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 .
Information Disclosure Statement
The information disclosure statements (IDS) submitted on 11/01/2025 are in accordance with the provisions of 37 CFR 1.97 and are considered by the Examiner.
Priority
Claim PCT/KR2023/010539 07/21/2023, FOREIGN APPLICATIONS KOREA, REPUBLIC OF 10-2022-0090767 07/22/2022 KOREA, REPUBLIC OF 10-2023-0093176 are acknowledged.
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.
Claim 12 invokes § 112(f) because recites a network unit configured to acquire electrocardiogram data and cardiac arrest data. Unit is a generic placeholder equivalent to means; network names the data environment, not the structure of the unit; and configured to acquire states pure function.
Because this/these claim limitation(s) is/are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, it/they is/are being interpreted to cover the corresponding structure described in the specification as performing the claimed function, and equivalents thereof.
network unit:
The spec supplies the structure: “network unit 130” receives data by “wired or wireless communication” from a “cloud server,” “smart watch,” or medical computing devices. Thus, it is hardware/network-interface structure with optional firmware/software control. The “processor including at least one core” is hardware executing “program codes”; Refer spec, pag. 13-14.
If applicant does not intend to have this/these limitation(s) interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, applicant may: (1) amend the claim limitation(s) to avoid it/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 limitation(s) recite(s) 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.
Non-Statutory Rejection:
Claim 11 is rejected under 35 U.S.C. 101 because recites non-statutory category under step 1. The claim recites “A computer program” stored in a “computer-readable storage medium,” but does not limit the medium to non-transitory storage. Under BRI, the medium encompasses transitory signal embodiments, including carrier waves. Because the claim covers both statutory storage articles and non-statutory transitory signals, the claim as a whole is not limited to a process, machine, manufacture, or composition of matter. For compact prosecution and in assumption of correct amendment to recite non-transitory storage, examiner continue with the subject eligibility analysis.
Subject Matter Eligibility Rejection:
Claims 1 to 12 are rejected under 35 U.S.C. 101. The claims are directed to an abstract idea, the clinically informed curation of electrocardiogram training data and prediction from it, carried out as a mental process combined with mathematical concepts. The only elements beyond that idea are a generic computing device, processor, memory, network unit, computer program, storage medium, and a generically applied deep learning model. Those elements neither integrate the exception into a practical application nor add an inventive concept, so the claims do not recite eligible subject matter.
Step 1.
Step 1 asks whether the claims fall within a statutory category under 35 U.S.C. § 101.
Claims 1 to 10 recite a method of acts and qualify as a process. Claim 12 recites a processor, a memory, and a network unit and qualifies as a machine. Claim 11 recites a computer program that is stored in a computer-readable storage medium and qualifies as a manufacture in view of compact prosecution.
Each claim falls within a statutory category under MPEP 2106.03, so the analysis proceeds to Step 2A.
Step 2A,
Independent Claims Analysis
Prong One asks whether the claim language recites a judicial exception.
Representative independent claim 1 (bold marks the additional elements; plain text is the abstract idea)
A method of predicting a health state using an electrocardiogram, performed by a computing device including at least one processor, comprising:
[1] acquiring electrocardiogram data and cardiac arrest data including whether cardiac arrest occurs in a subject whose electrocardiogram data is measured and a cardiac arrest occurrence time of the subject;
[2] analyzing a missing value of the electrocardiogram data or the cardiac arrest data to extract valid data from the electrocardiogram data; and
[3] labeling the extracted valid data based on whether the cardiac arrest occurs and the cardiac arrest occurrence time to generate training data.
Claims 11 and 12 mirror claim 1 in program and device form. Their only difference is the bold structural packaging, a computer program ... stored in a computer-readable storage medium ... executed on one or more processors (claim 11) and a processor including at least one core; a memory including program codes ...; and a network unit (claim 12). The plain-text acquiring, analyzing, and labeling recites similar functions under BRI. Claim 9 is a second independent claim on the inference side: it adds the bold inputting the first electrocardiogram data to a pre-trained deep learning model, while its plain-text wherein clauses recite the same labeling valid data ... based on whether cardiac arrest occurs curation.
Limitations 1 to 3 recite a mental process under MPEP 2106.04(a)(2)(III). acquiring electrocardiogram data and cardiac arrest data is collecting information; analyzing a missing value ... to extract valid data is an evaluation and judgment of which records are usable; and labeling the extracted valid data based on whether the cardiac arrest occurs and the cardiac arrest occurrence time is a classification. Claim 9's predict the health state is likewise an evaluative judgment.
For example, A physician handed a stack of electrocardiogram printouts and an arrest log performs every plain-text step by hand. She reads each tracing and its arrest note (acquire). She sets aside any tracing whose data is incomplete or unusable (analyze a missing value to extract valid data). She sorts the rest into an arrest pile and a normal pile keyed to when arrest occurred (label to generate training data). For claim 9, she looks at a new tracing and judges the patient's risk (predict). Pen, paper, and the clinician's judgment carry the work.
Dependent Claims Analysis
Each dependent claim narrows the same abstract idea, grouped here by mirror relationship:
Claims 2 to 4 recite, under their broadest reasonable interpretation, choosing a time boundary: analyzing the cardiac arrest occurrence time ... and clinical evidence ... to determine a cutoff time, picking that cutoff as one of three candidate times (claim 3), and selecting the largest value among the first candidate time, the second candidate time, or the third candidate time (claim 4). These fit mental processes and mathematical concepts under MPEP 2106.04(a)(2)(I) and (III), because a person can weigh clinical knowledge and compare three minute-values by hand. Per MPEP 2106.04(II)(B) they combine with claim 1 as a single abstract idea.
Claim 5 recites labeling, as a cardiac arrest group, and labeling, as a normal group, within a preset time, which is a classification and therefore a mental process.
Claims 6 and 7 recite excluding the electrocardiogram data when whole derivations are missing (claim 6) or when the missing value exists for a predetermined period of time or longer (claim 7). These are evaluation and filtering judgments, mental processes.
Claim 8 recites inputting the training data to a deep learning model to train the deep learning model. The training itself is a mathematical and evaluative act; the deep learning model is an additional element taken up in Prong Two.
Claim 10 recites a score value that indicates a healthy state as it approaches a first threshold value, and indicates an unhealthy state as it approaches a second threshold value, a mathematical relationship and a mental comparison.
Because every independent and dependent claim recites a mental process, in places combined with mathematical concepts, the claims recite an abstract idea. The analysis proceeds to Prong Two.
Step 2A, Prong Two: Integration Into a Practical Application
Prong Two asks whether the elements beyond the abstract idea impose a meaningful limit that integrates the exception into a practical application through a technical improvement, not a stated goal or result (MPEP 2106.04(d)).
Additional elements
The additional elements are the generic computing wrappers: claim 1's a computing device including at least one processor; claim 11's a computer program ... stored in a computer-readable storage medium ... executed on one or more processors; claim 12's a processor including at least one core, a memory including program codes, and a network unit; and claim 9's pre-trained deep learning model.
The computing device, processor, memory, and network unit are recited only as tools that carry out the acquiring, analyzing, and labeling. They supply no technical mechanism; the claim says what to compute, not an improvement in how a computer works, so they amount to applying the abstract idea on a general-purpose computer under MPEP 2106.05(f). The network unit's acquire electrocardiogram data is mere data gathering, insignificant extra-solution activity under MPEP 2106.05(g). The pre-trained deep learning model of claims 8 and 9 is invoked as a black box: the claim recites inputting data to it to predict, with no improvement to the model architecture or to electrocardiogram technology, which is apply-it under MPEP 2106.05(f). The asserted advance lies in smarter data curation, which refines the abstract idea itself rather than any technology, and so does not integrate under MPEP 2106.05(a).
Considered together, the device, processor, memory, network unit, and model do no more than perform the same mental curation. Their ordered combination still gathers, evaluates, classifies, and predicts on generic hardware and imposes no meaningful limit beyond the exception.
Dependent Claims Analysis
Claims 2 to 7 and 10 add only further abstract steps, namely cutoff selection, group labeling, lead exclusion, and scoring, and introduce no additional elements, so they cannot integrate the exception. Claims 8 and 9 add the deep learning model, which does not overcome Prong Two because the model is used as a generic predictive tool with no recited technical improvement (MPEP 2106.05(a) and (f)).
The additional elements, alone and combined, do not integrate the abstract idea into a practical application. The analysis proceeds to Step 2B.
Step 2B: Inventive Concept
Step 2B asks whether the additional elements, individually and as an ordered combination, add significantly more than the abstract idea (MPEP 2106.05).
The computing device and processor of claims 1 and 12, the memory and network unit of claim 12, the computer program and computer-readable storage medium of claim 11, and the pre-trained deep learning model of claims 8 and 9.
The specification admits each hardware element is generic. The processor may include a central processing unit (CPU), a general purpose graphics processing unit (GPGPU), a tensor processing unit (TPU), an application specific integrated circuit (ASIC), or a field programmable gate array (FPGA), etc. (Spec., pag.10). The memory may include at least one of storage media such as a flash memory type, a hard disk type (Spec., page 13.]). The network unit transmits and receives data through any known wired/wireless communication system (Spec., page 13 ll. 25 – page 14 ll. 5). These show the elements add nothing beyond generic computer function under MPEP 2106.05(d). The deep learning model is defined by the applicant as a system implemented using mathematical concepts and language to solve a specific problem (Spec., page 8), which is itself part of the exception rather than significantly more.
Taken as an ordered whole, the elements recite generic acquisition by a known network, generic storage in generic memory, generic computation by a generic processor, and generic prediction by a model the applicant defines in mathematical terms. Each piece performs its ordinary function and none operates in an unconventional way, so the combination supplies no inventive concept under MPEP 2106.05.
Dependent Claims Analysis
Claims 2 to 7 and 10 add no new additional elements; they only narrow the abstract idea through cutoff derivation, group labeling, lead exclusion, and score thresholds, and so add no inventive concept. Claims 8 and 9 add the deep learning model, which does not overcome Step 2B because the applicant describes it as implementing mathematical concepts and language and uses it as a generic predictive tool, adding nothing significantly more (Spec., para. [0033]; MPEP 2106.05(a) and (f)).
Claims 1 to 12 are directed to an abstract idea, the clinically informed curation of electrocardiogram training data and prediction from it, performed as a mental process combined with mathematical concepts. The additional elements are generic computing components and a generically applied model that neither integrate the exception into a practical application nor add an inventive concept. Claims 1 to 12 are therefore rejected under 35 U.S.C. 101.
Claim Rejections - 35 USC § 102
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
Claim(s) 1-3, and 8-10 are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Kwon et al., Artificial intelligence algorithm for predicting cardiac arrest using electrocardiography, Scandinavian Journal of Trauma, Resuscitation and Emergency Medicine, 28:98, 2020. Refer to PTO-892-U
[Claim 1]
Kwon teaches,
A method of predicting a health state using an electrocardiogram, performed by a computing device including at least one processor, comprising: (Kwon, abstract, Title: Artificial intelligence algorithm for predicting cardiac arrest using electrocardiography, fig 2, page 4))
acquiring electrocardiogram data and cardiac arrest data including whether cardiac arrest occurs in a subject whose electrocardiogram data is measured and a cardiac arrest occurrence time of the subject; (Kwon, fig. 1, Input (ECG), ECG data were recorded, page 2, methods, endpoint of this research was cardiac arrest …We reviewed electronic health records to identify the exact time of each endpoint, page 3 methods, fig 2, page 4))
analyzing a missing value of the electrocardiogram data or the cardiac arrest data to extract valid data from the electrocardiogram data; (Kwon, 12-lead ECG acquired in the supine position during the admission period. We excluded subjects with missing demographic or electrocardiographic information, page 2, We also excluded subjects in hospital B with missing value, page 3, fig 2, page 4))
and labeling the extracted valid data based on whether the cardiac arrest occurs and the cardiac arrest occurrence time to generate training data.( We made a dataset using the entire 12-lead ECG data., page 2, For a patient with cardiac arrest, the ECGs belonging to the prediction window were labeled as cardiac arrest and other ECGs were labeled as a nonevent, page 3, fig 2, page 4)
[Claim 2]
Kwon teaches,
The method of claim 1, wherein the analyzing of the missing value of the
electrocardiogram data or the cardiac arrest data to extract the valid data from the electrocardiogram data includes:
analyzing the cardiac arrest occurrence time recorded in the cardiac arrest data and clinical evidence regarding cardiac arrest to determine a cutoff time for filtering the electrocardiogram data; (Kwon, We reviewed electronic health records to identify the exact time of each endpoint, page 2, ECG was within the prediction time window of cardiac arrest, which is the 24 h interval before cardiac arrest, page 3)
Kwon expressly records "the exact time of each endpoint" (cardiac arrest occurrence time) and grounds the 24-hour window in the clinical observation that "80% of patients show signs of deterioration before cardiac arrest." From that clinical evidence and the recorded endpoint time, Kwon determines "the prediction time window of cardiac arrest, which is the 24 h interval before cardiac arrest" that 24-hour interval is the cutoff time for filtering the ECG data. The claim adds no structural feature absent from Kwon.
and filtering the electrocardiogram data of the subject who suffers from the cardiac arrest within the cutoff time among the acquired electrocardiogram data to extract the valid data.(Kwon, ECGs belonging to the prediction window were labeled as cardiac arrest and other ECGs were labeled as a nonevent., page 3, 12-lead ECG acquired in the supine position during the admission period. We excluded subjects with missing demographic or electrocardiographic information, page 2)
Kwon teaches filtering ECG data by cardiac-arrest time: ECGs within the selected pre-arrest window are labeled as cardiac arrest, while others are labeled differently. Kwon also excludes records with missing demographic or electrocardiographic information.
[Claim 3]
Kwon teaches,
The method of claim 2, wherein the analyzing of the cardiac arrest occurrence time recorded in the cardiac arrest data and the clinical evidence regarding the cardiac arrest to determine the cutoff time for filtering the electrocardiogram data includes determining, as the cutoff time, one of a first candidate time derived based on clinical determination on a pattern of change in the electrocardiogram that occurs before the cardiac arrest occurs, a second candidate time clinically determined as a time at which treatment is possible to perform to prevent the cardiac arrest when the cardiac arrest is predicted to occur, and a third candidate time corresponding to an error between the cardiac arrest occurrence time recorded in the cardiac arrest data and an actual cardiac arrest occurrence time of the subject whose electrocardiogram data is measured.(Kwon, “We reviewed electronic health records to identify the exact time of each endpoint” (page 2), ECG was within the prediction time window of cardiac arrest, which is the 24 h interval before cardiac arrest. For a patient with cardiac arrest, the ECGs belonging to the prediction window were labeled as cardiac arrest” (page 3), 80% of patients show signs of deterioration before cardiac arrest, diverse rapid response systems (RRSs) have been implemented to prevent cardiac arrest in the past (page 1, Previous studies found QT prolongation, QRS prolongation, fragmented QRS complexes, and early repolarization to be associated with cardiac arrest (page 2)), figure 2, figure 5, patterns for predicting cardiac arrest (page 6))
Kwon teaches the first candidate time because Kwon identifies the exact cardiac-arrest endpoint time from electronic health records and defines the prediction time window as the 24-hour interval before cardiac arrest.
[Claim 8]
Kwon teaches,
The method of claim 1, further comprising inputting the training data to a deep learning model to train the deep learning model so that the deep learning model calculates the possibility of occurrence of the cardiac arrest. (Kwon, For a patient with cardiac arrest, the ECGs belonging to the prediction window were labeled as cardiac arrest and other ECGs were labeled as a nonevent… DLA calculated the possibility of cardiac arrest, page 3)
[Claim 9]
Kwon teaches,
A method of predicting a health state using an electrocardiogram, performed by a computing device including at least one processor, comprising: (refer claim 1)
acquiring first electrocardiogram data; (refer claim 1) and
inputting the first electrocardiogram data to a pre-trained deep learning model to predict the health state of a subject whose electrocardiogram data is measured, (Kwon, p.3–4: At each input (ECG, age, and sex) of the validation data, the DLA calculated the possibility of cardiac arrest in the range from 0 (nonevent) to 1 (cardiac arrest))
Kwon's validation step is the claimed inference: the validation ECG is the first electrocardiogram data, the pre-trained DLA is the model, and computing the arrest possibility is predicting the health state.
wherein the deep learning model is trained based on training data generated by labeling valid data extracted from second electrocardiogram data based on whether cardiac arrest occurs and a cardiac arrest occurrence time (Kwon, labeled as cardiac arrest and other ECGs were labeled as a nonevent, …identify the exact time of each endpoint …page 3, )
Kwon labels every training ECG by whether it falls inside the 24-hour pre-arrest window — a label that is simultaneously outcome-based (arrest vs. no arrest) and time-anchored (exact endpoint time). The DLA is then trained on that labeled corpus.
, and
the valid data is extracted from the second electrocardiogram data based on results of analyzing a missing value of the second electrocardiogram data or cardiac arrest data including whether the cardiac arrest occurs in a subject whose second electrocardiogram data is measured and a cardiac arrest occurrence time of the subject. (Kwon, missing demographic or electrocardiographic, page 2, excluded subjects in hospital B page 3, occurrence of cardiac arrest within 24, page 4, figure 2)
Kwon screens the training corpus on both sides of the "or." It checks the ECG data (excludes records with missing electrocardiographic information) and checks the cardiac arrest data (endpoint records noting whether arrest occurred and its time). Records passing both screens form the valid training set
[Claim 10]
Kwon teaches,
The method of claim 9, wherein the prediction result derived by inputting the first electrocardiogram to the pre-trained deep learning model includes a score value indicating the health state of the subject whose electrocardiogram data is measured, (Kwon, the values of the output node represent the possibility of developing cardiac arrest, and the output node uses a sigmoid function as an activation function, as the output of the sigmoid function is between 0 and 1 page 2 the DLA calculated the possibility of cardiac arrest in the range from 0 (nonevent) to 1 (cardiac arrest) page 4
calculation of the risk score by the DLA page 7)
Kwon's DLA produces a continuous numerical output that directly reflects the patient's cardiac health state.
and the score value indicates a healthy state as it approaches a first threshold value, (Kwon, the range from 0 (nonevent) to 1 (cardiac arrest) page 4 true negative (low risk) versus the false positive (high risk) groups over time page 4 all ECGs were labeled as a nonevent" (for subjects without cardiac arrest) page 3)
Kwon maps score 0 to "nonevent” the non-arrest, healthy state. Patients whose scores fall below the cutoff are classified as low risk. A score approaching 0 signals no cardiac arrest, which is the healthy condition. Kwon expressly separates this population as the "true negative (low risk)" group.
and indicates an unhealthy state as it approaches . Kwon, the DLA calculated the possibility of cardiac arrest in the range from 0 (nonevent) to 1 (cardiac arrest).” Page 3 and 4, figure 2)
Kwon discloses a single prediction score whose low endpoint indicates nonevent/healthy state and whose high endpoint indicates cardiac arrest/unhealthy state.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries 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 4 and 5 rejected under 35 U.S.C. 103 as being unpatentable over Kwon et al. in view of US20200337622A1-Zhang et al.
[Claim 4]
Kwon teaches,
The method of claim 3,
.
Kwon teaches the base cutoff framework refer claim 3.
Kwon does not teaches,
Zhang teaches the missing maximum-selection rule. Zhang first explains why assumed timing can be wrong: If the test is too long, or not long enough, the linear descension from alert to fatigue is inaccurate. Zhang, [0006]. Zhang then teaches comparing candidate physiological signal segments and replacing the assumed maximum when another candidate is greater: the score for the segment of maximum fatigue is greater than that for the assumed maximum fatigue segment; and setting the segment of maximum fatigue as a revised assumed maximum fatigue segment. Zhang, [0017], [0012], fig.1. Zhang states the same rule again in process form: The score of the segment with the highest score is compared and the higher-scoring maximum is a more appropriate upper bound. Zhang, [0090]-[0092].
A POSITA would apply Zhang’s maximum-candidate replacement rule to Kwon’s ECG cutoff candidates because both references solve the same data-quality problem: avoiding an inaccurate assumed boundary in physiological-signal model training. The predictable result is a cleaner ECG training set using the largest cutoff that excludes all data any candidate identifies as unreliable.
[Claim 5]
Kwon teaches,
The method of claim 4, wherein the labeling of the extracted valid data based on whether the cardiac arrest occurs and the cardiac arrest occurrence time to generate the training data includes:
labeling, as a cardiac arrest group, data of the subject who suffers from the cardiac arrest within a preset time after the cutoff time in the valid data; (Kwon, For a patient with cardiac arrest, the ECGs belonging to the prediction window were labeled as cardiac arrest. page 3)
Kwon, identifying ECGs that fall within the 24-hour prediction window before a cardiac-arrest event and labeling those ECGs as cardiac arrest.
and labeling, as a normal group, data of the subject who does not suffer from cardiac arrest or die within the preset time after the cutoff time in the valid data. (Kwon, other ECGs were labeled as a nonevent. For a patient without cardiac arrest, all ECGs were labeled as a nonevent, page 3)
Kwon performs the same classification by labeling ECGs outside the cardiac-arrest window, and ECGs from patients without cardiac arrest, as nonevents.
Claim(s) 6, 7, 11, and 12 are rejected under 35 U.S.C. 103 as being unpatentable over Kwon et al. in view of US20200305799A1-Cao et al.
[Claim 6]
Kwon teaches,
The method of claim l, wherein the analyzing of the missing value of the electrocardiogram data or the cardiac arrest data to extract the valid data from the electrocardiogram data .
Kwon teaches missing-ECG-data exclusion generally, but does not expressly teach lead-level exclusion.
Cao teaches analyzes ECG data by lead, uses lead parameters, obtains classification information for each lead, and screens ECG event data according to a signal-quality evaluation index. Cao, [0013]-[0014], [0019], [0032]-[0033], [0100].
It would have been obvious to exclude, from Kwon’s valid training data, ECG records in which a required derivation/lead itself is absent, because Kwon already excludes missing ECG information and Cao teaches lead-specific screening to prevent invalid or unreliable ECG data from entering downstream analysis.
[Claim 7]
Kwon teaches,
The method of claim1, wherein the analyzing of the missing value of the electrocardiogram data or the cardiac arrest data to extract the valid data from the electrocardiogram data includes excluding, from the valid data, the electrocardiogram data
Kwon teaches missing-ECG-data exclusion generally, but does not expressly teach strikethrough language above.
Cao teaches identifies data segments meeting a preset interval threshold, judges whether those segments are abnormal, performs sliding sampling using a preset time width, and evaluates signal quality using threshold-based screening. Cao, [0025]-[0028], [0084]-[0089], [0144]. Cao’s abnormal-signal examples include electrode peeling off, low voltage, and signal overflow, which correspond to absent or invalid ECG signal values over a segment of time. Cao, [0086], [0089].
A person of ordinary skill would have found it obvious to apply Cao’s preset-interval/time-width abnormal-signal screening to Kwon’s missing-value exclusion, thereby excluding ECG data when a lead/derivation contains absent or invalid values for a predetermined duration. The combination merely applies Cao’s known time-resolved ECG quality control to Kwon’s known ECG training-data curation, with the predictable result of retaining only complete and reliable ECG data for model training. Claim 7 is therefore unpatentable over Kwon in view of Cao.
.
[Claim 11]
Kwon teaches,
acquiring electrocardiogram data and cardiac arrest data including whether cardiac arrest occurs in a subject whose electrocardiogram data is measured and a(refer claim 1)
cardiac arrest occurrence time of the subject; (refer claim 1)
analyzing a missing value of the electrocardiogram data or the cardiac arrest data to extract valid data from the electrocardiogram data; (refer claim 1) and
labeling the extracted valid data based on whether the cardiac arrest occurs and the cardiac arrest occurrence time to generate training data. (refer claim 1)
Kwon teaches the claimed operations because Kwon uses a DLA to predict cardiac arrest from ECG data, excludes missing ECG information, and labels ECGs in the arrest prediction window as cardiac arrest or nonevent. Kwon does not expressly teach a computer program stored in a computer-readable storage medium.
Cao teaches the conventional storage implementation: a “computer readable storage medium stores computer programs” executed by a processor. Cao [0149]. A POSITA would store Kwon’s ECG-training program on Cao’s medium because both run AI ECG-analysis methods, yielding the predictable result of executing Kwon’s known operations in stored software.
[Claim 12]
Kwon teaches,
A computing device for predicting a health state using an electrocardiogram,
comprising: (refer claim 1)
a processor including at least one core; (refer claim 1)
a memory including program codes executable in the processor; (refer claim 1)and
wherein the processor analyzes a missing value of the electrocardiogram data
or the cardiac arrest data to extract valid data from the electrocardiogram data, (refer claim 1) and
labels the extracted valid data based on whether the cardiac arrest occurs and the cardiac arrest occurrence time to generate training data. (refer claim 1)
Kwon teaches the ECG/cardiac-arrest data pipeline and the DLA logic, but does not expressly recite the claimed device package: processor, memory including executable program codes, and network unit configured to acquire ECG and cardiac-arrest data.
Cao supplies the device architecture: “memory and a processer,” stored programs, ECG input transmitted through “WIFI, Bluetooth, USB, 3G/4G/5G,” and “network and interface drivers.” Cao 0044], [0064], [0148]. A POSITA would implement Kwon’s software on Cao’s ECG-analysis apparatus to network-acquire ECG/EHR data and execute the same AI workflow.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JOSHUA DAMIAN RUIZ whose telephone number is (571)272-0409. The examiner can normally be reached 0800-1800.
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.
/JOSHUA DAMIAN RUIZ/Examiner, Art Unit 3684
/Shahid Merchant/Supervisory Patent Examiner, Art Unit 3684
Prosecution Insights