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 .
Response to Amendment
Applicant's submission filed on 2025-07-29 has been entered. The status of claims is as follows:
Claims 1-8 and 10-14 remain pending in the application.
Claims 1 and 8 are amended.
Claim 9 is cancelled.
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 11/08/2025 has been entered.
Response to Arguments
Applicant's arguments filed in response to rejections under 35 USC 101 have been fully considered but they are not persuasive.
Regarding applicant’s reliance on the decision of the Appeals Review Panel in Ex parte Desjardins, No. 2024-000567 (P.T.A.B. Sept. 26, 2025), Examiner notes that, in Desjardins, unlike in the claims at issue here, the appellants specifically argued that the claimed invention “address[es] challenges in continual learning and model efficiency by reducing storage requirements and preserving task performance across sequential training”. Desjardins, op. at 7. That is, the appellant in Desjardins specifically alleged that the claimed subject matter improves machine learning itself. By contrast, Applicant in the instant case does not point to any specific claim language that characterizing an improvement, and does not point to any claim language that is analogous to the claims at issue in Desjardins.
Applicant argues on Remarks Pages 8-9 that “the present application solves the aforementioned systemic technical problems through the specialized integration of the sensor and the computing device, and the present application uses an optimized deep learning model to execute a variational autoencoder algorithm.” Applicant also states “the claimed invention integrates the judicial exception into a practical application by providing an improved signal detection method.”
Examiner respectfully disagrees, and points out that the claims do not recite any sort of “specialized integration” of the sensor and the computing device that would amount to significantly more than the abstract idea. The claim merely recites “collecting initial data” in Claim 1 and broadly recites “electrically connected to the sensor” in Claim 8. If there is a specialized unconventional interface, this is not recited in the claims. These limitations are followed by an abstract idea of reconstructing the original signal with an algorithm. The fact that the algorithm uses “a variational autoencoder algorithm using an optimized deep learning model”, broadly recited at a high level of generality, amounts to nothing more than nothing more than an instruction to apply the abstract idea using a generic computer. Furthermore, an “improved signal detection method” is not an improvement to technology, but rather an improvement to the abstract idea of detecting a signal. Examiner notes MPEP 2106.05(a) states: “It is important to note, the judicial exception alone cannot provide the improvement.” In this case, Applicant is providing an improved algorithm to detect signals, wherein performing the algorithm is an abstract idea.
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 1-7 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1:
Claims 1-7 are directed to a method. Therefore, each of the claims is directed to one of the four statutory categories of patent eligible subject matter.
Step 2A Prong 1:
Claim 1 recites:
“reconstructing the original signal by executing a variational autoencoder algorithm using an optimized deep learning model to generate a reconstructed signal, wherein the variational autoencoder algorithm comprises an encoder portion for performing data compression and a decoder portion for performing data decompression, wherein the encoder portion performs an operation on the input values of the original signal through computing conditions of a hidden layer to output two vectors with mean and standard deviations, generates a third vector with errors by a normal distribution, performs exponential processing on the standard deviations, and adds products to the means after the standard deviations are multiplied by the errors to become low-dimensional vectors in an intermediate layer, a general equation of the low-dimensional vector ci is represented as ci = exp(σi)*ei + mi, where σi is the standard deviation, ei is the error, and the mi is the mean, and when the low-dimensional vectors are obtained, computing conditions of a hidden layer in the decoder portion performs operations according to the low-dimensional vectors to perform the signal reconstruction to obtain output values as the reconstructed signal”; this limitation describes the details of a mathematical algorithm, and is thus mathematical calculations under 2106.04(a)(2)(I)(C)
“comparing the original signal with the reconstructed signal to determine whether there is an abnormality in the original signal”; comparing to make a determination is a mental process
Step 2A Prong 2:
This judicial exception is not integrated into a practical application because the additional elements are as follows:
“sensor… computing device”; this amounts to mere instructions to implement the abstract idea on a computer as per MPEP 2106.05(f)
“collecting initial data”; this amounts to insignificant extra solution activity, mere data gathering as per MPEP 2106.05(g)
“pre-processing the initial data by filtering a noise from the initial data to obtain an original signal”; this is merely another aspect of insignificant extra solution activity, mere data gathering as per MPEP 2106.05(g), as “filtering a noise” from initial data is similar to “Selecting a particular data source or type of data to be manipulated” such as “(iii) Selecting information, based on types of information”
Step 2B:
The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional elements are as follows:
“collecting initial data”; this amounts to insignificant extra solution activity, mere data gathering as per MPEP 2106.05(g); furthermore, this amounts to well-understood, routine, and conventional activity per MPEP 2106.05(d): “i. Receiving or transmitting data over a network”
“pre-processing the initial data by filtering a noise from the initial data to obtain an original signal”; this is merely another aspect of insignificant extra solution activity, mere data gathering as per MPEP 2106.05(g), as “filtering a noise” from initial data is similar to “Selecting a particular data source or type of data to be manipulated” such as “(iii) Selecting information, based on types of information”; furthermore, pre-processing signal data to remove noise is well-understood, routine, and conventional activity per MPEP 2106.05(d) as indicated by Berkheimer evidence supplied by the 2015 paper Jambukia et al. (“Classification of ECG signals using machine learning techniques: A survey”) which states on Page 715 Section II “Background Knowledge”: “Preprocessing, feature extraction, normalization, and classification are main sequential steps of ECG classification. Researchers have applied different preprocessing techniques for ECG classification. For noise removal, techniques such as low pass linear phase filter, linear phase high pass filter etc. are used.”
Dependent Claims
Claim 2 recites: “wherein the abnormality in the original signal is determined when a difference between the original signal and the reconstructed signal is greater than a preset threshold”; comparing a difference to a threshold is a mental process
Claim 3 recites: “wherein the original signal is a sound signal, an image signal, or an oscillatory wave signal”; this amounts to merely indicating a field of use or technological environment in which to apply a judicial exception under Steps 2A Prong 2 and 2B as per MPEP 2106.05(h)
Claim 4 recites: “wherein an establishment of the optimized deep learning model further comprises:”
“collecting a plurality of pieces of sample data”; this amounts to insignificant extra solution activity, mere data gathering under Steps 2A Prong 2 and 2B as per MPEP 2106.05(g)
“pre-processing the plurality of pieces of sample data to obtain a plurality of sample signals”; this amounts to insignificant extra solution activity, mere data gathering under Steps 2A Prong 2 and 2B as per MPEP 2106.05(g) (see Rejection to Claim 1 for details)
“training a deep learning model by the sample signals, to generate the optimized deep learning model”; this amounts to insignificant extra-solution activity, “(2) whether the limitation is significant (i.e. it imposes meaningful limits on the claim such that it is not nominally or tangentially related to the invention)” under Step 2A Prong 2 as per MPEP 2106.05(g), as Examiner notes that in order to use any machine learning model, it must be trained first, and therefore when “training” is broadly recited at a high level of generality, it does not impose a meaningful limit on the claim; furthermore, under Step 2B training a ML model amounts to well-understood, routine, and conventional activity per MPEP 2106.05(d) as indicated by Berkheimer evidence Yufeng (“The 7 Steps of Machine Learning”) which states: “Now we move onto what is often considered the bulk of machine learning — the training.”
Claim 5 recites:
“wherein the initial data and the plurality of pieces of sample data are collected”; this amounts to insignificant extra solution activity, mere data gathering under Steps 2A Prong 2 and 2B as per MPEP 2106.05(g)
“by a same test condition”; evaluating a condition is a mental process
Claim 6 recites: “wherein the step of training the deep learning model by the sample signals further comprises”
“performing feature extraction on each sample signal to obtain feature data”; this is merely another aspect of insignificant extra solution activity, mere data gathering as per MPEP 2106.05(g), as “filtering a noise” from initial data is similar to “Selecting a particular data source or type of data to be manipulated” such as “(iii) Selecting information, based on types of information”; furthermore, performing feature extraction on signal data is well-understood, routine, and conventional activity per MPEP 2106.05(d) as indicated by Berkheimer evidence supplied by the 2015 paper Jambukia et al. (“Classification of ECG signals using machine learning techniques: A survey”) which states on Page 715 Section II “Background Knowledge”: “Preprocessing, feature extraction, normalization, and classification are main sequential steps of ECG classification. Feature extraction techniques used by researchers are Discrete Wavelet Transform (DWT), Continuous Wavelet Transform (CWT), Discrete Cosine Transform (DCT), STransform (ST), Discrete Fourier transform (DFT), Principal Component Analysis (PCA), Daubechies wavelet (Db4), Pan-Tompkins algorithm, Independent Component Analysis (ICA) etc.”
“reconstructing a training signal according to the feature data”; Examiner notes that Claim 1 details the reconstructing as a mathematical algorithm, and as explained in Claim 1, this is a mathematical calculation
“calculating a difference between the training signal and each sample signal, to adjust a model parameter of the deep learning model according to the difference”; calculating a difference and adjusting a parameter based on the difference is a mathematical calculation
“obtaining an optimized model parameter when the difference between the training signal and each sample signal converges to a minimum value”; determining whether a difference converges can be performed in the human mind and is a mental process, and Examiner notes that the convergence criterion could also even be a mathematical calculation
“using the optimized model parameter in the deep learning model, to generate the optimized deep learning model”; determining to set parameters of a model can be performed in the human mind with the aid of pen and paper and is thus a mental process, this could also be viewed as constructing a mathematical calculation
Examiner further notes that the training process described in Claim 6 does not remedy the observation in Claim 4 that the generically-recited training is insignificant extra-solution activity, “(2) whether the limitation is significant (i.e. it imposes meaningful limits on the claim such that it is not nominally or tangentially related to the invention)” under Step 2A Prong 2 as per MPEP 2106.05(g), because these limitations of the steps of training in Claim 6 are still broadly recited, and describe the general training practice of adjusting parameters based on a loss. Furthermore, the method of training is still well-understood, routine, and conventional activity per MPEP 2106.05(d), because reconstruction error is a standard part of VAE training loss functions, as indicated by Berkheimer evidence Zimmerer et al. (“Context-encoding Variational Autoencoder for Unsupervised Anomaly Detection”) which states in Abstract, “However, state-of-the-art anomaly scores are still based on the reconstruction error”, and on Page 2: “A common choice for the reconstruction error Lrec(x, xˆ) is the mean-squared error (MSE)”
Claim 7 recites: “subtracting the reconstructed signal from the original signal to obtain an abnormal pattern”; subtracting is a mathematical calculation
Claims 8 and 10-14 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1:
Claims 8-14 are directed to an electronic device. Therefore, each of the claims is directed to one of the four statutory categories of patent eligible subject matter.
Step 2A Prong 1:
Claim 8 recites:
“reconstructs the original signal by executing a variational autoencoder algorithm using the optimized deep learning model to generate a reconstructed signal, wherein the variational autoencoder algorithm comprises an encoder portion for performing data compression and a decoder portion for performing data decompression, and compare the original signal with the reconstructed signal to determine whether there is an abnormality in the original signal; wherein the encoder portion performs an operation on the input values of the original signal through computing conditions of a hidden layer to output two vectors with mean and standard deviations, generates a third vector with errors by a normal distribution, performs exponential processing on the standard deviations, and adds products to the means after the standard deviations are multiplied by the errors to become low-dimensional vectors in an intermediate layer, and when the low-dimensional vectors are obtained, computing conditions of a hidden layer in the decoder portion performs operations according to the low-dimensional vectors to perform the signal reconstruction to obtain output values as the reconstructed signal”; this limitation describes the details of a mathematical algorithm, and is thus mathematical calculations under 2106.04(a)(2)(I)(C)
“and adds products to the means after the standard deviations are multiplied by the errors to become low-dimensional vectors in an intermediate layer, a general equation of the low-dimensional vector c is represented as ci=exp(ai)*ei+mi, where az is the standard deviation, e is the error, and the m, is the mean,”; This limitation is directed to the abstract idea of a mental process (concepts performed in the human mind, including observation and evaluation or using pen and paper [see MPEP 2106.04(a)(2) III. C.]).
Step 2A Prong 2:
This judicial exception is not integrated into a practical application because the additional elements are as follows:
“An electronic device, comprising … a computing device … wherein the computing device configures to”; this amounts to mere instructions to implement the abstract idea on a computer as per MPEP 2106.05(f)
“a sensor; and a computing device, electrically connected to the sensor”; this amounts to merely indicating a field of use or technological environment in which to apply a judicial exception as per MPEP 2106.05(h)
“collect a plurality of pieces of sample data and initial data”; this amounts to insignificant extra solution activity, mere data gathering as per MPEP 2106.05(g)
“pre-process the plurality of pieces of sample data by filtering a noise from the plurality of pieces of sample data to obtain a plurality of sample signals; pre-process the initial data to obtain an original signal”; this is merely another aspect of insignificant extra solution activity, mere data gathering as per MPEP 2106.05(g), as “filtering a noise” from initial data is similar to “Selecting a particular data source or type of data to be manipulated” such as “(iii) Selecting information, based on types of information”
“trains a deep learning model by the sample signals to generate an optimized deep learning model … when the optimized deep learning model is established”; “ this amounts to insignificant extra-solution activity, “(2) whether the limitation is significant (i.e. it imposes meaningful limits on the claim such that it is not nominally or tangentially related to the invention)” under Step 2A Prong 2 as per MPEP 2106.05(g), as Examiner notes that in order to use any machine learning model, it must be trained first, and therefore when “training” is broadly recited at a high level of generality, it does not impose a meaningful limit on the claim
“first inputted into the optimized deep learning model … by the optimized deep learning model”; performing an abstract idea by a broadly recited machine learning model at a high level of generality amounts to nothing more than an instruction to apply the abstract idea using a generic computer as per MPEP 2106.05(f)
Step 2B:
The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional elements are as follows:
“An electronic device, comprising … a computing device … wherein the computing device configures to”; this amounts to mere instructions to implement the abstract idea on a computer as per MPEP 2106.05(f)
“a sensor; and a computing device, electrically connected to the sensor”; this amounts to merely indicating a field of use or technological environment in which to apply a judicial exception as per MPEP 2106.05(h)
“collect a plurality of pieces of sample data and initial data”; this amounts to insignificant extra solution activity, mere data gathering as per MPEP 2106.05(g); furthermore, this amounts to well-understood, routine, and conventional activity per MPEP 2106.05(d): “i. Receiving or transmitting data over a network”
“pre-process the plurality of pieces of sample data by filtering a noise from the plurality of pieces of sample data to obtain a plurality of sample signals; pre-process the initial data to obtain an original signal”; this is merely another aspect of insignificant extra solution activity, mere data gathering as per MPEP 2106.05(g), as “filtering a noise” from initial data is similar to “Selecting a particular data source or type of data to be manipulated” such as “(iii) Selecting information, based on types of information”; furthermore, pre-processing signal data to remove noise is well-understood, routine, and conventional activity per MPEP 2106.05(d) as indicated by Berkheimer evidence supplied by the 2015 paper Jambukia et al. (“Classification of ECG signals using machine learning techniques: A survey”) which states on Page 715 Section II “Background Knowledge”: “Preprocessing, feature extraction, normalization, and classification are main sequential steps of ECG classification. Researchers have applied different preprocessing techniques for ECG classification. For noise removal, techniques such as low pass linear phase filter, linear phase high pass filter etc. are used.”
“trains a deep learning model by the sample signals to generate an optimized deep learning model … when the optimized deep learning model is established”; this amounts to insignificant extra-solution activity, “(2) whether the limitation is significant (i.e. it imposes meaningful limits on the claim such that it is not nominally or tangentially related to the invention)” under Step 2A Prong 2 as per MPEP 2106.05(g), as Examiner notes that in order to use any machine learning model, it must be trained first, and therefore when “training” is broadly recited at a high level of generality, it does not impose a meaningful limit on the claim; furthermore, under Step 2B training a ML model amounts to well-understood, routine, and conventional activity per MPEP 2106.05(d) as indicated by Berkheimer evidence Yufeng (“The 7 Steps of Machine Learning”) which states: “Now we move onto what is often considered the bulk of machine learning — the training.”
“first inputted into the optimized deep learning model … by the optimized deep learning model”; performing an abstract idea by a broadly recited machine learning model at a high level of generality amounts to nothing more than an instruction to apply the abstract idea using a generic computer as per MPEP 2106.05(f)
Dependent Claims
Claim 10 recites: “wherein the original signal is a sound signal, an image signal, or an oscillatory wave signal”; this amounts to merely indicating a field of use or technological environment in which to apply a judicial exception under Steps 2A Prong 2 and 2B as per MPEP 2106.05(h)
Claim 11 recites: “wherein the computing device further configures to”
“perform feature extraction on each sample signal to obtain feature data”; this is merely another aspect of insignificant extra solution activity, mere data gathering as per MPEP 2106.05(g), as “filtering a noise” from initial data is similar to “Selecting a particular data source or type of data to be manipulated” such as “(iii) Selecting information, based on types of information”; furthermore, performing feature extraction on signal data is well-understood, routine, and conventional activity per MPEP 2106.05(d) as indicated by Berkheimer evidence supplied by the 2015 paper Jambukia et al. (“Classification of ECG signals using machine learning techniques: A survey”) which states on Page 715 Section II “Background Knowledge”: “Preprocessing, feature extraction, normalization, and classification are main sequential steps of ECG classification. Feature extraction techniques used by researchers are Discrete Wavelet Transform (DWT), Continuous Wavelet Transform (CWT), Discrete Cosine Transform (DCT), STransform (ST), Discrete Fourier transform (DFT), Principal Component Analysis (PCA), Daubechies wavelet (Db4), Pan-Tompkins algorithm, Independent Component Analysis (ICA) etc.”
“reconstruct a training signal according to the feature data”; Examiner notes that Claim 8 details the reconstructing as a mathematical algorithm, and as explained in Claim 8, this is a mathematical calculation
“calculate a difference between the training signal and each sample signal to adjust a model parameter of the deep learning model according to the difference”; calculating a difference and adjusting a parameter based on the difference is a mathematical calculation
“obtain an optimized model parameter when the difference between the training signal and each sample signal converges to a minimum value”; determining whether a difference converges can be performed in the human mind and is a mental process, and Examiner notes that the convergence criterion could also even be a mathematical calculation
“generate the optimized deep learning model based on the optimized model parameter in the deep learning model”; determining to set parameters of a model can be performed in the human mind with the aid of pen and paper and is thus a mental process, this could also be viewed as constructing a mathematical calculation
Examiner further notes that the training process described in Claim 11 does not remedy the observation in Claim 8 that the generically-recited training is insignificant extra-solution activity, “(2) whether the limitation is significant (i.e. it imposes meaningful limits on the claim such that it is not nominally or tangentially related to the invention)” under Step 2A Prong 2 as per MPEP 2106.05(g), because these limitations of the steps of training in Claim 11 are still broadly recited, and describe the general training practice of adjusting parameters based on a loss. Furthermore, the method of training is still well-understood, routine, and conventional activity per MPEP 2106.05(d), because reconstruction error is a standard part of VAE training loss functions, as indicated by Berkheimer evidence Zimmerer et al. (“Context-encoding Variational Autoencoder for Unsupervised Anomaly Detection”) which states in Abstract, “However, state-of-the-art anomaly scores are still based on the reconstruction error”, and on Page 2: “A common choice for the reconstruction error Lrec(x, xˆ) is the mean-squared error (MSE)”
Claim 12 recites:
“wherein the initial data and the plurality of pieces of sample data are collected”; this amounts to insignificant extra solution activity, mere data gathering under Steps 2A Prong 2 and 2B as per MPEP 2106.05(g)
“by a same test condition”; evaluating a condition is a mental process
Claim 13 recites: “wherein the computing device further subtracts the reconstructed signal from the original signal to obtain an abnormal pattern”; subtracting is a mathematical calculation
Claim 14 recites:
“wherein the computing device further marks the abnormal pattern”; marking an abnormal pattern can be performed by a human in the mind and with pen and paper, and is therefore a mental process
“provides the marked abnormal pattern to a supervised learning model as input data”; this amounts to insignificant extra solution activity, mere data gathering and outputting under Step 2A Prong 2 as per MPEP 2106.05(g); and furthermore gathering labeled data for training is well-understood, routine, and conventional activity under Step 2B as per Berkheimer evidence Yufeng (“The 7 Steps of Machine Learning”) which states: “Once we have our equipment and booze, it’s time for our first real step of machine learning: gathering data. This step is very important because the quality and quantity of data that you gather will directly determine how good your predictive model can be. In this case, the data we collect will be the color and the alcohol content of each drink … A few hours of measurements later, we have gathered our training data. Now it’s time for the next step of machine learning: Data preparation, where we load our data into a suitable place and prepare it for use in our machine learning training.”
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to VAN C MANG whose telephone number is (571)270-7598. The examiner can normally be reached Mon - Fri 8:00-5:00pm.
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 Yi can be reached at 5712707519. 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.
/VAN C MANG/Primary Examiner, Art Unit 2126