the American Society of Anesthesiologists Physical Status (ASA-PS) score [7, 8]…” (pg. 2, col. 1, para. 2)
Regarding claim 8
The method for classification of pre-anesthetic physical status of patients of claim 1, wherein for learning the medical classification model, the method comprises:
a training step using a first dataset;
Chung et al. teaches:
Train models…
Ref. U-5: “… Note section headers were excluded so that only the body of text from each section is included. We used text from each section to train models for ASA-PS prediction, resulting in 8 prediction tasks: Diagnosis, Procedure, HPI, PMSH, ROS, Medications (Meds), Note, Truncated Note (Note512). “Note” refers to using the whole note text as the predictor to train a model. When BioClinicalBERT is applied to the “Note” task, the WordPiece tokenizer [19–21] truncates input text to 512 tokens. This truncation does not occur for other models. For equitable comparison across models, we define the “Note512” task, which truncates the note text to the first 512 tokens used by the BioClinicalBERT model.” (pg. 3, col. 2, para. 1)
a tuning step using a second dataset; and
Model is tuned…
Ref. U-7: “Each model was trained on the training dataset. Model hyperparameters were tuned using Tune [23] with the BlendSearch [24, 25] algorithm to maximize Matthew’s Correlation Coefficient (MCC) computed on the validation dataset…” (pg. 3, col. 2, para. 3)
a testing step using a third dataset,
Test dataset…
Ref. U-9: “Final models were evaluated on the held-out test dataset by computing both class-specific and class-aggregate performance metrics…” (pg. 4, col. 2, para. 2)
wherein the second dataset and the third dataset comprise samples selected through a sampling method that takes into account each pre-anesthesia physical status class, and
Sample…
Ref. U-9: “… Class-aggregate performance metrics include MCC and AUCμ, [26] a multiclass generalization of the binary AUROC. Additionally, macro-average AUROC, AUPRC, precision, recall and F1 were also computed. Each metric and model-task combination was computed with 1000 bootstrap iterations each with 100,000 bootstrap samples on the test set. For each metric, p-values were computed for all 400 pairwise model-task comparisons with the Mann–Whitney U test followed by Benjamini–Hochberg procedure to control false discovery rate with α = 0.01.” (pg. 4, col. 2, para. 2 – pg. 5, col. 1, para. 1)
Pre-anesthesia…
Ref. U-4: “A unit of analysis is defined as a single case with an anesthesia preoperative evaluation note filed within 90 days of the procedure. This unit was chosen because ASA-PS is typically recorded on a per-case basis by the anesthesiologist to reflect the patient’s pre-anesthesia medical comorbidities at the time of the procedure…” (pg. 2, col. 2, para. 5)
are configured such that no data from the same patient is overlapped between the second dataset and the third dataset.
Held-out dataset (data withheld from training)…
Ref. U-9: “Final models were evaluated on the held-out test dataset by computing both class-specific and class-aggregate performance metrics. Class-specific metrics include: receiver operator characteristic (ROC) curve, area under receiver operator curve (AUROC), precision-recall curve, area under precision-recall curve (AUPRC), precision (positive predictive value), recall (sensitivity), and F1…” (pg. 4, col. 2, para. 2)
Regarding claim 9
The method for classification of pre-anesthetic physical status of patients of
Claim 8,
wherein a plurality of board-certified anesthesiologists independently perform an evaluation on the second dataset and the third dataset, and
Chung et al. teaches:
Ref. U-3: “… We compare the model’s prediction against the ASA-PS assigned by the anesthesiologist on the day of surgery and assess catastrophic errors made by one of these models…” (pg. 2, col. 1, para. 2)
See Board Certified below.
when the evaluation result is inconsistent, a reference label is generated by final agreement through additional board-certified anesthesiologist consultation,
Assess errors (result in inconsistent)…
Ref. U-3: “… We compare the model’s prediction against the ASA-PS assigned by the anesthesiologist on the day of surgery and assess catastrophic errors made by one of these models…” (pg. 2, col. 1, para. 2)
Visualize (therefore label) errors…
Ref. U-10: “Figure 5 depicts 4-by-4 contingency tables to visualize distribution of model errors on the Note512 task…” (pg. 7, col. 2, para. 2)
wherein the evaluation is performed after removing a part explicitly mentioned in the ASA-PS classification information from the medical summary information.
See Removing below.
Board Certified
The combined references teach sample. They also teach anesthesiologists. They do not teach board certified anesthesiologists.
Li et al. (Ref. V) also in the business of sample teaches:
Ref. V-3: “Private dataset: Dataset consisting of 6600 hospital admissions from 5000 unique patients at the inpatient neurology unit at New York Presbyterian/Weill Cornell Medical Center” (pg. 17,bottom para.)
Blind evaluation by 2 physicians…
Ref. V-3: “Best performance: ROUGE scores with an R-2 of 13.76. 62% of the auto-mated summaries meet the standard of care based on a blind-evaluation from 2 board-certified physicians.” (pg. 17, bottom para.)
It would have been obvious to one of ordinary skill in the art before the effective filing date to include in the method and system of the combined references the ability to use board-certified physicians as taught by Li et al. since the claimed invention is merely a combination of old elements and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable. Further motivation is provided by Li et al. who teaches the benefits of evaluating data with board-certified physicians.
The combined references teach anesthesiologist. They do not explicitly teach board certified anesthesiologist. However one of ordinary skill in the art would recognize that anesthesiologists are physicians.
It would have been obvious to one of ordinary skill in the art before the effective filing date of Applicant’s filing to modify the combined references with the knowledge available to such an artisan that anesthesiologist are physicians. This would have been known work in the field of endeavor prompting variations of it in the same field based on use of verifying data to ensure a models accuracy and would provide predictable results.
Removing
The combined references teach error. They do not teach removing from a summary.
Li et al. (Ref. V) also in the business of error teaches:
Filter out (remove) irrelevant information…
“LLMs can facilitate the extraction of information in a structured manner through named entity recognition, information extraction, and text classification. LLMs can help filter out irrelevant information and standardize relevant information (e.g., diseases, symptoms, medications, and treatments) into a consistent format213, which enables accurate interpretation as well as re-duces the risk of human-made errors.” (pg. 28, para. last)
It would have been obvious to one of ordinary skill in the art before the effective filing date to include in the method and system of the combined references the ability to filter out information as taught by Li et al. since the claimed invention is merely a combination of old elements and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable. Further motivation is provided by Li et al. who teaches the benefits of improving accuracy by filtering out irrelevant information.
It would have been obvious to one of ordinary skill in the art before the effective filing date to include in the method and system of the combined references the ability to filter out irrelevant information as taught by Li et al. since the claimed invention is merely a combination of old elements and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable. Further motivation is provided by Li et al. who teaches the benefits of accuracy by removing such information.
Regarding claim 10
The method for classification of pre-anesthetic physical status of patients of claim 1, wherein
a plurality of predictions are performed to estimate an uncertainty for the prediction of the medical classification model,
Chung et al. teaches:
Rarely made (estimate of uncertainty) of predictions…
Ref. U-14: “These models rarely made catastrophic errors. Erroneous predictions are typically adjacent to the ASA-PS assigned by the anesthesiologist, suggesting the model is making appropriate associations between freeform text predictors and the outcome variable (Fig. 5). Furthermore, when new anesthesiologist raters were asked to assign ASA-PS to the cases where catastrophic errors occurred from the BioclinicalBERT model on the Note512 task, there was greater concordance between the model predictions and the new anesthesiologist rather than the original anesthesiologist (Fig. 6)…” (pg. 11, col. 2, para .2)
wherein the uncertainty is estimated by distinguishing between the uncertainty due to variability inherent in the input data and the uncertainty due to the model parameters.
Example of longer text (input) and accurate prediction (input data caused uncertainty)…
Ref. U-13: “Longer text length provides more information for the model to make an accurate prediction. Even though text snippets such as Diagnosis or Procedure may have high relevance for the illness severity of the patient, the better performance on longer input text sequences indicate that more information is generally better. This is similar to what is observed in the multifaceted practice of clinical medicine–where a patient’s overall clinical status is often better understood as the sum of many weaker but synergistic signals rather than a single descriptor. The limited input sequence length for BioClinicalBERT creates a performance ceiling as it limits the amount of information available to the model. Comparing Note and Note512 tasks, all other models that can utilize the full note have better performance when this input length is lifted with fastText being the top performer…” (pg. 9, col. 2, para. 2 – pg. 10, col. 2, para. 1)
Model is tuned…
Ref. U-7: “Each model was trained on the training dataset. Model hyperparameters were tuned using Tune [23] with the BlendSearch [24, 25] algorithm to maximize Matthew’s Correlation Coefficient (MCC) computed on the validation dataset…” (pg. 3, col. 2, para. 3)
Example of parameter tuning to decrease parameter error…
Ref. U-7: “Each model was trained on the training dataset. Model hyperparameters were tuned using Tune [23] with the BlendSearch [24, 25] algorithm to maximize Matthew’s Correlation Coefficient (MCC) computed on the validation dataset. The number of hyperparameter tuning trials was selected to be 20 times the number of model hyperparameters with early stopping if the MCC of the last 3 trials reaches a plateau with standard deviation < 0.001. The best model was then evaluated on the held-out test dataset. Details on the approach taken for each of the four model architectures is available in Supplemental methods.” (pg. 3, col. 2, para. 3 – pg. 4, col. 1, para. 1)
Regarding claim 11
The method for classification of pre-anesthetic physical status of patients of claim 1, further comprising:
classifying the input text into subgroups according to the length of the text, and evaluating the performance of the medical classification model for each subgroup.
Chung et al. teaches:
Sections (subgroups)…
Ref. U-5: “Free-form text from the anesthesia preoperative evaluation note is organized into many sections. Regular expressions are used to extract HPI, PMSH, ROS, and medications from the note. While diagnosis and procedure sections exist within the note, they were less frequently documented than in the procedural case booking data from the surgeon. Therefore, free-form text for these sections were taken from the case booking…” (pg. 3, col. 1, para. 3)
Example of longer text vs snippets…
Ref. U-13: “Longer text length provides more information for the model to make an accurate prediction. Even though text snippets such as Diagnosis or Procedure may have high relevance for the illness severity of the patient, the better performance on longer input text sequences indicate that more information is generally better. This is similar to what is observed in the multifaceted practice of clinical medicine–where a patient’s overall clinical status is often better understood as the sum of many weaker but synergistic signals rather than a single descriptor. The limited input sequence length for BioClinicalBERT creates a performance ceiling as it limits the amount of information available to the model. Comparing Note and Note512 tasks, all other models that can utilize the full note have better performance when this input length is lifted with fastText being the top performer…” (pg. 9, col. 2, para. 2 – pg. 10, col. 2, para. 1)
Regarding claim 19
The system for classification of pre-anesthetic physical status of patients of claim 13, wherein the medical classification model unit comprises a transformer-based model capable of processing a long medical text.
Chung et al. teaches:
Ref. U-12: Transformer for large language models…
“Modern NLP techniques have overcome many of these challenges with vector space representation of words [12, 13, 34–36] and subword components [13, 19, 20, 37] as seen in the fastText model, attention mechanism [38, 39], and pretrained deep autoregressive neural networks [40–42] such as transformer neural networks [43]. This has resulted in successful large language models such as BERT [21, 44] and the domain-specific BioClinical-BERT [22]. Perhaps the most widely known large language model is ChatGPT (OpenAI, San Francisco, CA), a general purpose chatbot based on the GPT-3 model which contains 175 billion parameters [45]. In contrast, BioClinicalBERT used in this feasibility study contains roughly 1500 times fewer parameters, but has been trained specifically on clinical notes which makes it well suited for the ASA-PS prediction task [46].” (pg. 9, col. 1, para. 3 – col. 2, para. 1)
Longer text input…
Ref. U-13: “…These findings suggest that future development of a large language model similar to BioClinicalBERT capable of accepting a longer input context would likely have superior performance characteristics…” (pg. 10, col. 2, para. 1)
Regarding claim 20
The system for classification of pre-anesthetic physical status of patients of claim 13, wherein
the learning data of the medical classification model comprises samples selected through a sampling method that takes into account each pre-anesthesia physical status class, and
Chung et al. teaches:
Sample…
Ref. U-9: “… Class-aggregate performance metrics include MCC and AUCμ, [26] a multiclass generalization of the binary AUROC. Additionally, macro-average AUROC, AUPRC, precision, recall and F1 were also computed. Each metric and model-task combination was computed with 1000 bootstrap iterations each with 100,000 bootstrap samples on the test set. For each metric, p-values were computed for all 400 pairwise model-task comparisons with the Mann–Whitney U test followed by Benjamini–Hochberg procedure to control false discovery rate with α = 0.01.” (pg. 4, col. 2, para 2 – pg. 5, col. 1, para. 1)
Pre-anesthesia…
Ref. U-4: “A unit of analysis is defined as a single case with an anesthesia preoperative evaluation note filed within 90 days of the procedure. This unit was chosen because ASA-PS is typically recorded on a per-case basis by the anesthesiologist to reflect the patient’s pre-anesthesia medical comorbidities at the time of the procedure…” (pg. 2, col. 2, para. 5)
the samples comprise reference labels generated through evaluation by a plurality of board-certified anesthesiologists.
Assess errors (result in inconsistent)…
Ref. U-3: “… We compare the model’s prediction against the ASA-PS assigned by the anesthesiologist on the day of surgery and assess catastrophic errors made by one of these models…” (pg. 2, col. 1, para. 2)
Visualize (therefore label) errors…
Ref. U-10: “Figure 5 depicts 4-by-4 contingency tables to visualize distribution of model errors on the Note512 task…” (pg. 7, col. 2, para .2)
Board Certified
The combined references teach sample. They also teach anesthesiologists. They do not teach board certified anesthesiologists.
Li et al. (Ref. V) also in the business of sample teaches:
Ref. V-3: “Private dataset: Dataset consisting of 6600 hospital admissions from 5000 unique patients at the inpatient neurology unit at New York Presbyterian/Weill Cornell Medical Center” (pg. 17,bottom para.)
Blind evaluation by 2 physicians…
Ref. V-3: “Best performance: ROUGE scores with an R-2 of 13.76. 62% of the auto-mated summaries meet the standard of care based on a blind-evaluation from 2 board-certified physicians.” (pg. 17, bottom para.)
It would have been obvious to one of ordinary skill in the art before the effective filing date to include in the method and system of the combined references the ability to use board-certified physicians as taught by Li et al. since the claimed invention is merely a combination of old elements and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable. Further motivation is provided by Li et al. who teaches the benefits of evaluating data with board-certified physicians.
The combined references teach anesthesiologist. They do not explicitly teach board certified anesthesiologist. However one of ordinary skill in the art would recognize that anesthesiologists are physicians.
It would have been obvious to one of ordinary skill in the art before the effective filing date of Applicant’s filing to modify the combined references with the knowledge available to such an artisan that anesthesiologist are physicians. This would have been known work in the field of endeavor prompting variations of it in the same field based on use of verifying data to ensure a models accuracy and would provide predictable results.
Claims 2 and 14 are rejected under 35 U.S.C. 103 as being unpatentable over the combined references in section (11) above in further view of Kim et al. – Ref. X (KR 20230093754).
Regarding claims 2 and 14
(claim 2) The method for classification of pre-anesthetic physical status of patients of claim 1, wherein
the medical summary information comprises a text in which Korean and English are used interchangeably,
See Korean and English below.
generating the medical summary information of the patients further comprises performing an accurate translation of the medical terms with reference to a medical terminology dictionary before inputting into the medical classification model, and
See Korean and English below.
further comprising correcting a meaning of the medical terms according to the context and translating the text into English.
See Korean and English below.
Korean and English
The combined references teach medical summary. They do not teach translation.
Kim et al. also in the business of medical summary teaches:
Text summary…
Ref. X-1: “The deep learning model 112 may be designed to process text according to various purposes, such as text classification, text summary, text recognition/conversion, and text-related conversation generation.” (pg. 4, para. 7)
Korean-English translation with dictionary…
Ref. X-3 “In this case, the electronic device 100 may utilize at least one artificial intelligence model for performing Korean-English translation in addition to the aforementioned word dictionaries 111 and 405 (eg, conventional machine translation or statistical translation).” (pg. 5, para. 11)
Ref. X-4: “In this case, the electronic device 100 may identify (low quality) unified text resulting from an output having a difference of more than a threshold value from the above-described output closest to the correct answer, among the unified texts other than the selected unified text. In addition, the electronic device 100 may delete the second keyword related to the identified (low quality) integrated text from the keyword list matching the first keyword in the word dictionary 111 .” (pg. 7, para. 3)
Example of accuracy increase by using object (correcting according to context) into English…
Ref. X-2: “Here, the electronic device 100 may convert the first text into English in a state where “vital sign” is set as one object in the first text in which “vital sign” is replaced with “vital sign”. As such, when translation is performed in a state in which the “vital sign” itself secured according to the word dictionary 405 is fixed as an entity, the accuracy of translation may increase.” (pg. 5, para. 10)
It would have been obvious to one of ordinary skill in the art before the effective filing date to include in the method and system of the combined references the ability to translate text as taught by Kim et al. since the claimed invention is merely a combination of old elements and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable. Further motivation is provided by Kim et al. who teaches the benefits of translation and translation allows for more use pre-anesthesia models which provides a financial benefit of using correct anesthesia.
Claims 3 and 15 are rejected under 35 U.S.C. 103 as being unpatentable over the combined references in section (11) above in further view of Pub. No. US 2020/0381127 to Silverman.
Regarding claims 3 and 15
(claim 3) The method for classification of pre-anesthetic physical status of patients of claim 1 wherein generating the medical summary information of the patients comprises:
analyzing the extracted medical records to automatically identify the primary diagnosis and clinical condition of the patients;
Chung et al teaches:
Diagnosis loaded into prediction models (analyzing)…
Ref. U-1: “Machine learning and natural language processing (NLP) techniques, coupled with adoption of electronic health records (EHR), and widespread availability of high-performance computational resources offer new avenues for perioperative risk stratification whereby free-form text sources, such as medical notes, may be directly loaded into prediction models without the need to define, input or abstract predetermined data elements (e.g. diagnoses, medications, etc.)…” (pg. 2, col. 1, para 1)
matching the relevant drug use history and the previous surgical history based on the diagnosis content of the patients;
Including upload medications…
Ref. U-1: “Machine learning and natural language processing (NLP) techniques, coupled with adoption of electronic health records (EHR), and widespread availability of high-performance computational resources offer new avenues for perioperative risk stratification whereby free-form text sources, such as medical notes, may be directly loaded into prediction models without the need to define, input or abstract predetermined data elements (e.g. diagnoses, medications, etc.)…” (pg. 2, col. 1, para 1)
See Drug and Surgical below.
comparing and contrasting a plurality of test results with each other, and classifying the test results by diseases; and
See Drug and Surgical below.
See Classifying below.
integrating the analysis, matching and classification results into a standardized form of medical summary information.
See Drug and Surgical below.
See Classifying below.
Drug and Surgical
The combined references teach summary. They do not teach details of drug and surgical.
Silverman also in the business of summary teaches:
Medical records with diagnosis and condition…
“Whereas the detailed clinical information within the SHAPE™ database would be accessible for integration with a hospital's data network via established links based upon criteria such as name, birthdate and medical record number, inclusion of all data on each patient in a common database likely would be cumbersome. The present invention's consolidation of patient data enables efficient integration of vital data with links to sources of greater detail. FIG. 13 shows how the addition of approximately 15-30 columns increases the robustness of the clinical information in the database from the common listing of demographic data, costs and established codes (e.g., codes for International Classification of Disease (ICD-9) and Current Procedural Terminology (CPT) classifications)—the limitations of which are discussed later in this text—to body system by body system delineation of medical conditions (nature and severity) and, if deemed indicated, comparable delineation of surgical impact, risk indicators and integrated assessments. This provides ready access to such information for functions such as billing, resource allocation, predicting personnel and time requirements, quality assurance, auditing and research. The coding and scoring described herein enable simple linking to the details within the inclusive SHAPE™ database; e.g., more specifics as to the nature of a patient's class 3 disorder(s) of the Cardiac system as summarized at the end of the sample note (FIG. 4) and enabled by the ASPIRIN™ display (FIGS. 6 and 7). The optional linking to the detailed information—as may be indicated for detailed billing, quality assurance (e.g., to confirm adherence to diagnostic and treatment algorithms), medicolegal documentation and research—is mediated by a single cell (e.g., a patient's score for the Cardiac system) in the integrated hospital database, as opposed to a complex array of codes (e.g, ICD-9) which in and of themselves do not indicate disease severity and require a dictionary to identify the nature of the given disorder. To limit the size of the integrated database long-term, a patient's surgical procedure(s) could simply be consolidated according to its overall severity with an inventive code that integrates the risk/invasiveness score and the CPT code, with a link from that single cell to the itemization provided by the SICU™ score when more details are needed about a specific procedure.” [0663]
Cardiac risk and surgery and meds (drugs)…
“Looking at the currently available and the aforementioned SHAPE™ improvements, the six Cardiac Risk Indices variables are listed below along with quotes from relevant sections of FIG. 32 and Tables 11a and 11b that either are the given Risk Index variable or list a feature (with its score) which is suited to the given Risk Index variable. Note, the score provides additional information compared to the Cardiac Risk Index—precise features as well as score—while being programmed to maintain the simplicity of use. high-risk surgery: co-populate with entry in accordance with overall surgical risk/invasiveness (SOCU™) score of Table 11a; or more specifically, can co-populate with the score for the effect of surgery on the Cardiac system as a component of the SICU™ score (Table 11b). ischemic heart disease: co-populate with relevant features from “Ischemic Heart Disease” in FIG. 32 and their scores (in parentheses), including: exertional angina (3); old MI (3); stable (3); old wall motion abnormalities (3); extensive wall motion abnormalities and ischemia on scan (4); unstable angina (4); angina at ≤2 METS (4) congestive heart failure: co-populate with relevant features from “Congestive Heart Failure” in FIG. 32 and their scores, including: compensated CHF (3), stable on current meds (3); history of pulmonary edema in past—presently stable on current Rx (3); symptomatic CHF with moderate limitations (3), EF 25-50% (3); severe CHF (4); severe pulmonary edema (4); cannot perform any physical activity (4); requires cardiac meds to function (4) life-threatening pulmonary & hepatic congestion (4)EF<25% (4).” [0364] – [0367]
Classification based on testing severity score (disease)…
“ASA Physical Status. Because of its time-tested clinical utility, the ASA PS classification system has been chosen for use as the foundation for the present SHAPE™ representation of patient condition. However, in order to provide additional depth of information, while not forsaking its simple elegance, the present invention introduces the practice of assigning a 1 to 5 (or 0 to 5) severity score for each major organ system based on information learned from the patient history and physical examination and associated testing. The inventive system further provides the methods and algorithms to do so, thereby generating what is termed herein SHAPE™ Individual Systems Status (SISS™ or SIS™) scores (Table 4 and FIG. 32). The scoring criteria for each body system parallel the overall ASA PS classification but, by being body system-specific, are uniquely different and enable quantification of the ability of the given body system (for example, cardiac or respiratory) to withstand potential surgery-induced demands and insults—what is referred to herein as a SHAPE™ assessment of resilience.” [0157]
Standardized notes…
“Moreover, TUNAS can be adapted to exceed the objectives put forth in Table 1. In accordance with the question raised more than 10 years ago—“is it time for our specialty (anesthesiology) to develop a standardized preoperative assessment . . . and postoperative visit?” (T1116) TUNAS can provide the foundation for universal preoperative and post-operative notes as well as notes throughout a patient's healthcare management. One also can provide for additions (e.g., nonscaled descriptors) so TUNAS can generate a standalone document for handoffs, morning rounds or even discharge summaries as well as wallet card and chips based on cutoffs (individual and integrated). This can drive testing as well as prompt review of relevant textual descriptions (as per current progress notes) and potentially link to relevant entries within TUNAS and throughout the electronic health record. In addition, TUNAS can provide the patient with a succinct, readily viewable scaling “selfie.” It thus achieves (and exceeds) Center for Medicare and Medicaid Services' recommendations/requirements for meaningful clinical, administrative and investigative electronic health record use (Blumenthal D, Tavenner M. The “meaningful use” regulation for electronic health records. N Engl J Med. 2010; 363(6):501-4).” [0803]
It would have been obvious to one of ordinary skill in the art before the effective filing date to include in the method and system of the combined references the ability use surgical and drug data as taught by Silverman since the claimed invention is merely a combination of old elements and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable. Further motivation is provided by Silverman who teaches the benefits of analyzing surgical and drug information for anesthesia purposes.
Classifying
The combined references teach drug and surgical. They do not specifically teach classifying and standardize text.
Li et al. also in the business of classifying teaches:
Classify text…Ref. V-4: “Text classification for EHR data involves categorizing unstructured or semi-structured text data into predefined categories or classes. Given that a significant portion of EHR data is in the form of free text, such as clinical notes written by healthcare providers, text classification can help obtain meaningful insights from EHR data. However, it is important to note that text classification generally requires pre defined categories and assigns a category to the entire text, which differentiates it from either NER or information extraction. The pre-defined categories used in text classification can vary widely depend ing on specific use cases, ranging from medical events158 to disease types to patient statuses120.” (pg. 23, para. 2)
Standardize relevant information…
Ref. V-5: “LLMs can facilitate the extraction of information in a structured manner through named entity recognition, information extraction, and text classification. LLMs can help filter out irrelevant information and standardize relevant information (e.g., diseases, symptoms, medications, and treatments) into a consistent format213, which enables accurate interpretation as well as reduces the risk of human-made errors.” (pg. 28, para. last)
It would have been obvious to one of ordinary skill in the art before the effective filing date to include in the method and system of the combined references the ability classify and standardize text as taught by Li et al. since the claimed invention is merely a combination of old elements and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable. Further motivation is provided by Li et al. who teaches the benefits of classifying and standardizing text for analysis purposes.
Claims 7 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over the combined references in section (11) above in further view of Curatolo et al. – Ref. W (Curatolo et al, “ASA physical status assignment by non-anthesia providers: Do surgeons consistently downgrade the ASA score preoperativelty?,” 2017, Journal of Clinical Anesthesia,” pp. 123-128)
Regarding claims 7 and 18
(claim 7) The method for classification of pre-anesthetic physical status of patients of claim 6, wherein the ASA-PS class is classified into one of classes I, II, III, IV-V, and further comprising
See Classes below.
determining whether to apply additional anesthesia fees for the patients classified as ASA-PS class III or higher.
See Classes below.
Classes
The combined references teach ASA-PS class. They do not teach fees.
Curatolo et al. also in the business of ASA-PS class teaches:
ASA-PS and class 2 vs 3 and costly (fees)…
Ref. W-1: “… However, the requirement does not always include a visit to the anesthesia clinic prior to surgery. As a result, patients are assigned a presumed ASA-PS by a nonanesthesia provider (e.g., surgeons and physician extenders) that may not reflect the A')A-PS chosen by the anesthesiologist on the day of surgery. Errors in the accuracy of this assignment (especially when assigning ASA 2 vs. 3) can lead to over-prescribing of preoperative testing (resulting in unnecessary and costly tests) or under prescription of testing (potentially resulting in operative delays or case cancellations on the day of surgery) [ 12, 13 ]. Cost concerns are becoming more salient and prior studies have demonstrated operating room case cancellations as representing thousands of dollars lost per cancelled case and were noted to be higher among patients without a pre-operative clinic visit ! 14 ]. Thus we aimed to study assignment by non-anesthesia providers as they are often assigning the ASA-PS classification to patients that will dictate whether or not a pre-operative clinic visit is required at our institution.” (pg. 124, col. 1, para. 1)
It would have been obvious to one of ordinary skill in the art before the effective filing date to include in the method and system of the combined references the ability to determine anesthesia costs as taught by Curatolo et al. since the claimed invention is merely a combination of old elements and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable. Further motivation is provided by the financial benefits of Curatolo et al. and the costs of using a higher class for anesthesia.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
The following prior art teaches at least anesthesia:
CN-119339954-B; AU-2022481770-A1; CN-119007922-A; CN-119339954-A; WO-2024080977-A1; EP-4478370-A1; CN-120048442-A; CN-119423705-A; CN-118737438-A; CN-118236072-A; US-11288445-B2; US-10796801-B2; US-20150039339-A1
Any inquiry concerning this communication or earlier communications from the examiner should be directed to KENNETH BARTLEY whose telephone number is (571)272-5230. The examiner can normally be reached Mon-Fri: 7:30 - 4:00 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 at (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 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.
/KENNETH BARTLEY/Primary Examiner, Art Unit 3684
Prosecution Insights