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 .
Claim Rejections - 35 USC § 101
Claims 1-11 are rejected under 35 U.S.C 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1:
Independent claims 1, 10, and 11 are directed to a method, a computer program product, and a device for predicting a health state. Thus, the claims are directed to statutory categories of invention. (Step 1: YES)
Step 2A, Prong 1
Independent claims 1, 10, and 11 recite the following limitations:
“predicting a health state using an electrocardiogram segment” (mental process)
“dividing electrocardiogram data to generate a plurality of segments;” (mathematical calculation and/or mental process)
“calculate possibilities of diseases corresponding to each of the plurality of segments” (mental process and/or mathematical calculation)
“removing outliers from among the possibilities of the diseases corresponding to each of the plurality of segments” (mental process and/or mathematical calculation
“combining the possibilities of the diseases with the outliers removed to” (mental process and/or mathematical calculation)
“generate a result value of the prediction of the health state” (mental process and/or mathematical calculation)
The dependent claims recite the following limitations:
“wherein the dividing of the electrocardiogram data to generate the plurality of segments includes dividing a signal in the electrocardiogram data to have a predetermined length without an overlapping area to generate the plurality of segments” (further limiting the abstract idea of “dividing of the electrocardiogram data”) (claim 2)
“wherein the dividing of the electrocardiogram data to generate the plurality of segments includes dividing the electrocardiogram data while moving a fixed-size window by a preset area in the electrocardiogram data to generate the plurality of segments” (further limiting the abstract idea of “dividing of the electrocardiogram data”)(claim 3)
“calculate uncertainty of output corresponding to each of the plurality of segments” (mental process and/or mathematical calculation) (claim 4)
“quantify the uncertainty of the output” (mental process and/or mathematical calculation) (claim 5)
“wherein the removing of the outliers from among the possibilities of the diseases corresponding to each of the plurality of segments includes considering and removing the possibility of the disease corresponding to the uncertainty of the output that is greater than or equal to a first preset reference value, when the uncertainty of the output is greater than or equal to the first preset reference value.” (further limiting the abstract idea of “removing of the outliers…”) (claim 6)
“wherein the combining of the possibilities of the diseases with the outliers removed to generate the result value of the prediction of the health state includes comparing an average value of the possibilities of the diseases with the outliers removed with a preset second reference value to generate the result value of the prediction of the health state.” (further limiting the abstract idea of “combining the possibilities…”) (claim 7)
“wherein the combining of the possibilities of the diseases with the outliers removed to generate the result value of the prediction of the health state includes comparing each of the possibilities of the diseases with the outliers removed with a preset third reference value to generate the result value of the prediction of the health state.” (further limiting the abstract idea of “combining the possibilities…”) (claim 8)
“wherein the comparing of each of the possibilities of the diseases with the outliers removed with the preset third reference value to generate the result value of the prediction of the health state includes: converting the possibilities of the diseases with the outliers removed into binary values based on whether the possibilities of the diseases with the outliers removed are greater than or equal to the preset third reference value; and” (further limiting the abstract idea of “comparing of each of the possibilities”) (claim 9)
“generating the result value of the prediction of the health state based on a value having a highest ratio among the converted binary values” (mental process and/or mathematical calculation) (claim 9)
Regarding the limitations directed to a mental process, the courts consider a mental process (thinking) that “can be performed in the human mind, or by a human using a pen and paper to be an abstract idea. Examples of mental processes include observation, judgement, evaluation and opinion. Examiner notes that the courts do not distinguish between claims that recite mental processes performed by humans and claims that recite mental processes performed on a computer. See MPEP 2106.04(a)(2)(III). The limitations above are nothing more than a heath care professional analyzing electrocardiogram data to determine a patient’s current health state.
Regarding the limitations directed to a mathematical calculation, under the broadest reasonable interpretation a mathematical calculation is a mathematical operation or an act of calculating using mathematical methods to determine a number of variable. See MPEP 2106.04(a)(2)(I). Examiner notes, as discussed in MPEP 2106.04(a)(2)(I), the claim does not need to recite the word “calculating” in order to be considered a mathematical calculation.
For the reasons above, Examiner asserts that the claims recite a judicial exception, specifically an abstract idea (Step 2A, Prong 1: Yes).
Step 2A, Prong 2
Claims 1, 10 and 11 recite the following additional elements:
“a computing device including at least one processor” (generic computer component)
“inputting the plurality of segments into a pre-trained machine learning model” (instruction to implement the abstract idea of “calculate possibility of disease”)
"at least one processor” (generic computer component)
“computer readable storage medium” (generic computer component)
“one or more processors” (generic computer component)
“ a processor including at least one core” (generic computer component)
“a memory including program codes executable in the processor” (generic computer component)
“a network unit” (generic computer component)
The above additional elements with the exception of “inputting the plurality of segments into a pre-trained machine learning model”, are examples of generic computer components. These additional elements fail to provide significantly more because they amount to merely applying the abstract idea using generic computer components.
The additional element, “inputting the plurality of segments into a pre-trained machine learning model” is an instruction to implement the abstract idea of “calculate possibilities of diseases”. This additional element fails to provide significantly more because it amounts to adding the word “apply it”. See MPEP 2106.05(f).
The dependent claims recite the following additional elements:
“inputting the plurality of segments into a pre-trained machine learning model” (instructions to implement the abstract idea of “calculate uncertainty of output…”) (claim 4)
“wherein the machine learning model includes a neural network having a probability distribution of neural network weights” (further limiting aspect of an additional element and/or generally linking to a specific technological environment)
The additional elements listed above fail to provide significantly more because they amount to either adding the word “apply it” (or an equivalent) with the judicial exception and generally linking the use of a judicial exception to a particular technological environment or field of use. See MPEP 2106.05(f) and 2106.05(h).
Step 2B
The claims do not include any additional elements that amount to significantly more than the judicial exception.
As discussed above, in Step 2A, Prong 2 the remaining additional elements amount to no more than generic computer components, applying the abstract idea using generic computer components, and generally linking the use of judicial exception to a particular technological environment or field of use.
Moreover, reconsidering the claim limitations individually and as an ordered combination, the claims fail to meet the requirements for eligibility under 35 U.S.C. 101. (Step 2B: NO).
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.
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claims 1, 3 and 8-11 are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Crespin et al. (US 2021/0386354).
In re claim 1, Crespin discloses a method (FIG. 2; [0089]) of predicting a health state [0120]: “normal heart rhythm” or “abnormal heart rhythm”) using an electrocardiogram segment (“Episode”), which is performed by a computing device including at least one processor [0125], the method comprising:
dividing (103) electrocardiogram data (“Segment(s), R waves”) to generate a plurality of segments (Subsegment 1….Subsegment N; [0106]);
inputting the plurality of segments to a pre-trained machine learning model (105) to calculate possibilities of diseases (106, [0112]: “a score vector for each of the sub-segment”, where diseases include [0110]: “asystole”, “Bradycardia”, “atrial fibrillation or atrial tachycardia”, “ventricular tachycardia”, “an artifact”, or “normal heart rhythm”) corresponding to each of the plurality of segments [0111-0112];
removing outliers from among the possibilities of the diseases corresponding to each of the plurality of segments ([0120]: “score vectors” being filled with “0” which “corresponds to the absence of one specific label”) and
combining ([0120]: “merged at episode level”) the possibilities of the diseases with the outliers removed to generate a result value ([0120]: “normal heart rhythm” or “abnormal heart rhythm”) of the prediction of the health state [0120].
In re claim 3, Crespin discloses, wherein the dividing of the electrocardiogram data to generate the plurality of segments includes dividing the electrocardiogram data while moving a fixed-size window by a preset area in the electrocardiogram data to generate the plurality of segments [0106].
In re claim 8, Crespin discloses, wherein the combining of the possibilities of the diseases with the outliers removed to generate the result value of the prediction of the health state includes comparing each of the possibilities of the diseases with the outliers removed with a preset third reference value ([0120]: “…set to “1” if all the sub-segments in the episode have been labelled as “normal heart rhythm”) to generate the result value of the prediction of the health state (FIG. 1: “Classification”; [0120].
In re claim 9, wherein the comparing of each of the possibilities of the diseases with the outliers removed with the preset third reference value to generate the result value of the prediction of the health state includes:
converting the possibilities of the diseases with the outliers removed into binary values based on whether the possibilities of the diseases with the outliers removed are greater than or equal to the preset third reference value (107, [0120]; and
generating the result value of the prediction of the health state (FIG.1 : “Classification”) based on a value having a highest ratio among the converted binary values ([0120]: “normal heart rhythm” if “all sub-segments in the episode are labeled as “normal heart rhythm””).
In re claim 10, see above (In re claim 1).
Crespin also discloses, a computer program [0130] that is stored in a computer-readable storage medium [0129] and performs operations for predicting a health state [0130].
In re claim 11, see above (In re claim 1).
Crespin also discloses the computing device comprising:
a processor [0123] including at least one core (apparent);
a memory [0129] including program codes executable in the processor [0129, 0131]; and
a network unit [0087] through which electrocardiogram data is acquired [0087].
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.
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.
Claim 2 is rejected under 35 U.S.C. 103 as being unpatentable over Crespin et al. (US 2021/0386354), in view of Anastasia et al. (US 2021/0315506).
In re claim 2, Crespin discloses, wherein the dividing of the electrocardiogram data to generate the plurality of segments includes dividing a signal in the electrocardiogram data to generate the plurality of segments (103, “Subsegment 1…N”)
Crespin does not disclose,
dividing a signal in the electrocardiogram data to have a predetermined length without an overlapping area to generate the plurality of segments.
Anastasia discloses a method of predicting a heath state using electrocardiogram data (abstract). The method, in part, involves dividing the electrocardiogram data into a plurality of segments prior to inputting the plurality of segments into a pretrained machine learning model for further evaluation (FIG. 3). As disclosed by Anastasia, the electrocardiogram data can be divided by 1) using a window based on ECG peak identifiers (e.g. an R peak of a QRS complex) [0105] or 2) using a window with a predetermined length ([0105]: “previously set”) that adequately captures a heartbeat without overlapping adjacent ones [0105].
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the dividing of the electrocardiogram data to generate a plurality of segments taught by Crespin to include dividing a signal in the electrocardiogram data to have a predetermined length without an overlapping area to generate the plurality of segments, as taught by Anastasia. One would have been motivated to make this modification because the division techniques disclosed by Crespin and Anastasia are functionally equivalent, that is, they both divide an electrocardiogram signal into a plurality of segments. Moreover, one ordinary skill in the art would have the ability to choose the division technique that would best meet their needs.
Claims 4 and 6 are rejected under 35 U.S.C. 103 as being unpatentable over Crespin et al. (US 2021/0386354), in view of Smith et al. (US 2022/0330838).
In re claim 4, Crespin discloses, further comprising inputting the plurality of segments into the pre-trained machine learning model (106; [0112, 0113]).
Crespin does not disclose inputting the plurality of segments into the pre-trained machine learning model to calculate uncertainty of output corresponding to each of the plurality of segments.
Smith discloses a method of predicting a heath state (abstract: “cardiac event”) that, like Crespin, involves inputting electrocardiogram data into a machine learning model. Smith further discloses the machine learning model having the ability to determine a confidence level, that is, an indication of certainty or uncertainty in the accuracy of the machine learning models classification/identification [0024]. In some instances, Smith discloses filtering out events with a low confidence level [0029].
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the method of Crespin to include inputting the plurality of segments into the pretrained machine learning model to calculate uncertainty of output, as taught by Smith. One would have been motivated to make this modification to ensure that all electrocardiogram data being analyzed is both clinically significant and have confidence levels that indicate that the machine learning models classification/identification is accurate (Smith, [0029]). Accordingly, such a modification would yield “inputting the plurality of segments into the pretrained machine learning model to calculate uncertainty of output corresponding to each of the plurality of segments.”
In re claim 6, Crespin does not disclose, wherein the removing of the outliers from among the possibilities of the diseases corresponding to each of the plurality of segments includes considering and removing the possibility of the disease corresponding to the uncertainty of the output that is greater than or equal to a first preset reference value, when the uncertainty of the output is greater than or equal to the first preset reference value.
As previously discussed, Smith discloses removing cardiac events with a low confidence level, i.e., high degree of uncertainty, from further analysis [0029].
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the processes of removing the outliers from among the possibilities of the diseases corresponding to each of the plurality of segments to include considering and removing the possibility of the disease corresponding to the uncertainty of the output that is greater than or equal to a first preset reference value, when the uncertainty of the output is greater than or equal to the first preset reference value, as taught by Smith. One would have been motivated to make this modification because confidence thresholding is a known strategy used in machine learning to ensure system reliability, as evidenced by Smith. Examiner notes that such a strategy is particularly useful in medical diagnostics as a wrong prediction could have negative and even life-threatening consequences.
Claim 5 is rejected under 35 U.S.C. 103 as being unpatentable over Crespin et al. (US 2021/0386354), in view of Smith et al. (US 2022/0330838), [in view of Patel et al. (US 2007/0239043)].
In re claim 5, the proposed combination yields (all mapping directed to Smith),
wherein the machine learning model includes a neural network [0023] having *a probability distribution of neural network weights to quantify the uncertainty of the output [0024].
*It is apparent that the neural network of the proposed combination would have “a probability distribution of neural network weights”, given that standard neural networks output a single predictive value. In other words, as known in the art, to get both an “uncertainty” and a predictive value a neural network with a probability distribution of weights (e.g., a Bayesian network) must be used. However, in so far as the proposed combination does not explicitly state whether or not the neural network has “a probability distribution of neural network weights” claim 5 is alternatively rejected under 35 U.S.C 103 in view of Crespin, in view of Smith, in view of Patel as follows:
Patel discloses an implantable cardiac device (abstract) that processes episode data using a Bayesian network to produce a list of potential diagnosis along with an indication (e.g., a probability) of each potential diagnosis being correct [0076].
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the neural network of the proposed combination to be a Bayesian neural network, as taught by Patel. One would have been motivated to make this modification given that using a Bayesian network is a known way of incorporating expert knowledge into decision making processes while allowing for uncertainty through the use of probabilities (Patel, [0038]). Accordingly, such a modification would yield the neural network “having a probability distribution of neural network weights”.
Claim 7 is rejected under 35 U.S.C. 103 as being unpatentable over Crespin et al. (US 2021/0386354), in view of Nishihara et al. (US 2025/0160723).
In re claim 7, Crespin discloses, wherein the combining of the possibilities of the diseases with the outliers removed to generate the result value of the prediction of the health state includes comparing the possibilities of the diseases with the outliers removed with a preset second reference value to generate the result value of the prediction of the health state ([0120]; where the “second reference value” is “all of the sub-segments in the episode have been labeled as “normal heart rhythm”)
Crespin does not disclose, comparing an average value of the possibilities of diseases with the outliers removed with a present second reference value.
Nishihara discloses an analogous method (abstract) of predicting a heath state ([0077]: “whether or not…electrocardiogram is a disease”) using an electrocardiogram segment [0077], that like Crespin, evaluates an electrocardiogram segment based on a number of positive example labels and a number of negative example labels output by a pretrained machine learning model [0075]. Nishihara also discloses alternatively evaluating the electrocardiogram segment using an average evaluation value instead of the number of positive/negative example labels [0075].
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the method of Crespin to include comparing an average value of the possibilities of disease removed, as taught by Nishiara. One would have been motivated to make this modification given that the types of model evaluation (i.e., positive/negative labels and average values) are known alternatives as evidenced by Nishiara (see Nishiara, [0075]).
Conclusion
The following prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
Ahmed et al. (US 2022/0039727) discloses a method (abstract) that automatically detects cardiovascular disease based on electrocardiogram signals using machine learning [0002].
De Chazal et al. (US 2020/0107775) discloses a method for detecting and diagnosing sleep apnea using electrocardiogram data (abstract). The method involves dividing the electrocardiogram data into a plurality of sections ([0084]: “corrected signals are divided into epochs”) and feeding the plurality of segments into a machine learning model [0087].
Contact
Any inquiry concerning this communication or earlier communications from the examiner should be directed to OLIVIA WALKER whose telephone number is (571)272-7052. The examiner can normally be reached M-F: 7-4pm CT.
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, David Hamaoui can be reached at (571)-270-5625. 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.
/OLIVIA WALKER/Examiner, Art Unit 3796
/DAVID HAMAOUI/SPE, Art Unit 3796
Prosecution Insights