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 .
Priority
The present application is a continuation of an International Application No. PCT/IN2023/051047, with a filing date of Nov. 10, 2023, the entire disclosure of which is incorporated herein by reference for all purposes. The present application claims the benefit of Indian priority application No. 202241065024, with a filing date of Nov. 14, 2022, the entire disclosure of which is incorporated herein by reference for all purposes.
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 04/04/2025, and 07/10/2025 were filed along with and after the mailing date of the Non-Provisional Patent Application on 04/04/2025. The submission is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
DETAILED ACTION
This Office Action is in response to a Non-Provisional Patent Application received on 04/04/2025. In the application, claims 1-16 have been cancelled. Claims 17-25 were added under preliminary amendments.
For this Office Action, claims 17-25 have been received for consideration and have been examined.
Specification
Applicant’s submitted specification has been reviewed and found to be in compliance.
Drawings
Applicant’s submitted drawings have been reviewed and found to be in compliance.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 18 and 21 recites terms which have insufficient antecedent basis for in the claims.
Claim 18 recite “an abscissa corresponding to the spectral maxima (β), a width of the spectrum (ω), and a bias or asymmetry parameter of the spectrum (ϵ)”.
The underlined terms in claim 18 have insufficient antecedent basis for claim 18 which renders the claim indefinite.
Claim 21 recite “… or the like”. The recitation of “or the like” expression have insufficient antecedent basis for claim 21 which renders the claim indefinite. The expression indicate that the items already mentioned are representative examples and that other similar items may also be included but not explicitly listed.
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 17–25 are rejected under 35 U.S.C. §101 because the claimed invention is directed to a judicial exception—an abstract idea—without reciting additional elements sufficient to integrate the exception into a practical application or amount to significantly more than the judicial exception.
The analysis follows Alice Corp. v. CLS Bank International, 573 U.S. 208, 217–18 (2014); Mayo Collaborative Services v. Prometheus Laboratories, Inc., 566 U.S. 66, 72–73 (2012); and MPEP §§2106 and 2106.04.
Claim 17
Claim 17 recites a method for user authentication based on exhaled breath. The claim includes:
Collecting exhaled breath and acquiring corresponding velocity time-series signals;
Segmenting, filtering, and normalizing the signals;
Extracting features from the normalized signals;
Building a model library and generating training data;
Comparing newly obtained breath data with data in the model library; and
Confirming or identifying a user based on the comparison.
Step 2A, Prong One: Judicial exception
Claim 17 recites the abstract idea of:
Collecting physiological data, mathematically processing and organizing the data, comparing the processed data with previously stored data, and determining a person’s identity based on the comparison.
The “segmenting, filtering, and normalizing” limitations encompass mathematical manipulation of numerical time-series values. The specification explains that normalization may be performed by subtracting a mean from each value and dividing by the standard deviation. It further describes detrended fluctuation functions, Hurst exponents, mass exponents, singularity exponents, polynomial fitting, and multifractal spectra. Specification ¶¶ [0064]–[0067].
Similarly, “extracting a plurality of features,” “building a model library,” “generating training data,” and comparing user data with model-library data encompass mathematical calculations and statistical classification. Specification ¶¶ [0069], [0072]–[0074], [0081]–[0083], and [0094].
Accordingly, these limitations fall within the “mathematical concepts” grouping of abstract ideas because they encompass:
Mathematical calculations;
Mathematical relationships;
Statistical analysis;
Mathematical modeling; and
Comparison of numerical values to make a classification.
See MPEP §2106.04(a)(2); SAP America, Inc. v. InvestPic, LLC, 898 F.3d 1161, 1167–70 (Fed. Cir. 2018) (mathematical and statistical analysis of data held abstract); In re Board of Trustees of Leland Stanford Junior University, 991 F.3d 1245, 1250–53 (Fed. Cir. 2021) (mathematical processing and comparison of data to improve classification held abstract).
The limitations of “confirming” and “identifying” the user also recite the evaluation of information and formation of a judgment based on a comparison. Although the volume of data may make manual performance impractical, merely automating data evaluation and classification does not remove the claimed concept from the abstract-idea category. See Electric Power Group, LLC v. Alstom S.A., 830 F.3d 1350, 1353–54 (Fed. Cir. 2016).
The claim therefore recites a judicial exception.
Step 2A, Prong Two: Practical application
The additional elements include:
A mouthpiece;
A hot-wire anemometer (“HWA”);
A data-acquisition system;
Collecting breath through the mouthpiece;
Measuring breath-velocity signals using the HWA; and
Acquiring the measured signals using the data-acquisition system.
These elements do not integrate the abstract idea into a practical application.
First, the mouthpiece, HWA, and data-acquisition system are used principally to obtain the data upon which the mathematical analysis is performed. Collecting data for subsequent analysis constitutes insignificant pre-solution activity when the focus of the claim remains the analysis and classification of the collected information. See Electric Power Group, 830 F.3d at 1354–55; MPEP §2106.05(g).
The specification acknowledges that:
Respiratory-flow measurements were conventionally performed using spirometers and pneumotachographs;
Hot-wire anemometers previously had been used for respiratory-flow measurements;
Constant-temperature HWAs previously had been used to measure respiratory gas-flow rates; and
HWAs were known as robust tools for obtaining time-resolved turbulence measurements.
Specification ¶¶ [0008]–[0013].
Second, claim 17 does not recite a technological improvement to the mouthpiece, HWA, or data-acquisition system. It does not specify a new sensor architecture, calibration procedure, probe configuration, signal-acquisition circuit, or physical modification that improves how the HWA measures velocity. The claimed hardware performs its ordinary data-acquisition function and supplies information to the mathematical classification process.
Merely limiting an abstract analytical process to data collected by a particular sensor does not necessarily constitute integration into a practical application. See iLife Technologies, Inc. v. Nintendo of America, Inc., 839 F. App’x 534, 536–38 (Fed. Cir. 2021) (receiving sensor data and evaluating the data to determine body activity held abstract); Electric Power Group, 830 F.3d at 1355.
Third, “authenticating a user” states the intended field of use or result of the mathematical comparison. Claim 17 does not require controlling access to a secured device, locking or unlocking equipment, initiating a transaction, or otherwise changing the operation of another physical system based on the authentication result. Limiting an abstract idea to the field of biometric authentication does not itself integrate the exception into a practical application. See Parker v. Flook, 437 U.S. 584, 590 (1978); MPEP §2106.05(h).
Thus, claim 17 is directed to the abstract idea of mathematical processing and comparing physiological data to classify or identify a person.
Step 2B: Inventive concept
The additional elements, considered individually and as an ordered combination, do not amount to significantly more than the abstract idea.
Individually:
The mouthpiece merely receives exhaled breath;
The HWA performs its known function of measuring airflow velocity;
The data-acquisition system performs its generic function of receiving sensor data; and
The collection and acquisition steps merely supply input data for the abstract analysis.
The specification’s acknowledgment that HWAs previously were used for respiratory-flow measurements supports a finding that using an HWA to acquire breath-flow data was well-understood, routine, and conventional. Specification ¶¶ [0009]–[0013]. This finding is based on the applicant’s description of the prior art and is therefore consistent with Berkheimer v. HP Inc., 881 F.3d 1360, 1368–69 (Fed. Cir. 2018), and MPEP §2106.05(d).
As an ordered combination, the claim merely applies conventional breath-data collection before performing mathematical feature extraction, model construction, comparison, and classification. Arranging conventional data collection followed by abstract data analysis does not provide an inventive concept. See Electric Power Group, 830 F.3d at 1355; TLI Communications LLC v. AV Automotive, L.L.C., 823 F.3d 607, 613–15 (Fed. Cir. 2016).
Furthermore, the alleged novelty of using exhaled-breath turbulence for authentication does not establish eligibility. A novel or nonobvious abstract idea does not become patent eligible solely because its particular informational subject matter was previously unknown. See Diamond v. Diehr, 450 U.S. 175, 188–89 (1981); SAP America, 898 F.3d at 1163.
Claim 17 therefore does not recite significantly more than the judicial exception.
Claim 18
Claim 18 further recites that the extracted features include:
An abscissa corresponding to the spectral maximum;
A width of the spectrum; and
A bias or asymmetry parameter.
These limitations further specify mathematical properties extracted from the time-series data. The specification describes these parameters as dimensionless characteristics of a multifractal spectrum. Specification ¶¶ [0064]–[0069].
Calculating or extracting spectral maxima, spectrum widths, and asymmetry parameters constitutes mathematical analysis rather than an additional practical application. See SAP America, 898 F.3d at 1167–70. Claim 18 therefore does not overcome the rejection of claim 17.
Claim 19
Claim 19 recites using binary random-forest classifiers to select features.
A random-forest classifier is a mathematical model used to evaluate data, determine feature importance, and classify information. The specification explains that the classifier estimates the potential loss in performance if a feature is removed and selects frequently occurring features. Specification ¶ [0074].
The claim does not specify whether a new random-forest architecture or an improvement in the operation of a computer. It merely uses a known mathematical model to perform the abstract feature-selection process. Generic use of machine learning to analyze data does not, without more, provide an inventive concept. See Stanford, 991 F.3d at 1251–53; SAP America, 898 F.3d at 1167–70.
Claim 19 therefore does not overcome the rejection.
Claim 20
Claim 20 recites using data derived from velocity time-series signals, alone or with other signals, to train random-forest models.
Training a mathematical model using collected data is part of the abstract data-analysis process. The limitation identifies the information supplied to the model but does not improve the operation of the sensor, data-acquisition system, or computer.
The inclusion of signals “unrelated to breath related measurements” further demonstrates that the claimed model-training operation is not limited to a particular physical measurement technique. The limitation instead broadly covers supplying different categories of information to a mathematical model.
Claim 20 therefore does not integrate the exception into a practical application or provide significantly more.
Claim 21
Claim 21 recites obtaining breath-measurement data from:
A hot-wire anemometer;
Laser Doppler velocimetry;
Particle-tracking velocimetry; or
Particle-imaging velocimetry.
These devices and techniques merely identify alternative sources from which the data to be analyzed may be collected. The claim does not improve any of the recited measurement technologies or require an unconventional configuration of those technologies.
Selecting a particular data source or measurement environment does not make an otherwise abstract data-analysis process patent eligible. See Electric Power Group, 830 F.3d at 1355; TLI Communications, 823 F.3d at 613.
Claim 21 therefore does not overcome the rejection.
Claim 22
Claim 22 recites using random-forest models configured to capture complex decision boundaries between classes.
A decision boundary is a mathematical division of a feature space used to classify data. The specification explains that the feature space is divided into decision regions and that the random-forest model classifies samples according to those regions. Specification ¶ [0092].
Thus, the added limitation further refines the mathematical classification technique itself. It does not recite an improvement to a computer or other technology outside the abstract idea. Improving the accuracy of a mathematical model, without improving the underlying computer technology, remains within the abstract idea. See SAP America, 898 F.3d at 1167–70; Stanford, 991 F.3d at 1251–53.
Claim 22 therefore does not overcome the rejection.
Claim 23
Claim 23 recites a multi-model identification approach using a user-confirmation block comprising either:
A hypothesis-test-based model; or
A machine-learning-based model.
Both alternatives are mathematical or statistical techniques for evaluating data and making a classification. The specification identifies Hotelling’s hypothesis testing and random-forest classifiers as alternative analytical approaches. Specification ¶¶ [0083], [0094], and [0099].
Combining multiple abstract mathematical models and weighting their results remains mathematical analysis. See Stanford, 991 F.3d at 1251–53. The claim does not specify a technological improvement in how a computer executes either model.
Claim 23 therefore does not overcome the rejection.
Claim 24
Claim 24 recites applying the method individually or in combination with other biometric systems or time-series inputs, including:
Heart rate;
Fingerprint;
Gait;
Face;
Iris or retina;
Speech or voice;
Body temperature;
Breathing rate; and
Brain signals.
The limitation merely identifies additional types of information that may be collected and analyzed. It does not require a particular technological integration among the listed sensors, identify how the signals are combined, or cause a concrete physical action based on the authentication result.
Limiting an abstract data-analysis process to particular categories of biometric information constitutes a field-of-use limitation or data-source restriction. See Flook, 437 U.S. at 590; Electric Power Group, 830 F.3d at 1355.
Claim 24 therefore does not overcome the rejection.
Claim 25
Claim 25 recites classifying users in a manner that “supports diagnosis for personalized medication and treatment.”
The claim does not require:
Diagnosing a particular disease;
Identifying a particular medical condition;
Selecting a particular medication;
Determining or administering a particular dosage; or
Performing any treatment step.
Instead, the claim merely classifies users and states that the resulting classification “supports” diagnosis and personalized treatment. The limitation therefore represents a field of use or intended use of the classification result rather than an affirmative treatment application.
The specification similarly characterizes the clustering procedure as a database-based tool that may “potentially” serve as a diagnostic tool when baseline health data is available. Specification ¶ [0127].
Merely analyzing physiological information and providing information that may be used in diagnosis or treatment does not integrate the abstract analysis into a practical medical application. See Stanford, 991 F.3d at 1251–53; MPEP §§2106.05(e) and 2106.05(h).
Claim 25 therefore does not overcome the rejection.
Conclusion
Claims 17–25 are directed to collecting physiological data, mathematically processing and organizing that data, comparing it with stored data, and classifying or identifying a person. The mouthpiece, HWA, data-acquisition system, and alternative sensors merely supply data to the analytical process and perform their ordinary data-gathering functions.
The claims do not require improvement to measurement devices, to computer operation, or a concrete physical action resulting from authentication or diagnosis. Therefore, the additional elements, individually and in combination, neither integrate the abstract idea into practical application nor amount to significantly more than the judicial exception.
Accordingly, claims 17–25 are rejected under 35 U.S.C. §101.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claim(s) 17 and 19–25 are rejected under 35 U.S.C. 103 as being unpatentable over Sezan et al., (US20160106935A1) in view of Wolf et al., (US5518002A) and further in view of Khanzada., (US20220037022A1).
Regarding claim 17, Sezan discloses:
A method for authenticating a user based on exhaled breath:
receiving exhaled breath through an apparatus configured to receive breath samples, such as an inhaler or an over-the-mouth or over-the-nose mask (¶¶[0040], [0070]);
using breath-responsive sensors to identify persons, recognize a particular user, or confirm the identity of a person administering a drug (¶¶[0040], [0047]);
receiving sensor data and preprocessing the sensor outputs, including analog-to-digital conversion and filtering (¶¶[0069], [0088]);
extracting feature vectors from test, target, and control breath data (¶¶[0072], [0083], [0088]);
training a pattern classifier using positive breath samples belonging to a particular person and negative breath samples obtained from control subjects or other sources (¶¶[0070]–[0072], [0083]–[0084], [0089]); and
authenticating the person by comparing a test feature vector with a target feature vector according to a difference, distance, similarity, or classification criterion (¶¶[0072], [0084], [0089]).
Sezan et al. therefore teach collecting exhaled-breath information from users, extracting features, generating training data, training a classifier, and confirming or identifying a user by comparing newly obtained breath data with enrolled data.
Sezan et al. also teach both verification and identification. Specifically, Sezan et al. disclose that the electronic nose may “identify persons,” “recognize a particular user,” or determine that sensor data “identifies the person.” See ¶¶[0040], [0047], [0072]. These disclosures teach or suggest identifying a user from stored breath information without requiring the user first to declare an identity.
Sezan et al. do not expressly teach measuring exhaled-breath velocity using a hot-wire anemometer.
Wolf et al., however, teach:
an airflow chamber having a mouthpiece;
a hot wire acting as an anemometer positioned in the airflow chamber;
sensing exhaled airflow passing through the mouthpiece;
signal-generating circuitry connected to the hot wire; and
an electronic package that computes and records the flow rate and volume of each exhalation.
See Wolf et al., claims 1, 7, and 15.
Wolf et al. further disclose that the exhaled breath cools the heated wire and produces a waveform proportional to airflow. The waveform is supplied to an analog-to-digital converter and microprocessor for sampling and processing. See Wolf et al., col. 6, approximately ll. 10–65.
Wolf et al. also teach sampling the flow sensor, storing the sampled values in a buffer, smoothing the samples using a digital filter, and calculating respiratory-flow parameters from the sampled waveform. See Wolf et al., col. 9, approximately ll. 25–46.
The combination of Sezan and Wolf fails to disclose:
a respiratory-signal processing pipeline that includes preprocessing, filtering, feature extraction, model training, and model inference.
Khanzada further teaches a respiratory-signal processing pipeline that includes preprocessing, filtering, feature extraction, model training, and model inference. See Khanzada ¶¶[0016], [0042]–[0044]. Khanzada teaches dividing respiratory time-series data into chunks or segments, extracting multiple features from the segments, partitioning the data into training, validation, and test sets, and training multiple machine-learning models. See ¶¶[0031], [0042]–[0044], [0055]–[0057], [0065]–[0066]. Khanzada also teaches using normalization in the machine-learning processing architecture. See ¶¶[0048]–[0049].
It would have been obvious to one of ordinary skill in the art before the effective filing date to employ the mouthpiece, hot-wire airflow sensor, and digital signal-acquisition arrangement of Wolf et al. as an additional or alternative breath-sensing channel in the authentication system of Sezan et al.
Sezan et al. seek sensor-derived characteristics that reliably distinguish one person's exhaled breath from another person's breath. Wolf et al. teach a compact and established breath sensor that produces a quantitative, time-dependent waveform representative of a person's exhaled airflow. Employing Wolf et al.'s hot-wire sensor would predictably provide additional breath-related features for the classifier of Sezan et al.
It further would have been obvious to process the sampled breath waveform using Khanzada's segmentation, preprocessing, normalization, feature extraction, training-set generation, and multiple-model techniques. A skilled artisan would have been motivated to do so to place breath signals on a comparable scale, increase the number of training samples, reduce irrelevant sensor variation, and improve classification accuracy.
Regarding claim 19, the combination of Sezan, Wolf and Khanzada teaches:
random forests as suitable machine-learning classifiers. Khanzada explains that a random forest:
constructs a collection of decision trees;
selects variables for splitting groups of data;
repeatedly selects the best split;
processes a large number of input variables; and
identifies the most significant variables.
See Khanzada ¶¶[0060], [0063].
Khanzada further teaches generating classifiers using selected feature subsets and evaluating the classifiers to determine whether the selected features provide an acceptable classification result. See ¶¶[0065]–[0066].
Sezan et al. teach positive training samples belonging to a particular user and negative training samples obtained from other persons or subjects. See Sezan et al. ¶¶[0070]–[0072], [0083]–[0084]. It therefore would have been obvious to configure Khanzada's random forest as a binary classifier distinguishing a target user from a non-target user.
Using the random forest's variable-importance results to select the most discriminating breath features would have been an ordinary application of Khanzada's express teaching that random forests identify significant variables from large numbers of input variables.
Regarding claim 20, the combination of Sezan, Wolf and Khanzada teaches:
training random-forest models using information derived from the velocity time series, either alone or in combination with breath-related or unrelated signals.
Wolf et al. teach obtaining sampled exhaled-flow information from a hot-wire anemometer. Khanzada teaches machine-learning models, including random forests, using respiratory signal features and combining information from multiple signal modalities. Khanzada identifies the following suitable inputs:
breathing and cough signals;
temporal information;
facial and fingertip video;
pulse-oximetry information;
inertial or gait information;
heart-rate information; and
temperature information.
See Khanzada ¶¶[0016]–[0019], [0031], [0033], [0060], [0071]–[0072].
It would have been obvious to train the random-forest models using the HWA-derived velocity information alone or in combination with Khanzada's additional respiratory or physiological modalities. Khanzada teaches that using multiple modalities can enhance accuracy and provide additional information to an ensemble classifier. See ¶¶[0017], [0031], [0033].
Regarding claim 21, the combination of Sezan, Wolf and Khanzada teaches:
The breath signal be selected from a signal obtained using a hot-wire anemometer, LDV, PTV, PIV, or a similar measurement technique.
Wolf et al. expressly teach obtaining an exhaled-breath airflow signal using a hot wire acting as an anemometer. See Wolf et al., claims 1, 7, and 15. Wolf et al. therefore directly teach the claimed HWA alternative.
Regarding claim 22, the combination of Sezan, Wolf and Khanzada teaches:
random-forest models configured to capture complex decision boundaries.
Khanzada teaches that a random forest constructs multiple decision trees, selects the best split from candidate variables, and recursively partitions groups of data until terminal nodes are reached. See Khanzada ¶[0063].
The resulting collection of recursive decision-tree partitions defines nonparametric class regions and is capable of representing complex decision boundaries between classes. Khanzada also teaches that decision trees identify relationships between multiple variables and that random forests are suitable for classification involving large numbers of input variables. See ¶¶[0062]–[0063].
It would have been obvious to use such random-forest decision boundaries to distinguish the breath-feature vectors of different users because Sezan et al. already teach classifying and comparing user and control feature vectors, while Khanzada teaches random forests as suitable classifiers for large, multidimensional feature sets.
Regarding claim 23, the combination of Sezan, Wolf and Khanzada teaches:
implementing identification using a multi-model approach having a user-confirmation block comprising a hypothesis-test-based model or a machine-learning-based model.
Sezan et al. teach a user-confirmation classifier that determines whether test breath information verifies the identity of a person. See Sezan et al. ¶¶[0069]–[0072], [0083]–[0089].
Khanzada teaches a multi-model ensemble comprising separately trained upstream submodels whose outputs are combined to produce an ultimate classification. Khanzada expressly identifies decision trees, random forests, neural networks, and other machine-learning classifiers as suitable models. See Khanzada ¶¶[0031], [0033], [0060]–[0063].
It would have been obvious to use Khanzada's ensemble arrangement for the user-confirmation classifiers of Sezan et al. to aggregate decisions derived from multiple breath features or sensing modalities, thereby improving classification robustness and accuracy.
Regarding claim 24, the combination of Sezan, Wolf and Khanzada teaches:
authenticating a user using the exhaled-breath biometric alone or in combination with another biometric or physiological time-dependent signal.
Sezan et al. teach combining breathprint authentication with fingerprint authentication. See Sezan et al. ¶¶[0043], [0074].
Khanzada teaches combining respiratory information with:
facial imagery;
fingertip imagery;
gait or inertial signals;
heart-rate signals;
temperature signals;
pulse-oximetry signals; and
other biometric information.
See Khanzada ¶¶[0017], [0019], [0071]–[0072].
It would have been obvious to combine breath authentication with one or more of these known biometric modalities to provide redundant identity evidence, reduce false acceptance and rejection rates, and allow authentication when one sensing modality provides insufficient signal quality.
Regarding claim 25, the combination of Sezan, Wolf and Khanzada teaches:
classifying users using the identification method and using the classification to support diagnosis and personalized medication or treatment.
Sezan et al. teach analyzing breath information both to identify a person and to determine a biological or respiratory condition. Sezan et al. further teach selecting or adjusting a recommended drug dose based on the detected condition, previous administrations, and the user's response to treatment. See Sezan et al. ¶¶[0043], [0055]–[0056], [0079], [0094], [0099].
Khanzada teaches using trained classifiers and ensemble models to classify respiratory conditions, including asthma and respiratory syndromes. See Khanzada ¶¶[0016]–[0019], [0055]–[0060].
Wolf et al. likewise teach evaluating measured expiratory airflow and providing medication recommendations when respiratory performance indicates a possible chronic or asthmatic episode. See Wolf et al., claim 5 and col. 2, approximately ll. 33–60.
It would have been obvious to use enrolled respiratory-flow features and trained classifiers for both user classification and respiratory diagnosis because the references teach that respiratory signals contain information relevant to both user-specific characteristics and medical conditions.
It further would have been obvious to use the resulting medical-condition classification to individualize medication because Sezan et al. and Wolf et al. expressly teach adjusting or recommending medication based on breath-derived respiratory information.
Claim(s) 18 rejected under 35 U.S.C. 103 as being unpatentable over Sezan et al., (US20160106935A1) in view of Wolf et al., (US5518002A) in view of Khanzada., (US20220037022A1) and further in view of Raghunathan et al – NPL document titled “Multifractal analysis of flame dynamics during transition to thermoacoustic instability in a turbulent combustor” dated 06 February 2020.
Regarding claim 18, the combination of Sezan, Wolf and Khanzada fails to disclose:
The method as claimed in claim 17, wherein the extracted features include an abscissa corresponding to the spectral maxima (β), a width of the spectrum (ω), and a bias or asymmetry parameter of the spectrum (ϵ).
However, Raghunathan discloses:
characterizing a multifractal spectrum using a plurality of numerical features, including the most dominant singularity exponent , the spectrum width , and the asymmetry parameter . See Raghunathan et al., §3, “Multifractal analysis: Page A14-6.
Raghunathan et al. define the most dominant singularity exponent as the value of the horizontal-axis variable at which the multifractal spectrum reaches its maximum. Thus, is the abscissa corresponding to the spectral maximum and teaches the claimed parameter β. The difference between the symbol used by Raghunathan et al. and the symbol β used in claim 18 is merely a difference in nomenclature.
Raghunathan et al. further teach the spectrum width , which represents the horizontal extent or range of the multifractal spectrum and corresponds to the claimed spectrum width ω. Raghunathan et al. additionally teach an asymmetric parameter , which represents the skewness or imbalance between the left and right portions of the multifractal spectrum and therefore corresponds to the claimed bias or asymmetry parameter ε.
Raghunathan et al. apply these parameters to characterize different dynamical states associated with turbulent flow. Although Raghunathan et al. analyze turbulent combustion rather than exhaled breath, the reference is reasonably pertinent to the problem addressed by claim 18 because it teaches how to obtain compact numerical features from a multifractal spectrum representing turbulent-flow behavior.
It would have been obvious to one of ordinary skill in the art to apply the multifractal-spectrum parameters of Raghunathan et al. to the exhaled-airflow signal produced by the hot-wire anemometer of Wolf et al. in the breath-authentication system of Sezan et al., as processed using the feature-extraction and machine-learning techniques of Khanzada.
A skilled artisan would have been motivated to employ these known features because Raghunathan et al. teach that the peak location, width, and asymmetry characterize differences in the complexity, scaling behavior, and dynamical states of turbulent signals. Applying those features to an exhaled-flow signal would predictably produce a compact feature vector capable of being supplied to the authentication classifier.
Raghunathan does not independently disclose extracting these parameters from human breath.
Its role is to teach the three specific multifractal features; Wolf supplies the exhaled-flow signal, and Sezan and Khanzada supply the authentication and machine-learning context.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SYED M AHSAN whose telephone number is (571)272-5018. The examiner can normally be reached 8:30 AM - 6:00 PM.
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, William Korzuch can be reached at 571-272-7589. 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.
/SYED M AHSAN/Primary Examiner, Art Unit 2491