DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application is being examined under the pre-AIA first to invent provisions.
Claim Objections
The Claims are objected to because of the following informalities:
In Claim 23, line 16, the term “an FFT spectrum” should be replaced with –[[an]] a fast Fourier transform (FFT) spectrum-- for claim clarity.
In Claim 32, line 18, the term “an FFT spectrum” should be replaced with –[[an]] a fast Fourier transform (FFT) spectrum-- for claim clarity.
In Claim 39, line 24, the term “an FFT spectrum” should be replaced with –[[an]] a fast Fourier transform (FFT) spectrum-- for claim clarity.
Appropriate correction is required and applicant should carefully review the Claims for any other informalities.
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 23, 32, and 39 and claims dependent thereon rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor, or for pre-AIA the applicant regards as the invention.
Regarding Claim 23, lines 14-15, the term “(i) extracting a signal envelope;
(ii) Computing an auto-correlation function;" renders the claim indefinite because it is unclear how the auto-correlation function is being applied, whether on the extracted signal envelope or the audio respiratory signal of each frame. Upon review of the Specification, Examiner believes the auto-correlation function is applied to the envelope and suggest amending the claim to read -- extracting a signal envelope;
Computing an auto-correlation function of the signal envelope;--.
Similar changes should be applied to independent claims 32 and 39.
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 23-42 are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception (i.e., a law of nature, a natural phenomenon, or an abstract idea) without significantly more.
Regarding Claim 23, the claim(s) recites “analyzing the plurality of audio files to extract a plurality of audio respiratory signals;
training the deep learning process using the plurality of audio respiratory signals and the metadata;”
“for each frame of the plurality of overlapping frames, performing the following:
…
(ii) computing an auto-correlation function,
(iii) computing an FFT spectrum from the auto-correlation function,
(iv) computing a respiratory rate of the subject using the FFT spectrum, and
outputting the rate of progression of the lung, throat, and/or heart pathology of the subject using the deep learning process.” which amounts to an abstract idea (mental process and mathematical calculations).
This judicial exception is not integrated into a practical application because:
- The claims fail to outline an improvement to the technical field.
- The claims fail to apply the judicial exception to effect a particular treatment.
- The claims fail to apply the judicial exception with a particular machine.
- The claims fail to effect a transformation or reduction of a particular article to a different state or thing.
Next, the claim as a whole is analyzed to determine whether any element or a combination of elements, integrates judicial exception into a practical application.
For this part of the 101 analysis, the following additional limitations are considered:
“inputting a plurality of audio files comprising a training set into a deep learning process, wherein the plurality of audio files includes sessions with subjects with known pathologies of varying degrees of severity;
annotating the plurality of audio files with metadata associated with the subjects with known pathologies;”
“capturing an audio respiratory signal generated by a subject using a microphone;
segmenting the audio respiratory signal into a plurality of overlapping frames;
for each frame of the plurality of overlapping frames, performing the following:
(i) extracting a signal envelope,
…
(v) storing respiratory rates for the plurality of overlapping frames in computer memory;
inputting the stored respiratory rates of the subject into the deep learning process;”
The additional elements are insufficient to amount to significantly more than the judicial exception because they seem to merely generally link the use of the judicial exception to a particular technological environment.
Moreover, the claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception because they pertain merely to insignificant extrasolution data gathering activities and generic postsolution activity.
Furthermore, microphones are general field of use and computer memories are generic computer elements used to perform generic computer functions and don’t add significantly more and are well-understood, routine, and previously known to the industry.
None of these limitations, considered as an ordered combination provide eligibility because the claim taken as a whole, does not amount to significantly more than the underlying abstract idea of
Inputting training data to a deep learning process, processing the data to discretize the training data to generate input respiratory audio data, training the deep learning process using the input respiratory audio data and a broad classification of algorithms of deep learning, detecting respiratory data from overlapping frames of captured audio respiratory signals, computing a respiratory rate from the respiratory data, detecting a rate of progression of a lung, throat, and/or heart pathology using the trained deep learning, and outputting the pathology data from the trained deep learning and does not purport to improve the functioning of the signal processing, or to improve any other technology or technical field. Use of a generic signal processing does not amount to significantly more than the abstract idea itself. Dependent claims 24-31 also do not add significantly more to the exception as they merely add details to the mental steps, add details to the extrasolution data gathering steps, add general field of use components to facilitate the extrasolution data gathering, and add mental steps.
Regarding Claim 32, the claim(s) recites “analyzing the plurality of audio files to extract a plurality of audio respiratory signals;
training the deep learning process using the plurality of audio respiratory signals and the metadata;”
“for each frame of the plurality of overlapping frames, performing the following:
…
(ii) computing an auto-correlation function,
(iii) computing an FFT spectrum from the auto-correlation function,
(iv) computing a respiratory rate of the subject using the FFT spectrum, and
outputting the rate of progression of the lung, throat, and/or heart pathology of the subject using the deep learning process.” which amounts to an abstract idea (mental process and mathematical calculations).
This judicial exception is not integrated into a practical application because:
- The claims fail to outline an improvement to the technical field.
- The claims fail to apply the judicial exception to effect a particular treatment.
- The claims fail to apply the judicial exception with a particular machine.
- The claims fail to effect a transformation or reduction of a particular article to a different state or thing.
Next, the claim as a whole is analyzed to determine whether any element or a combination of elements, integrates judicial exception into a practical application.
For this part of the 101 analysis, the following additional limitations are considered:
“inputting a plurality of audio files comprising a training set into a deep learning process, wherein the plurality of audio files includes sessions with subjects with known pathologies of varying degrees of severity;
annotating the plurality of audio files with metadata associated with the subjects with known pathologies;”
“capturing an audio respiratory signal generated by a subject using a microphone;
segmenting the audio respiratory signal into a plurality of overlapping frames;
for each frame of the plurality of overlapping frames, performing the following:
(i) extracting a signal envelope,
…
(v) storing respiratory rates for the plurality of overlapping frames in computer memory;
inputting the stored respiratory rates of the subject into the deep learning process;”
The additional elements are insufficient to amount to significantly more than the judicial exception because they seem to merely generally link the use of the judicial exception to a particular technological environment.
Moreover, the claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception because they pertain merely to insignificant extrasolution data gathering activities and generic postsolution activity.
Furthermore, microphones are general field of use and non-transitory computer-readable storage media are generic computer elements used to perform generic computer functions and don’t add significantly more and are well-understood, routine, and previously known to the industry.
None of these limitations, considered as an ordered combination provide eligibility because the claim taken as a whole, does not amount to significantly more than the underlying abstract idea of
Inputting training data to a deep learning process, processing the data to discretize the training data to generate input respiratory audio data, training the deep learning process using the input respiratory audio data and a broad classification of algorithms of deep learning, detecting respiratory data from overlapping frames of captured audio respiratory signals, computing a respiratory rate from the respiratory data, detecting a rate of progression of a lung, throat, and/or heart pathology using the trained deep learning, and outputting the pathology data from the trained deep learning and does not purport to improve the functioning of the signal processing, or to improve any other technology or technical field. Use of a generic signal processing does not amount to significantly more than the abstract idea itself. Dependent claims 33-38 also do not add significantly more to the exception as they merely add details to the mental steps, add details to the extrasolution data gathering steps, add general field of use components to facilitate the extrasolution data gathering, and add mental steps.
Regarding Claim 39, the claim(s) recites “analyze the plurality of audio files to extract a plurality of audio respiratory signals;
train the deep learning process using the plurality of audio respiratory signals and the metadata;”
“for each frame of the plurality of overlapping frames, performing the following:
…
(ii) compute an auto-correlation function,
(iii) compute an FFT spectrum from the auto-correlation function,
(iv) compute a respiratory rate of the subject using the FFT spectrum, and
outputt the rate of progression of the lung, throat, and/or heart pathology of the subject using the deep learning process.” which amounts to an abstract idea (mental process and mathematical calculations).
This judicial exception is not integrated into a practical application because:
- The claims fail to outline an improvement to the technical field.
- The claims fail to apply the judicial exception to effect a particular treatment.
- The claims fail to apply the judicial exception with a particular machine.
- The claims fail to effect a transformation or reduction of a particular article to a different state or thing.
Next, the claim as a whole is analyzed to determine whether any element or a combination of elements, integrates judicial exception into a practical application.
For this part of the 101 analysis, the following additional limitations are considered:
“input a plurality of audio files comprising a training set into a deep learning process, wherein the plurality of audio files includes sessions with subjects with known pathologies of varying degrees of severity;
annotate the plurality of audio files with metadata associated with the subjects with known pathologies;”
“capture an audio respiratory signal generated by a subject using a microphone;
segment the audio respiratory signal into a plurality of overlapping frames;
for each frame of the plurality of overlapping frames, performing the following:
(i) extract a signal envelope,
…
(v) store respiratory rates for the plurality of overlapping frames in computer memory;
input the stored respiratory rates of the subject into the deep learning process;”
The additional elements are insufficient to amount to significantly more than the judicial exception because they seem to merely generally link the use of the judicial exception to a particular technological environment.
Moreover, the claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception because they pertain merely to insignificant extrasolution data gathering activities and generic postsolution activity.
Furthermore, nebulizers and microphones are general field of use and processors and computer memories are generic computer elements used to perform generic computer functions and don’t add significantly more and are well-understood, routine, and previously known to the industry.
None of these limitations, considered as an ordered combination provide eligibility because the claim taken as a whole, does not amount to significantly more than the underlying abstract idea of Inputting training data to a deep learning process, processing the data to discretize the training data to generate input respiratory audio data, training the deep learning process using the input respiratory audio data and a broad classification of algorithms of deep learning, detecting respiratory data from overlapping frames of captured audio respiratory signals, computing a respiratory rate from the respiratory data, detecting a rate of progression of a lung, throat, and/or heart pathology using the trained deep learning, and outputting the pathology data from the trained deep learning and does not purport to improve the functioning of the signal processing, or to improve any other technology or technical field. Use of a generic signal processing does not amount to significantly more than the abstract idea itself. Dependent claims 40-42 also do not add significantly more to the exception as they merely add details to the mental steps, add details to the extrasolution data gathering steps, add general field of use components to facilitate the extrasolution data gathering, and add mental steps.
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.
Claim(s) 23-25, 27, 30-34, and 36 is/are rejected under 35 U.S.C. 103 as being unpatentable over Umeki et al ("Classification between Normal and Abnormal Lung Sounds Using Unsupervised Subject-Adaptation”) (“Umeki”) in view of Alshaer et al (US 2014/0188006) (“Alshaer”) and further in view of Abe et al (US 2012/0150054) (“Abe”) and further in view of Yang et al (US 2011/0295139) (“Yang”) and further in view of Joeken (US 2010/0130874) and further in view of Snyder et al (US 7,463,922) (“Snyder”).
Regarding Claim 23, while Umeki teaches a method of detecting a rate of progression of lung, throat, and/or heart pathology (Abstract), the method comprising:
inputting a plurality of audio files comprising a training set into a pattern recognition process (p214, A. Training and Evaluation Data, “In the R2 auscultation point, 53 lung sound samples from 53 patients with pulmonary emphysema, and 53 samples from 53 healthy subjects were collected. Sixty-two samples from patients and healthy subjects were also collected in the R4 auscultation point, respectively. Each lung sound sample consisted of successive respiratory phase segments, and the average number of respiratory (inspiratory/expiratory) segments was eight. Each sample from the patients contained at least one phase segment that included adventitious sounds.” A. Architecture of Classification System, “In the training process, original acoustic HMMs Ɵ for each kind of segment were generated for each respiratory phase (inspiration/expiration) using the segment labels and their periods detected manually ((i) in Figure 2). Segment bigrams with reference to the frequency of occurrence sequences of the segments in abnormal respiratory periods were also estimated according to the segment labels [6].”), wherein the plurality of audio files includes sessions with subjects with a known pathology of a known degree of severity (p214, A. Training and Evaluation Data, normal subject and pulmonary emphysema patients);
annotating the plurality of audio files with metadata associated with the subjects with known pathologies (p214, A. Training and Evaluation Data “We tagged the segments according to the respiratory phase (inspiratory or expiratory), diagnostic state (normal or abnormal), auscultation point, and subjects’ health (healthy or patient). The subject’s health was determined by a doctor on the basis of auscultation and the existence of other medical conditions.”;
analyzing the plurality of audio files to extract a plurality of audio respiratory signals (p214, A. Training and Evaluation Data and B. Manual Segmentation of Acoustic Segments);
training the pattern recognition process using the plurality of audio respiratory signals and the metadata (Fig. 2, p214 A. Architecture of Classification System, “In the training process, original acoustic HMMs Ɵ for each kind of segment were generated for each respiratory phase (inspiration/expiration) using the segment labels and their periods detected manually ((i) in Figure 2). Segment bigrams with reference to the frequency of occurrence sequences of the segments in abnormal respiratory periods were also estimated according to the segment labels [6].”);
inputting test respiration data of the subject into the pattern recognition process (Fig. 2, p214 A. Architecture of Classification System, “The test process comprised three steps: first, for a test respiratory input, the likelihood for a normal candidate and that for an abnormal candidate were calculated for each respiration period using original HMMs. For the abnormal candidate, the most likely segment sequence and their segment periods were obtained. Second, subject-adaptation was performed using the obtained segment labels of confident respiratory phases ((ii) in Figure 2) and the manual labels of the training data (i) on the basis of the ML approach. Third, the classification was performed using the adapted HMMs.”) and
outputting the rate of progression of the lung, throat, and/or heart pathology of the subject using the deep learning process (Fig. 2, p214 A. Architecture of Classification System, “The test process comprised three steps: first, for a test respiratory input, the likelihood for a normal candidate and that for an abnormal candidate were calculated for each respiration period using original HMMs. For the abnormal candidate, the most likely segment sequence and their segment periods were obtained.” Where each respiration period was given a likelihood of normal or abnormal and thus reflects a progression of respiratory pathology and the rate of change), where audio data capture comprises using a microphone and segmenting the audio respiratory signal into a plurality of frames (p214, A. Training and Evaluation Data and B. Manual Segmentation of Acoustic Segments),
Umeki fails to teach the pattern recognition being performed by a deep learning process; and
wherein the plurality of audio files includes sessions with subjects with known pathologies of varying degrees of severity;
However Alshaer teaches a breathing disorder identification (Abstract) where pattern recognition for breathing disorder classification can be done with either hidden Markov models or artificial neural networks ([0213]-[0214] pattern recognition applied to breath sounds, [0242] where the pattern recognition can be performed alternatively by hidden Markov models or artificial neural networks);
teaches that multiple pathologies can be recognized from a breathing sound pattern recognition method (Abstract, [0059] central sleep apnea and obstructive sleep apnea); and
further teaches storing data results on a memory ([0091]).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to perform the pattern recognition method of Umeki as a deep learning process by artificial neural network taught by Alshaer as a simple substitution of one form of pattern recognition for another to obtain predictable results of accurately classified breath data as recognized by Alshaer ([0242]). Furthermore, it would be obvious to add pathologies taught by Alshaer to the pattern recognition of Umeki as this increase the utility of the analysis of Umeki. Finally, it would be obvious to use a memory to store analysis results for later application in trend monitoring or increase available training data.
Yet their combined efforts fail to teach
inputting the stored respiratory rates of the subject into the deep learning process; and
outputting the pathology from the results.
However Abe teaches a respiratory condition pattern recognition method (Abstract) where respiratory rate is a provided feature for the pattern recognition ([0033], [0037]-[0039]) and teaches that the different features can be combined to provide both a respiratory abnormality and a degree of respiratory abnormality ([0075]-[0079]).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, for the features of Umeki to further include respiration rate as taught by Abe as this and other features also enable the determination of a degree of abnormality in a respiratory condition, information that enables timely intervention for a patient.
Yet their combined efforts fail to teach the testing phase comprising:
segmenting the audio respiratory signal into a plurality of frames;
for each frame of the plurality of frames, performing the following:
(i) extracting a signal envelope,
(ii) computing an auto-correlation function,
(iii) computing an FFT spectrum from the auto-correlation function,
(iv) computing a respiratory rate of the subject using the FFT spectrum, and
(v) storing respiratory rates for the plurality of frames in computer memory.
However Yang teaches a method of determining respiratory rate from an audio respiratory signal (Abstract, [0004]), the method comprising:
capturing the audio respiratory signal generated by a subject using a microphone ([0018], [0031]);
segmenting the audio respiratory signal into a plurality of frames ([0037]);
for each frame of the plurality of frames performing the following:
extracting a signal envelope ([0018]);
computing an auto-correlation function ([0018]);
computing a fast Fourier (FFT) spectrum from the auto-correlation function;
computing a respiratory rate of the subject ([0035]); and
storing respiratory rate in computer memory ([0048] data management system logs respiration rate).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, that the required input of respiration rate from Umeki and can be derived from the method of Yang as can be performed on the basis of a microphone, hardware already included with the system of Umeki.
Yet their combined efforts fail to teach
computing a fast Fourier (FFT) spectrum from the auto-correlation function; and
computing a respiratory rate of the subject using the FFT spectrum.
However Joeken teaches a physiological parameter determining method (Abstract) wherein physiological data can be processed for cyclical physiological data ([0051]) and further teaches that a fast Fourier transform can be applied to an autocorrelation function and help differentiate between a respiratory rate and multiples of the frequency of the respiratory rate ([0051], [0054]).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to apply a fast Fourier transform of Joeken to the autocorrelation of Yang as an additional processing step to clarify the identification of the appropriate respiratory frequency.
Yet their combined efforts fail to teach
segmenting the audio respiratory signal into a plurality of overlapping frames;
for each frame of the plurality of overlapping frames performing the signal processing steps; and
storing respiratory rates for the plurality of overlapping frames.
However Snyder teaches a method for evaluating physiological data (Abstract) comprising segmenting the physiological signal into a plurality of overlapping frames ([0037]).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to segment the audio of Yang into overlapping frames as taught by Snyder as a way to optimize analysis accuracy and eliminate a boundary problem of discrete data windows (Snyder: Abstract).
Regarding Claim 24, Umeki, Alshaer, Abe, Yang, Joeken, and Snyder teach the method of claim 23, and Umeki further teaches the method comprising: updating the deep learning process based on new audio respiratory signals generated by new subjects (p214, A. Architecture of Classification System, “Second, subject-adaptation was performed using the obtained segment labels of confident respiratory phases ((ii) in Figure 2) and the manual labels of the training data (i) on the basis of the ML approach.” The process will be updated based on new audio respiratory signals from new subjects to optimize the detection in a subject-specific manner).
Regarding Claim 25, Umeki, Alshaer, Abe, Yang, Joeken, and Snyder teach the method of claim 23, wherein the deep learning process comprises a trained artificial neural network or a convolutional neural network (See Claim 23 Rejection)
Regarding Claim 27, while Umeki, Alshaer, Abe, Yang, Joeken, and Snyder teach the method of claim 32, and Snyder teaches that the overlap of the overlapping frames can be above 50% and can be optimized as desired by practitioner (Col. 5, L. 30 – Col. 6, L. 9), their combined efforts fail to teach wherein two or more frames of the plurality of overlapping frames overlap by at least 66%.
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to change the overlap of Snyder in Umeki, Alshaer, Abe, Yang, Joeken, and Snyder to above 66% as part of optimizing analysis accuracy and computational load. Further, in a case where the claimed ranges "overlap or lie inside ranges disclosed by the prior art" a prima facie case of obviousness exists -- here the prior art’s range of above 50% overlap intersects with a range of at least 66% overlap [In re Geisler, 116 F.3d 1465, 1469-71, 43 USPQ2d 1362, 1365-66 (Fed. Cir. 1997) (Claim reciting thickness of a protective layer as falling within a range of "50 to 100 Angstroms" considered prima facie obvious in view of prior art reference teaching that "for suitable protection, the thickness of the protective layer should be not less than about 10 nm [i.e., 100 Angstroms]." The court stated that "by stating that ‘suitable protection’ is provided if the protective layer is ‘about’ 100 Angstroms thick, [the prior art reference] directly teaches the use of a thickness within [applicant’s] claimed range.").]
Regarding Claim 30, Umeki, Alshaer, Abe, Yang, Joeken, and Snyder teach the method of claim 23, wherein computing the respiratory rate comprises:
determining a location of a peak magnitude of the FFT spectrum (See Claim 23 Rejection); and
computing one or more values associated with the respiratory rate using the peak magnitude (See Claim 23 Rejection, Yang: [0054]).
Regarding Claim 31, Umeki, Alshaer, Abe, Yang, Joeken, and Snyder teach the method of claim 30, and Yang further teaches the method comprising: applying median filtering to the one or more values associated with the respiratory rate to reduce inaccurate values ([0035] median filter applied to envelope).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to apply median filtering as taught Yang to the one or more values associated with the respiratory rate of Umeki, Alshaer, Abe, Yang, Joeken, and Snyder to reduce inaccurate values and lead to an improved pathology detecting method.
Regarding Claim 32, while Umeki teaches carrying out operations to detect a rate of progression of lung, throat, and/or heart pathology (Abstract), the method comprising:
inputting a plurality of audio files comprising a training set into a pattern recognition process (p214, A. Training and Evaluation Data, “In the R2 auscultation point, 53 lung sound samples from 53 patients with pulmonary emphysema, and 53 samples from 53 healthy subjects were collected. Sixty-two samples from patients and healthy subjects were also collected in the R4 auscultation point, respectively. Each lung sound sample consisted of successive respiratory phase segments, and the average number of respiratory (inspiratory/expiratory) segments was eight. Each sample from the patients contained at least one phase segment that included adventitious sounds.” A. Architecture of Classification System, “In the training process, original acoustic HMMs Ɵ for each kind of segment were generated for each respiratory phase (inspiration/expiration) using the segment labels and their periods detected manually ((i) in Figure 2). Segment bigrams with reference to the frequency of occurrence sequences of the segments in abnormal respiratory periods were also estimated according to the segment labels [6].”), wherein the plurality of audio files includes sessions with subjects with a known pathology of a known degree of severity (p214, A. Training and Evaluation Data, normal subject and pulmonary emphysema patients);
annotating the plurality of audio files with metadata associated with the subjects with known pathologies (p214, A. Training and Evaluation Data “We tagged the segments according to the respiratory phase (inspiratory or expiratory), diagnostic state (normal or abnormal), auscultation point, and subjects’ health (healthy or patient). The subject’s health was determined by a doctor on the basis of auscultation and the existence of other medical conditions.”;
analyzing the plurality of audio files to extract a plurality of audio respiratory signals (p214, A. Training and Evaluation Data and B. Manual Segmentation of Acoustic Segments);
training the pattern recognition process using the plurality of audio respiratory signals and the metadata (Fig. 2, p214 A. Architecture of Classification System, “In the training process, original acoustic HMMs Ɵ for each kind of segment were generated for each respiratory phase (inspiration/expiration) using the segment labels and their periods detected manually ((i) in Figure 2). Segment bigrams with reference to the frequency of occurrence sequences of the segments in abnormal respiratory periods were also estimated according to the segment labels [6].”);
inputting test respiration data of the subject into the pattern recognition process (Fig. 2, p214 A. Architecture of Classification System, “The test process comprised three steps: first, for a test respiratory input, the likelihood for a normal candidate and that for an abnormal candidate were calculated for each respiration period using original HMMs. For the abnormal candidate, the most likely segment sequence and their segment periods were obtained. Second, subject-adaptation was performed using the obtained segment labels of confident respiratory phases ((ii) in Figure 2) and the manual labels of the training data (i) on the basis of the ML approach. Third, the classification was performed using the adapted HMMs.”) and
outputting the rate of progression of the lung, throat, and/or heart pathology of the subject using the deep learning process (Fig. 2, p214 A. Architecture of Classification System, “The test process comprised three steps: first, for a test respiratory input, the likelihood for a normal candidate and that for an abnormal candidate were calculated for each respiration period using original HMMs. For the abnormal candidate, the most likely segment sequence and their segment periods were obtained.” Where each respiration period was given a likelihood of normal or abnormal and thus reflects a progression of respiratory pathology and the rate of change), where audio data capture comprises using a microphone and segmenting the audio respiratory signal into a plurality of frames (p214, A. Training and Evaluation Data and B. Manual Segmentation of Acoustic Segments),
Umeki fails to teach the operations being carried out by a non-transitory computer-readable storage medium having stored thereon, computer executable instructions;
the pattern recognition being performed by a deep learning process; and
wherein the plurality of audio files includes sessions with subjects with known pathologies of varying degrees of severity;
However Alshaer teaches a breathing disorder identification (Abstract) with operations being carried out by a non-transitory computer-readable storage medium having stored thereon, computer executable instructions ([0020]-[0022]);
where pattern recognition for breathing disorder classification can be done with either hidden Markov models or artificial neural networks ([0213]-[0214] pattern recognition applied to breath sounds, [0242] where the pattern recognition can be performed alternatively by hidden Markov models or artificial neural networks);
teaches that multiple pathologies can be recognized from a breathing sound pattern recognition method (Abstract, [0059] central sleep apnea and obstructive sleep apnea); and
further teaches storing data results on a memory ([0091]).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to perform the pattern recognition method of Umeki as a deep learning process by artificial neural network taught by Alshaer as a simple substitution of one form of pattern recognition for another to obtain predictable results of accurately classified breath data as recognized by Alshaer ([0242]). Further, it would have been obvious to apply the steps of Umeki with a processor and computer-readable media as these components can streamline the sampling and analysis of data. Furthermore, it would be obvious to add pathologies taught by Alshaer to the pattern recognition of Umeki as this increase the utility of the analysis of Umeki. Finally, it would be obvious to use a memory to store analysis results for later application in trend monitoring or increase available training data.
Yet their combined efforts fail to teach
inputting the stored respiratory rates of the subject into the deep learning process; and
outputting the pathology from the results.
However Abe teaches a respiratory condition pattern recognition method (Abstract) where respiratory rate is a provided feature for the pattern recognition ([0033], [0037]-[0039]) and teaches that the different features can be combined to provide both a respiratory abnormality and a degree of respiratory abnormality ([0075]-[0079]).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, for the features of Umeki to further include respiration rate as taught by Abe as this and other features also enable the determination of a degree of abnormality in a respiratory condition, information that enables timely intervention for a patient.
Yet their combined efforts fail to teach the testing phase comprising:
segmenting the audio respiratory signal into a plurality of frames;
for each frame of the plurality of frames, performing the following:
(i) extracting a signal envelope,
(ii) computing an auto-correlation function,
(iii) computing an FFT spectrum from the auto-correlation function,
(iv) computing a respiratory rate of the subject using the FFT spectrum, and
(v) storing respiratory rates for the plurality of frames in computer memory.
However Yang teaches a method of determining respiratory rate from an audio respiratory signal (Abstract, [0004]), the method comprising:
capturing the audio respiratory signal generated by a subject using a microphone ([0018], [0031]);
segmenting the audio respiratory signal into a plurality of frames ([0037]);
for each frame of the plurality of frames performing the following:
extracting a signal envelope ([0018]);
computing an auto-correlation function ([0018]);
computing a fast Fourier (FFT) spectrum from the auto-correlation function;
computing a respiratory rate of the subject ([0035]); and
storing respiratory rate in computer memory ([0048] data management system logs respiration rate).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, that the required input of respiration rate from Umeki and can be derived from the method of Yang as can be performed on the basis of a microphone, hardware already included with the system of Umeki.
Yet their combined efforts fail to teach
computing a fast Fourier (FFT) spectrum from the auto-correlation function; and
computing a respiratory rate of the subject using the FFT spectrum.
However Joeken teaches a physiological parameter determining method (Abstract) wherein physiological data can be processed for cyclical physiological data ([0051]) and further teaches that a fast Fourier transform can be applied to an autocorrelation function and help differentiate between a respiratory rate and multiples of the frequency of the respiratory rate ([0051], [0054]).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to apply a fast Fourier transform of Joeken to the autocorrelation of Yang as an additional processing step to clarify the identification of the appropriate respiratory frequency.
Yet their combined efforts fail to teach
segmenting the audio respiratory signal into a plurality of overlapping frames;
for each frame of the plurality of overlapping frames performing the signal processing steps; and
storing respiratory rates for the plurality of overlapping frames.
However Snyder teaches a method for evaluating physiological data (Abstract) comprising segmenting the physiological signal into a plurality of overlapping frames ([0037]).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to segment the audio of Yang into overlapping frames as taught by Snyder as a way to optimize analysis accuracy and eliminate a boundary problem of discrete data windows (Snyder: Abstract).
Regarding Claim 33, Umeki, Alshaer, Abe, Yang, Joeken, and Snyder teach the non-transitory computer-readable storage medium of claim 32, wherein the operations further include: updating the deep learning process based on new audio respiratory signals generated by new subjects (p214, A. Architecture of Classification System, “Second, subject-adaptation was performed using the obtained segment labels of confident respiratory phases ((ii) in Figure 2) and the manual labels of the training data (i) on the basis of the ML approach.” The process will be updated based on new audio respiratory signals from new subjects to optimize the detection in a subject-specific manner).
Regarding Claim 34, Umeki, Alshaer, Abe, Yang, Joeken, and Snyder teach the non-transitory computer-readable storage medium of claim 32, wherein the deep learning process comprises a trained artificial neural network or a convolutional neural network (See Claim 32 Rejection).
Regarding Claim 36, while Umeki, Alshaer, Abe, Yang, Joeken, and Snyder teach the non-transitory computer-readable storage medium of claim 32, and Snyder teaches that the overlap of the overlapping frames can be above 50% and can be optimized as desired by practitioner (Col. 5, L. 30 – Col. 6, L. 9), their combined efforts fail to teach wherein two or more frames of the plurality of overlapping frames overlap by at least 66%.
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to change the overlap of Snyder in Umeki, Alshaer, Abe, Yang, Joeken, and Snyder to above 66% as part of optimizing analysis accuracy and computational load. Further, in a case where the claimed ranges "overlap or lie inside ranges disclosed by the prior art" a prima facie case of obviousness exists -- here the prior art’s range of above 50% overlap intersects with a range of at least 66% overlap [In re Geisler, 116 F.3d 1465, 1469-71, 43 USPQ2d 1362, 1365-66 (Fed. Cir. 1997) (Claim reciting thickness of a protective layer as falling within a range of "50 to 100 Angstroms" considered prima facie obvious in view of prior art reference teaching that "for suitable protection, the thickness of the protective layer should be not less than about 10 nm [i.e., 100 Angstroms]." The court stated that "by stating that ‘suitable protection’ is provided if the protective layer is ‘about’ 100 Angstroms thick, [the prior art reference] directly teaches the use of a thickness within [applicant’s] claimed range.").]
Claim(s) 26 and 35 is/are rejected under 35 U.S.C. 103 as being unpatentable over Umeki in view of Alshaer and further in view of Abe and further in view of Yang and further in view of Joeken and further in view of Snyder and further in view of Halperin et al (US 2008/0275349) (“Halperin”).
Regarding Claim 26, while Umeki, Alshaer, Abe, Yang, Joeken, and Snyder teach the method of Claim 23, and Yang teaches wherein each of the plurality of frames has a duration of 15 seconds ([0037]), their combined efforts fail to teach wherein each of the plurality of overlapping frames has a duration of at least 30 seconds.
However Halperin teaches a physiological monitoring system (Abstract) and further teaches that the duration of a time frame for respiration judgement may be at least 30 seconds ([0781]).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to change the duration of time frame of Yang to at least 30 second as taught by Halperin as a larger time frame would be less likely to be skewed by noise or artifact causing a larger deviation in respiration rate. In turn, this prevents a treatment being applied unnecessarily.
Regarding Claim 35, while Umeki, Alshaer, Abe, Yang, Joeken, and Snyder teach the non-transitory computer-readable storage medium of claim 32, and Yang teaches wherein each of the plurality of frames has a duration of 15 seconds ([0037]), their combined efforts fail to teach wherein each of the plurality of overlapping frames has a duration of at least 30 seconds.
However Halperin teaches a physiological monitoring system (Abstract) and further teaches that the duration of a time frame for respiration judgement may be at least 30 seconds ([0781]).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to change the duration of time frame of Yang to at least 30 second as taught by Halperin as a larger time frame would be less likely to be skewed by noise or artifact causing a larger deviation in respiration rate. In turn, this prevents a treatment being applied unnecessarily.
Claim(s) 28 and 37 is/are rejected under 35 U.S.C. 103 as being unpatentable over Umeki in view of Alshaer and further in view of Abe and further in view of Yang and further in view of Joeken and further in view of Snyder and further in view of Desforges et al (US 2011/0034819) (“Desforges”).
Regarding Claim 28, while Umeki, Alshaer, Abe, Yang, Joeken, and Snyder teach the method of Claim 23, their combined efforts fail to teach wherein computing the auto-correlation function further comprises: filtering the auto-correlation function using low and high possible respiratory threshold values.
However Desforges teaches a respiratory-related medical device (Abstract and further teaches that a determination of a breathing frequency should utilize filtering of low and high possible threshold values ([0035]).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to apply a filter to the auto-correlation function of Yang using low and high possible respiratory threshold values as taught by Desforges as a means to ensure only relevant physiological data is processed, reducing computational load.
Regarding Claim 37, while Umeki, Alshaer, Abe, Yang, Joeken, and Snyder teach the non-transitory computer-readable storage medium of Claim 32, their combined efforts fail to teach wherein computing the auto-correlation function further comprises: filtering the auto-correlation function using low and high possible threshold respiratory values.
However Desforges teaches a respiratory-related medical device (Abstract and further teaches that a determination of a breathing frequency should utilize filtering of low and high possible threshold values ([0035]).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to apply a filter to the auto-correlation function of Yang using low and high possible respiratory threshold values as taught by Desforges as a means to ensure only relevant physiological data is processed, reducing computational load.
Claim(s) 29 and 38 is/are rejected under 35 U.S.C. 103 as being unpatentable over Umeki in view of Alshaer and further in view of Abe and further in view of Yang and further in view of Joeken and further in view of Snyder and further in view of Gamble et al (US 2013/0245347) (“Gamble”).
Regarding Claim 29, while Umeki, Alshaer, Abe, Yang, Joeken, and Snyder teach the method of Claim 23, and Yang teaches applying a high pass filter directly to the respiratory audio ([0033]), their combined efforts fail to teach wherein computing the auto-correlation function further comprises: filtering the auto-correlation function using a high-pass filter.
However Gamble teaches a physiological rate determination (Abstract, heart rate) comprising the use of segmenting radio-frequency data, an auto-correlation, high pass filtering the auto-correlation data, and FFT applied to the high-pass filtered auto-correlation data to output a heart rate (Fig. 3, [0098]).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to apply a high pass filter as taught by Gamble to the auto-correlation function of Yang as a way to remove unwanted frequencies as part of a similar physiological rate processing method (Gamble: heart rate, Yang: respiratory rate).
Regarding Claim 38, while Umeki, Alshaer, Abe, Yang, Joeken, and Snyder teach the non-transitory computer-readable storage medium of Claim 32, and Yang teaches applying a high pass filter directly to the respiratory audio ([0033]), their combined efforts fail to teach wherein computing the auto-correlation function further comprises: filtering the auto-correlation function using a high-pass filter.
However Gamble teaches a physiological rate determination (Abstract, heart rate) comprising the use of segmenting radio-frequency data, an auto-correlation, high pass filtering the auto-correlation data, and FFT applied to the high-pass filtered auto-correlation data to output a heart rate (Fig. 3, [0098]).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to apply a high pass filter as taught by Gamble to the auto-correlation function of Yang as a way to remove unwanted frequencies as part of a similar physiological rate processing method (Gamble: heart rate, Yang: respiratory rate).
Claim(s) 39-42 is/are rejected under 35 U.S.C. 103 as being unpatentable over Umeki in view of Alshaer and further in view of Abe and further in view of Yang and further in view of Joeken and further in view of Snyder and further in view of Vink et al (US 2014/0257126) (“Vink”).
Regarding Claim 39, while Umeki teaches a method for detecting a rate of progression of lung, throat, and/or heart pathology (Abstract), the method comprising:
inputting a plurality of audio files comprising a training set into a pattern recognition process (p214, A. Training and Evaluation Data, “In the R2 auscultation point, 53 lung sound samples from 53 patients with pulmonary emphysema, and 53 samples from 53 healthy subjects were collected. Sixty-two samples from patients and healthy subjects were also collected in the R4 auscultation point, respectively. Each lung sound sample consisted of successive respiratory phase segments, and the average number of respiratory (inspiratory/expiratory) segments was eight. Each sample from the patients contained at least one phase segment that included adventitious sounds.” A. Architecture of Classification System, “In the training process, original acoustic HMMs Ɵ for each kind of segment were generated for each respiratory phase (inspiration/expiration) using the segment labels and their periods detected manually ((i) in Figure 2). Segment bigrams with reference to the frequency of occurrence sequences of the segments in abnormal respiratory periods were also estimated according to the segment labels [6].”), wherein the plurality of audio files includes sessions with subjects with a known pathology of a known degree of severity (p214, A. Training and Evaluation Data, normal subject and pulmonary emphysema patients);
annotating the plurality of audio files with metadata associated with the subjects with known pathologies (p214, A. Training and Evaluation Data “We tagged the segments according to the respiratory phase (inspiratory or expiratory), diagnostic state (normal or abnormal), auscultation point, and subjects’ health (healthy or patient). The subject’s health was determined by a doctor on the basis of auscultation and the existence of other medical conditions.”;
analyzing the plurality of audio files to extract a plurality of audio respiratory signals (p214, A. Training and Evaluation Data and B. Manual Segmentation of Acoustic Segments);
training the pattern recognition process using the plurality of audio respiratory signals and the metadata (Fig. 2, p214 A. Architecture of Classification System, “In the training process, original acoustic HMMs Ɵ for each kind of segment were generated for each respiratory phase (inspiration/expiration) using the segment labels and their periods detected manually ((i) in Figure 2). Segment bigrams with reference to the frequency of occurrence sequences of the segments in abnormal respiratory periods were also estimated according to the segment labels [6].”);
inputting test respiration data of the subject into the pattern recognition process (Fig. 2, p214 A. Architecture of Classification System, “The test process comprised three steps: first, for a test respiratory input, the likelihood for a normal candidate and that for an abnormal candidate were calculated for each respiration period using original HMMs. For the abnormal candidate, the most likely segment sequence and their segment periods were obtained. Second, subject-adaptation was performed using the obtained segment labels of confident respiratory phases ((ii) in Figure 2) and the manual labels of the training data (i) on the basis of the ML approach. Third, the classification was performed using the adapted HMMs.”) and
outputting the rate of progression of the lung, throat, and/or heart pathology of the subject using the deep learning process (Fig. 2, p214 A. Architecture of Classification System, “The test process comprised three steps: first, for a test respiratory input, the likelihood for a normal candidate and that for an abnormal candidate were calculated for each respiration period using original HMMs. For the abnormal candidate, the most likely segment sequence and their segment periods were obtained.” Where each respiration period was given a likelihood of normal or abnormal and thus reflects a progression of respiratory pathology and the rate of change), where audio data capture comprises using a microphone and segmenting the audio respiratory signal into a plurality of frames (p214, A. Training and Evaluation Data and B. Manual Segmentation of Acoustic Segments),
Umeki fails to teach
a memory coupled to the nebulizer and operable to store the audio respiratory signal, wherein the memory further comprises an application for detecting the rate of progression of lung, throat, and/or heart pathology; and
a processor coupled to said memory, the processor being configured to operate in accordance with said application to:
the pattern recognition being performed by a deep learning process; and
wherein the plurality of audio files includes sessions with subjects with known pathologies of varying degrees of severity;
However Alshaer teaches a breathing disorder identification (Abstract) utilizing a memory coupled to the system and operable to store the audio respiratory signal, wherein the memory further comprises an application for detecting the rate of progression of lung, throat, and/or heart pathology ([0020]-[0022]); and
a processor coupled to said memory ([0020]-[0022]), the processor being configured to operate in accordance with said application to perform steps
where pattern recognition for breathing disorder classification can be done with either hidden Markov models or artificial neural networks ([0213]-[0214] pattern recognition applied to breath sounds, [0242] where the pattern recognition can be performed alternatively by hidden Markov models or artificial neural networks);
teaches that multiple pathologies can be recognized from a breathing sound pattern recognition method (Abstract, [0059] central sleep apnea and obstructive sleep apnea); and
further teaches storing data results on a memory ([0091]).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to perform the pattern recognition method of Umeki as a deep learning process by artificial neural network taught by Alshaer as a simple substitution of one form of pattern recognition for another to obtain predictable results of accurately classified breath data as recognized by Alshaer ([0242]). Further, it would have been obvious to apply the steps of Umeki with a processor and computer-readable media as these components can streamline the sampling and analysis of data. Furthermore, it would be obvious to add pathologies taught by Alshaer to the pattern recognition of Umeki as this increase the utility of the analysis of Umeki. Finally, it would be obvious to use a memory to store analysis results for later application in trend monitoring or increase available training data.
Yet their combined efforts fail to teach
inputting the stored respiratory rates of the subject into the deep learning process; and
outputting the pathology from the results.
However Abe teaches a respiratory condition pattern recognition method (Abstract) where respiratory rate is a provided feature for the pattern recognition ([0033], [0037]-[0039]) and teaches that the different features can be combined to provide both a respiratory abnormality and a degree of respiratory abnormality ([0075]-[0079]).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, for the features of Umeki to further include respiration rate as taught by Abe as this and other features also enable the determination of a degree of abnormality in a respiratory condition, information that enables timely intervention for a patient.
Yet their combined efforts fail to teach the testing phase comprising:
segmenting the audio respiratory signal into a plurality of frames;
for each frame of the plurality of frames, performing the following:
(i) extracting a signal envelope,
(ii) computing an auto-correlation function,
(iii) computing an FFT spectrum from the auto-correlation function,
(iv) computing a respiratory rate of the subject using the FFT spectrum, and
(v) storing respiratory rates for the plurality of frames in computer memory.
However Yang teaches a method of determining respiratory rate from an audio respiratory signal (Abstract, [0004]), the method comprising:
capturing the audio respiratory signal generated by a subject using a microphone ([0018], [0031]);
segmenting the audio respiratory signal into a plurality of frames ([0037]);
for each frame of the plurality of frames performing the following:
extracting a signal envelope ([0018]);
computing an auto-correlation function ([0018]);
computing a fast Fourier (FFT) spectrum from the auto-correlation function;
computing a respiratory rate of the subject ([0035]); and
storing respiratory rate in computer memory ([0048] data management system logs respiration rate).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, that the required input of respiration rate from Umeki and can be derived from the method of Yang as can be performed on the basis of a microphone, hardware already included with the system of Umeki.
Yet their combined efforts fail to teach
computing a fast Fourier (FFT) spectrum from the auto-correlation function; and
computing a respiratory rate of the subject using the FFT spectrum.
However Joeken teaches a physiological parameter determining method (Abstract) wherein physiological data can be processed for cyclical physiological data ([0051]) and further teaches that a fast Fourier transform can be applied to an autocorrelation function and help differentiate between a respiratory rate and multiples of the frequency of the respiratory rate ([0051], [0054]).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to apply a fast Fourier transform of Joeken to the autocorrelation of Yang as an additional processing step to clarify the identification of the appropriate respiratory frequency.
Yet their combined efforts fail to teach
segmenting the audio respiratory signal into a plurality of overlapping frames;
for each frame of the plurality of overlapping frames performing the signal processing steps; and
storing respiratory rates for the plurality of overlapping frames.
However Snyder teaches a method for evaluating physiological data (Abstract) comprising segmenting the physiological signal into a plurality of overlapping frames ([0037]).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to segment the audio of Yang into overlapping frames as taught by Snyder as a way to optimize analysis accuracy and eliminate a boundary problem of discrete data windows (Snyder: Abstract).
Yet their combined efforts fail to teach the system comprising a nebulizer communicatively coupled with a microphone, wherein the microphone is operable to capture audio respiratory signals from a subject;
the memory coupled to the nebulizer
a processor coupled to said memory and said nebulizer.
However Vink teaches a system for determining respiratory rate from an audio respiratory signal (Abstract), the system comprising:
a nebulizer communicatively coupled with a microphone, wherein the microphone is operable to capture the audio respiratory signal from a subject (Fig. 1, [0021] respiratory therapy device 103, monitoring device 105, [0024] therapy device may comprise nebulizer, [0026] monitoring device may comprise microphone);
a memory coupled to the nebulizer and operable to store the audio respiratory signal, wherein the memory further comprises an application for determining the respiratory rate from a breathing session stored therein ([0029] memory part of analysis portion 125); and
a processor coupled to said memory and said nebulizer ([0029] analysis portion comprises processor, coupled to memory and to nebulizer 113), the processor configured to operate in accordance with said application to:
capture the audio respiratory signal generated by the subject using the microphone ([0026]);
compute a respiratory parameter of the subject ([0037]).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, add a nebulizer of Vink to the system of Umeki, Alshaer, Abe, Yang, Joeken, and Snyder to enable combined monitoring and therapy and minimize the time between a recognized abnormality in a patient and an applied treatment.
Regarding Claim 40, Umeki, Alshaer, Abe, Yang, Joeken, Snyder, and Vink teach the system of claim 39, wherein the microphone, the processor and the memory are integrated with the nebulizer in a single device (See Claim 39 Rejection).
Regarding Claim 41, Umeki, Alshaer, Abe, Yang, Joeken, Snyder, and Vink teach the system of claim 39, wherein the processor is further configured to operate in accordance with said application to update the deep learning process based on new audio respiratory signals generated by new subjects (p214, A. Architecture of Classification System, “Second, subject-adaptation was performed using the obtained segment labels of confident respiratory phases ((ii) in Figure 2) and the manual labels of the training data (i) on the basis of the ML approach.” The process will be updated based on new audio respiratory signals from new subjects to optimize the detection in a subject-specific manner).
Regarding Claim 42, Umeki, Alshaer, Abe, Yang, Joeken, Snyder, and Vink teach the system of claim 39, wherein the deep learning process comprises a trained artificial neural network (See Claim 39 Rejection).
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JAIRO H PORTILLO whose telephone number is (571)272-1073. The examiner can normally be reached M-F 9:00 am - 5:15 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, Jacqueline Cheng can be reached at (571)272-5596. 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.
/JAIRO H. PORTILLO/
Examiner
Art Unit 3791
/PUYA AGAHI/Primary Examiner, Art Unit 3791
Prosecution Insights