DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Amendment
The proposed reply field on 12/29/2025 has been entered. Claims 1-20 remain pending in the current application. The amendments to the claims have overcome the claims objections and the 35 USC 112 rejections.
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 factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1-2, 4, 7-12, 14, 17-20 are rejected under 35 U.S.C. 103 as being unpatentable over Venkatraman et al (US Pub No. 2014/0288438) in the view of Ripoll (US Pub No. 2013/0012823).
Claim 1, Venkatraman teaches a wearable device configured to assess subject blood pressure, the wearable device comprising (figure 1, paragraphs 0121-0122; Portable biometric monitoring devices according to embodiments and implementations described herein may have shapes and sizes adapted for coupling to (e.g., secured to, worn, borne by, etc.) the body or clothing of a user. Indeed, the biometric monitoring device may measure or calculate a plurality of other physiological metrics in addition to, or in place of, the user's step count. These include, blood pressure):
a photoplethysmography (PPG) sensor configured to output a plurality of PPG waveforms (para. 0161; the signal processing logic receives two input signals: (1) heartbeat waveform signals from a PPG sensor 1703, for example, and (2) motion detection sensor 1705 output);
an inertial sensor configured to sense subject motion (para. 0161; the signal processing logic receives two input signals: (1) heartbeat waveform signals from a PPG sensor 1703, for example, and (2) motion detection sensor 1705 output); and
an assessment processor operatively connected to the PPG sensor and the inertial sensor, the assessment processor being configured to (para. 0161; the signal processing logic receives two input signals: (1) heartbeat waveform signals from a PPG sensor 1703, for example, and (2) motion detection sensor 1705 output.):
process inertial signals output by the inertial sensor to determine a data integrity of the plurality of PPG waveforms output by the PPG sensor representative of subject cadence (paras. 0136, 0166-0168, and 0176; the raw heartbeat waveform signal measured by a PPG sensor may be improved by using one or more algorithms to remove motion artifacts. In certain embodiments, data from a motion sensor is employed to gauge a user's motion and an adaptive filter is employed to remove the motion artifact from the heart rate signal when the user's motion is periodic. Movement of the user (for determining motion artifacts) may be measured using sensors including, but not limited to, accelerometers, gyroscopes, proximity detectors, magnetometers, etc. The goal of such algorithms is to remove components of the PPG signal attributable to movement (movement artifacts) using the movement signal captured from the other sensors as a guide. The examiner notes that the PPG waveform integrity is determined based on the received data of the motion sensor. The processor determines from the motion sensor the user’s activity level and cadence; when the processor determines that the user motion is not sedentary, then the processor determines that PPG waveform contains a motion artifact and has a poor data integrity. Additionally, when the processor determines from the motion sensor that the user is not moving, then the PPG waveform does not have a motion artifact and has good data integrity.);
modify at least one of the plurality of PPG waveforms output by the PPG sensor based on the determined data integrity representative of subject cadence exceeding at least one threshold (paras. 0132, 0136, and 0176-0177; Initially, the first activity mode is determined based on the user's motion. Many criteria may be employed to determine the user's activity mode. At a minimum, the devices should receive sensor output suggesting that the user is engaged in some activity and is not sedentary. In a typical implementation, a motion sensor output shows that the user or a user's limb is moving at a reasonable rate. Regardless of how the signal processing logic (e.g., block 1717) determines that the user is participating in the first activity, it begins processing of the buffered output data via the first activity channel. In this channel, the logic assumes that there is a motion artifact that should be removed or reduced before calculating the user's heart rate. The first activity channel may employ an adaptive filter to remove or reduce the motion artifacts from the time domain signal. See block 1721 in the embodiments of FIG. 17. In certain implementations, the adaptive filter attempts to predict the heartbeat waveform sensor output data from the motion sensor output data. This determines the motion artifact in the heartbeat waveform sensor output because the motion artifact is the only component common to the two output signals. Stated another way, the adaptive filter subtracts the motion artifact from the heart rate output signal to provide "cleaned" heart rate output data. Examples of adaptive filtering that may be employed include least mean square filtering and recursive least squares filtering. Optionally, the processing logic passes the cleaned heartbeat waveform output signal through the adaptive filter a second time, as explained above. The examiner notes that the processor determines the data integrity based on determining if the data comprise motion artifacts. When the processor determines that the motion of the user exceeds a certain threshold (Activity level 1, not stationary), then the PPG waveform has poor data integrity (motion artifacts) and needs modification. The modification is done by filtering out the motion artifact in the waveform.);
process the modified plurality of PPG waveforms to generate an assessment of the subject blood pressure (paras. 0122 and 0161-0183; Further processing of the heartbeat waveform signal in the first activity channel may proceed generally as described above for the stationary channel. However, identifying the heartbeat track may employ more sophisticated processing. The following are operations that may be employed in first activity channel. Employ an FFT or other frequency domain conversion technique to get an "estimate" of the user's heart rate from the time domain data output from the adaptive filter. Identify a heart rate track in the spectral representation of the filtered output data from the heartbeat waveform sensor. Smooth the estimate to get a smoothed output heart rate. Assess the "confidence" of the output heart rate. Present the calculated heart rate to the user. This information may be presented along with the confidence level and/or a recommendation for adjusting device position to improve confidence.).
However, Venkatraman fails to explicitly teach using a neural network comprising thousands of coefficients and/or a machine learning model to process the PPG waveforms and generate an assessment of the subject blood pressure.
Ripoll, in the same field of endeavor, teaches process the plurality of PPG waveforms using a neural network comprising thousands of coefficients and/or a machine learning model to generate an assessment of the subject blood pressure (para. 0040; This information may then be pre-processed to obtain a fixed length vector describing the PPG signal and incorporating the clinical parameters of the patient. This fixed length vector may be used by the trained machine learning algorithm to calculate the blood pressure of the patient.).
It would have been obvious to one in the ordinary skill in the art before the effective filling date of the claimed invention to have modified the processing steps of Venkatraman to incorporate the teachings of Ripoll to provide a machine learning model to process PPG waveforms and generate an assessment of the subject blood pressure. This modification will reduce the estimation of errors in the post-processing stage as taught within Ripoll in paragraph 0035. Additionally using machine learning models to estimate blood pressure will result in more accurate, personalized, and robust blood pressure estimation.
Claim 2, Venkatraman teaches the wearable device of claim 1, wherein the assessment processor is further configured to change a polling of the PPG sensor responsive to the assessment processor detecting a steady subject cadence from the inertial signals output by the inertial sensor and/or a steady subject heart rate from the plurality PPG waveforms output by the PPG sensor (paras. 0132 and 0135; a biometric monitoring device may employ data (for example, from one or more motion sensors) indicative of user activity or motion to adjust or modify characteristics of triggering, acquiring, and/or obtaining desired heart rate measurements or data (for example, to improve robustness to motion artifact). For instance, if the biometric monitoring device receives data indicative of user activity or motion, the biometric monitoring device may adjust or modify the sampling rate and/or resolution mode of sensors used to acquire heart rate data (for example, where the amount of user motion exceeds a certain threshold, the biometric monitoring device may increase the sampling rate and/or increase the sampling resolution mode of sensors employed to acquire heart rate measurement or data. For example, in one embodiment, a biometric monitoring device (or heart-rate measurement technique as disclosed herein may adjust and/or reduce the sampling rate of optical heart rate sampling when motion detector circuitry detects or determines that the biometric monitoring device wearer's motion is below a threshold (for example, if the biometric monitoring device determines the user is sedentary or asleep).).
Claim 4, Venkatraman teaches the wearable device of claim 1, wherein: the assessment processor is further configured to buffer the modified plurality of PPG waveforms and generate a plurality of representations of the modified plurality of PPG waveforms (paras. 0163-0167; the signal processing logic buffers the pre-processed heartbeat waveform sensor signal in a buffer 1713 and buffers the pre-processed motion sensor signal in a buffer 1715. In each of the channels, the signal processing logic converts the time domain signals acquired from the sensors to frequency domain signals.).
to process the modified plurality of PPG waveforms, the assessment processor is configured to process the plurality of representations of the modified plurality of PPG waveforms using the neural network comprising thousands of coefficients and/or the machine learning model to generate the assessment of the subject blood pressure (paras. 0122 and 0161-0183; Further processing of the heartbeat waveform signal in the first activity channel may proceed generally as described above for the stationary channel. However, identifying the heartbeat track may employ more sophisticated processing. The following are operations that may be employed in first activity channel. Employ an FFT or other frequency domain conversion technique to get an "estimate" of the user's heart rate from the time domain data output from the adaptive filter. Identify a heart rate track in the spectral representation of the filtered output data from the heartbeat waveform sensor. Smooth the estimate to get a smoothed output heart rate. Assess the "confidence" of the output heart rate. Present the calculated heart rate to the user. This information may be presented along with the confidence level and/or a recommendation for adjusting device position to improve confidence.)
However, Venkatraman fails to explicitly teach process the plurality of PPG waveforms using the neural network comprising thousands of coefficients and/or the machine learning model to generate the assessment of the subject blood pressure.
Ripoll, in the same field of endeavor, teaches process the plurality of PPG waveforms using a neural network comprising thousands of coefficients and/or a machine learning model to generate an assessment of the subject blood pressure (para. 0040; This information may then be pre-processed to obtain a fixed length vector describing the PPG signal and incorporating the clinical parameters of the patient. This fixed length vector may be used by the trained machine learning algorithm to calculate the blood pressure of the patient.).
It would have been obvious to one in the ordinary skill in the art before the effective filling date of the claimed invention to have modified the processing steps of Venkatraman to incorporate the teachings of Ripoll to provide a machine learning model to process PPG waveforms and generate an assessment of the subject blood pressure. This modification will reduce the estimation of errors in the post-processing stage as taught within Ripoll in paragraph 0035. Additionally using machine learning models to estimate blood pressure will result in more accurate, personalized, and robust blood pressure estimation.
Claim 7, Venkatraman teaches the wearable device of claim 4, wherein at least one of the plurality of representations comprises a spectral representation of at least one of the modified plurality PPG waveforms (para. 0180; Identify a heart rate track in the spectral representation of the filtered output data from the heartbeat waveform sensor. The examiner notes that the filtered PPG waveforms are used to generate a spectral representation of the PPG waveform.).
Claim 8, Venkatraman teaches the wearable device of claim 1, wherein: the assessment processor is further configured to determine a spectral representation of at least one of the modified plurality of PPG waveforms (para. 0180; Identify a heart rate track in the spectral representation of the filtered output data from the heartbeat waveform sensor.); and
to process the modified plurality of PPG waveforms, the assessment processor is configured to, responsive to the determined data integrity, process the spectral representation to generate the assessment of the subject blood pressure (paras. 0122 and 0161-0183; Further processing of the heartbeat waveform signal in the first activity channel may proceed generally as described above for the stationary channel. However, identifying the heartbeat track may employ more sophisticated processing. The following are operations that may be employed in first activity channel. Employ an FFT or other frequency domain conversion technique to get an "estimate" of the user's heart rate from the time domain data output from the adaptive filter. Identify a heart rate track in the spectral representation of the filtered output data from the heartbeat waveform sensor. Smooth the estimate to get a smoothed output heart rate. Assess the "confidence" of the output heart rate. Present the calculated heart rate to the user. This information may be presented along with the confidence level and/or a recommendation for adjusting device position to improve confidence.).
However, Venkatraman fails to explicitly teach using a neural network comprising thousands of coefficients and/or a machine learning model to process the PPG waveforms and generate an assessment of the subject blood pressure.
Ripoll, in the same field of endeavor, teaches process the plurality of PPG waveforms using a neural network comprising thousands of coefficients and/or a machine learning model to generate an assessment of the subject blood pressure (para. 0040; This information may then be pre-processed to obtain a fixed length vector describing the PPG signal and incorporating the clinical parameters of the patient. This fixed length vector may be used by the trained machine learning algorithm to calculate the blood pressure of the patient.).
It would have been obvious to one in the ordinary skill in the art before the effective filling date of the claimed invention to have modified the processing steps of Venkatraman to incorporate the teachings of Ripoll to provide a machine learning model to process PPG waveforms and generate an assessment of the subject blood pressure. This modification will reduce the estimation of errors in the post-processing stage as taught within Ripoll in paragraph 0035. Additionally using machine learning models to estimate blood pressure will result in more accurate, personalized, and robust blood pressure estimation.
Claim 9, Venkatraman teaches the wearable device of claim 1, wherein the assessment processor is further configured to change a sampling rate of the PPG sensor responsive to the assessment processor detecting a steady subject cadence from the inertial signals output by the inertial sensor and/or detecting a steady subject heart rate from the PPG waveforms output by the PPG sensor (paras. 0132 and 0135; a biometric monitoring device may employ data (for example, from one or more motion sensors) indicative of user activity or motion to adjust or modify characteristics of triggering, acquiring, and/or obtaining desired heart rate measurements or data (for example, to improve robustness to motion artifact). For instance, if the biometric monitoring device receives data indicative of user activity or motion, the biometric monitoring device may adjust or modify the sampling rate and/or resolution mode of sensors used to acquire heart rate data (for example, where the amount of user motion exceeds a certain threshold, the biometric monitoring device may increase the sampling rate and/or increase the sampling resolution mode of sensors employed to acquire heart rate measurement or data. For example, in one embodiment, a biometric monitoring device (or heart-rate measurement technique as disclosed herein may adjust and/or reduce the sampling rate of optical heart rate sampling when motion detector circuitry detects or determines that the biometric monitoring device wearer's motion is below a threshold (for example, if the biometric monitoring device determines the user is sedentary or asleep).).
Claim 10, Venkatraman teaches the wearable device of claim 1, wherein the wearable device comprises a device worn at the wrist, arm, leg, digits, nose, head, neck, or torso (para. 0120; Standalone biometric monitoring devices may be provided in a number of form factors and may be designed to be worn in a variety of ways. In some implementations, a biometric monitoring device may be designed to be insertable into a wearable case or into multiple, different wearable cases, e.g., a wristband case, a belt-clip case, a pendant case, a case configured to be attached to a piece of exercise equipment such as a bicycle, etc. Such implementations are described in more detail in, for example, U.S. patent application Ser. No. 14/029,764, filed Sep. 17, 2013, which is hereby incorporated by reference for such purpose. In other implementations, a biometric monitoring device may be designed to be worn in only one manner, e.g., a biometric monitoring device that is integrated into a wristband in a non-removable manner may be intended to be worn only on a person's wrist (or perhaps ankle).).
Claim 11, Venkatraman teaches a method of assessing subject blood pressure via a wearable device, the method comprising (figure 1, paragraphs 0121-0122; Portable biometric monitoring devices according to embodiments and implementations described herein may have shapes and sizes adapted for coupling to (e.g., secured to, worn, borne by, etc.) the body or clothing of a user. Indeed, the biometric monitoring device may measure or calculate a plurality of other physiological metrics in addition to, or in place of, the user's step count. These include, blood pressure):
Collecting a photoplethysmography (PPG) waveforms from a PPG sensor in the wearable device (para. 0161; the signal processing logic receives two input signals: (1) heartbeat waveform signals from a PPG sensor 1703, for example, and (2) motion detection sensor 1705 output);
collecting inertial data associated with subject motion from an inertial sensor in the wearable device (para. 0161; the signal processing logic receives two input signals: (1) heartbeat waveform signals from a PPG sensor 1703, for example, and (2) motion detection sensor 1705 output); and
processing the inertial data in an assessment processor operatively connected to the PPG sensor and the inertial sensor to determine a data integrity of the plurality of PPG waveforms representative of subject cadence (paras. 0136, 0166-0168, and 0176; the raw heartbeat waveform signal measured by a PPG sensor may be improved by using one or more algorithms to remove motion artifacts. In certain embodiments, data from a motion sensor is employed to gauge a user's motion and an adaptive filter is employed to remove the motion artifact from the heart rate signal when the user's motion is periodic. Movement of the user (for determining motion artifacts) may be measured using sensors including, but not limited to, accelerometers, gyroscopes, proximity detectors, magnetometers, etc. The goal of such algorithms is to remove components of the PPG signal attributable to movement (movement artifacts) using the movement signal captured from the other sensors as a guide. The examiner notes that the PPG waveform integrity is determined based on the received data of the motion sensor. The processor determines from the motion sensor the user’s activity level and cadence; when the processor determines that the user motion is not sedentary, then the processor determines that PPG waveform contains a motion artifact and has a poor data integrity. Additionally, when the processor determines from the motion sensor that the user is not moving, then the PPG waveform does not have a motion artifact and has good data integrity.);
filter at least one of the plurality of PPG waveforms from the PPG sensor based on the determined data integrity representative of subject cadence exceeding at least one threshold (paras. 0132, 0136, and 0176-0177; Initially, the first activity mode is determined based on the user's motion. Many criteria may be employed to determine the user's activity mode. At a minimum, the devices should receive sensor output suggesting that the user is engaged in some activity and is not sedentary. In a typical implementation, a motion sensor output shows that the user or a user's limb is moving at a reasonable rate. Regardless of how the signal processing logic (e.g., block 1717) determines that the user is participating in the first activity, it begins processing of the buffered output data via the first activity channel. In this channel, the logic assumes that there is a motion artifact that should be removed or reduced before calculating the user's heart rate. The first activity channel may employ an adaptive filter to remove or reduce the motion artifacts from the time domain signal. See block 1721 in the embodiments of FIG. 17. In certain implementations, the adaptive filter attempts to predict the heartbeat waveform sensor output data from the motion sensor output data. This determines the motion artifact in the heartbeat waveform sensor output because the motion artifact is the only component common to the two output signals. Stated another way, the adaptive filter subtracts the motion artifact from the heart rate output signal to provide "cleaned" heart rate output data. Examples of adaptive filtering that may be employed include least mean square filtering and recursive least squares filtering. Optionally, the processing logic passes the cleaned heartbeat waveform output signal through the adaptive filter a second time, as explained above. The examiner notes that the processor determines the data integrity based on determining if the data comprise motion artifacts. When the processor determines that the motion of the user exceeds a certain threshold (Activity level 1, not stationary), then the PPG waveform has poor data integrity (motion artifacts) and needs modification. The modification is done by filtering out the motion artifact in the waveform.);
processing the filtered plurality of PPG waveforms to generate an assessment of the subject blood pressure (paras. 0122 and 0161-0183; Further processing of the heartbeat waveform signal in the first activity channel may proceed generally as described above for the stationary channel. However, identifying the heartbeat track may employ more sophisticated processing. The following are operations that may be employed in first activity channel. Employ an FFT or other frequency domain conversion technique to get an "estimate" of the user's heart rate from the time domain data output from the adaptive filter. Identify a heart rate track in the spectral representation of the filtered output data from the heartbeat waveform sensor. Smooth the estimate to get a smoothed output heart rate. Assess the "confidence" of the output heart rate. Present the calculated heart rate to the user. This information may be presented along with the confidence level and/or a recommendation for adjusting device position to improve confidence.).
However, Venkatraman fails to explicitly teach using a neural network comprising thousands of coefficients and/or a machine learning model to process the PPG waveforms and generate an assessment of the subject blood pressure.
Ripoll, in the same field of endeavor, teaches process the plurality of PPG waveforms using a neural network comprising thousands of coefficients and/or a machine learning model to generate an assessment of the subject blood pressure (para. 0040; This information may then be pre-processed to obtain a fixed length vector describing the PPG signal and incorporating the clinical parameters of the patient. This fixed length vector may be used by the trained machine learning algorithm to calculate the blood pressure of the patient.).
It would have been obvious to one in the ordinary skill in the art before the effective filling date of the claimed invention to have modified the processing steps of Venkatraman to incorporate the teachings of Ripoll to provide a machine learning model to process PPG waveforms and generate an assessment of the subject blood pressure. This modification will reduce the estimation of errors in the post-processing stage as taught within Ripoll in paragraph 0035. Additionally using machine learning models to estimate blood pressure will result in more accurate, personalized, and robust blood pressure estimation.
Claim 12, Venkatraman teaches the method of claim 11, further comprising changing, by the assessment processor, a polling of the PPG sensor responsive to the assessment processor detecting a steady subject cadence from the inertial signals output by the inertial sensor and/or a steady subject heart rate from the PPG waveforms output by the PPG sensor (paras. 0132 and 0135; a biometric monitoring device may employ data (for example, from one or more motion sensors) indicative of user activity or motion to adjust or modify characteristics of triggering, acquiring, and/or obtaining desired heart rate measurements or data (for example, to improve robustness to motion artifact). For instance, if the biometric monitoring device receives data indicative of user activity or motion, the biometric monitoring device may adjust or modify the sampling rate and/or resolution mode of sensors used to acquire heart rate data (for example, where the amount of user motion exceeds a certain threshold, the biometric monitoring device may increase the sampling rate and/or increase the sampling resolution mode of sensors employed to acquire heart rate measurement or data. For example, in one embodiment, a biometric monitoring device (or heart-rate measurement technique as disclosed herein may adjust and/or reduce the sampling rate of optical heart rate sampling when motion detector circuitry detects or determines that the biometric monitoring device wearer's motion is below a threshold (for example, if the biometric monitoring device determines the user is sedentary or asleep).).
Claim 14, Venkatraman teaches the method of claim 11, further comprising the assessment processor: buffering the filtered plurality of PPG waveforms; and generating a plurality of representations of the filtered plurality of PPG waveforms (paras. 0163-0167; the signal processing logic buffers the pre-processed heartbeat waveform sensor signal in a buffer 1713 and buffers the pre-processed motion sensor signal in a buffer 1715. In each of the channels, the signal processing logic converts the time domain signals acquired from the sensors to frequency domain signals.),
wherein the processing of the filtered plurality of PPG waveforms comprises processing, by the assessment processor, the plurality of representations of the filtered plurality of PPG waveforms to generate the assessment of the subject blood pressure (paras. 0122 and 0161-0183; Further processing of the heartbeat waveform signal in the first activity channel may proceed generally as described above for the stationary channel. However, identifying the heartbeat track may employ more sophisticated processing. The following are operations that may be employed in first activity channel. Employ an FFT or other frequency domain conversion technique to get an "estimate" of the user's heart rate from the time domain data output from the adaptive filter. Identify a heart rate track in the spectral representation of the filtered output data from the heartbeat waveform sensor. Smooth the estimate to get a smoothed output heart rate. Assess the "confidence" of the output heart rate. Present the calculated heart rate to the user. This information may be presented along with the confidence level and/or a recommendation for adjusting device position to improve confidence.).
However, Venkatraman fails to explicitly teach process the of the plurality of PPG waveforms using the neural network comprising thousands of coefficients and/or the machine learning model to generate the assessment of the subject blood pressure.
Ripoll, in the same field of endeavor, teaches process the plurality of PPG waveforms using a neural network comprising thousands of coefficients and/or a machine learning model to generate an assessment of the subject blood pressure (para. 0040; This information may then be pre-processed to obtain a fixed length vector describing the PPG signal and incorporating the clinical parameters of the patient. This fixed length vector may be used by the trained machine learning algorithm to calculate the blood pressure of the patient.).
It would have been obvious to one in the ordinary skill in the art before the effective filling date of the claimed invention to have modified the processing steps of Venkatraman to incorporate the teachings of Ripoll to provide a machine learning model to process PPG waveforms and generate an assessment of the subject blood pressure. This modification will reduce the estimation of errors in the post-processing stage as taught within Ripoll in paragraph 0035. Additionally using machine learning models to estimate blood pressure will result in more accurate, personalized, and robust blood pressure estimation.
Claim 17, Venkatraman teaches the method of claim 14, wherein at least one of the plurality of representations comprises a spectral representation of at least one of the filtered plurality of PPG waveforms (para. 0180; Identify a heart rate track in the spectral representation of the filtered output data from the heartbeat waveform sensor. The examiner notes that the filtered PPG waveforms are used to generate a spectral representation of the PPG waveform.).
Claim 18, Venkatraman teaches the method of claim 11, further comprising determining a spectral representation of at least one of the filtered plurality of PPG waveforms, wherein the processing of the filtered plurality of PPG waveforms comprises (para. 0180; Identify a heart rate track in the spectral representation of the filtered output data from the heartbeat waveform sensor.); and
processing in the assessment processor the spectral representation to generate the assessment of the subject blood pressure (paras. 0122 and 0161-0183; Further processing of the heartbeat waveform signal in the first activity channel may proceed generally as described above for the stationary channel. However, identifying the heartbeat track may employ more sophisticated processing. The following are operations that may be employed in first activity channel. Employ an FFT or other frequency domain conversion technique to get an "estimate" of the user's heart rate from the time domain data output from the adaptive filter. Identify a heart rate track in the spectral representation of the filtered output data from the heartbeat waveform sensor. Smooth the estimate to get a smoothed output heart rate. Assess the "confidence" of the output heart rate. Present the calculated heart rate to the user. This information may be presented along with the confidence level and/or a recommendation for adjusting device position to improve confidence.).
However, Venkatraman fails to explicitly teach using a neural network comprising thousands of coefficients and/or a machine learning model to process the PPG waveforms and generate an assessment of the subject blood pressure.
Ripoll, in the same field of endeavor, teaches process the plurality of PPG waveforms using a neural network comprising thousands of coefficients and/or a machine learning model to generate an assessment of the subject blood pressure (para. 0040; This information may then be pre-processed to obtain a fixed length vector describing the PPG signal and incorporating the clinical parameters of the patient. This fixed length vector may be used by the trained machine learning algorithm to calculate the blood pressure of the patient.).
It would have been obvious to one in the ordinary skill in the art before the effective filling date of the claimed invention to have modified the processing steps of Venkatraman to incorporate the teachings of Ripoll to provide a machine learning model to process PPG waveforms and generate an assessment of the subject blood pressure. This modification will reduce the estimation of errors in the post-processing stage as taught within Ripoll in paragraph 0035. Additionally using machine learning models to estimate blood pressure will result in more accurate, personalized, and robust blood pressure estimation.
Claim 19, Venkatraman teaches the method of claim 11, further comprising the assessment processor changing a sampling rate of the PPG sensor responsive to the assessment processor detecting a steady subject cadence from the inertial data output by the inertial sensor and/or detecting a steady subject heart rate from the plurality PPG waveforms output by the PPG sensor (paras. 0132 and 0135; a biometric monitoring device may employ data (for example, from one or more motion sensors) indicative of user activity or motion to adjust or modify characteristics of triggering, acquiring, and/or obtaining desired heart rate measurements or data (for example, to improve robustness to motion artifact). For instance, if the biometric monitoring device receives data indicative of user activity or motion, the biometric monitoring device may adjust or modify the sampling rate and/or resolution mode of sensors used to acquire heart rate data (for example, where the amount of user motion exceeds a certain threshold, the biometric monitoring device may increase the sampling rate and/or increase the sampling resolution mode of sensors employed to acquire heart rate measurement or data. For example, in one embodiment, a biometric monitoring device (or heart-rate measurement technique as disclosed herein may adjust and/or reduce the sampling rate of optical heart rate sampling when motion detector circuitry detects or determines that the biometric monitoring device wearer's motion is below a threshold (for example, if the biometric monitoring device determines the user is sedentary or asleep).).
Claim 20, Venkatraman teaches the method of claim 11, wherein: the wearable device comprises a device worn at the wrist, arm, leg, digits, nose, head, neck, or torso; and the collecting of the plurality of PPG waveforms comprises collecting the plurality of PPG waveforms from the wrist, arm, leg, digits, nose, head, neck, or torso of the subject (para. 0120; Standalone biometric monitoring devices may be provided in a number of form factors and may be designed to be worn in a variety of ways. In some implementations, a biometric monitoring device may be designed to be insertable into a wearable case or into multiple, different wearable cases, e.g., a wristband case, a belt-clip case, a pendant case, a case configured to be attached to a piece of exercise equipment such as a bicycle, etc. Such implementations are described in more detail in, for example, U.S. patent application Ser. No. 14/029,764, filed Sep. 17, 2013, which is hereby incorporated by reference for such purpose. In other implementations, a biometric monitoring device may be designed to be worn in only one manner, e.g., a biometric monitoring device that is integrated into a wristband in a non-removable manner may be intended to be worn only on a person's wrist (or perhaps ankle).).
Claims 3, 5-6, 13, and 15-16 are rejected under 35 U.S.C. 103 as being unpatentable over Venkatraman et al (US Pub No. 2014/0288438) in the view of Ripoll et al (US Pub No. 2013/0012823) in further view of Unser et al (NPL: “Splines a perfect for signal and image processing”).
Claim 3, Venkatraman teaches the wearable device of claim 2, the assessment processor is further configured to create the modified plurality of PPG waveforms (paras. 0136, 0146-0149, and 0161-0183; Initially, the first activity mode is determined based on the user's motion. Many criteria may be employed to determine the user's activity mode. At a minimum, the devices should receive sensor output suggesting that the user is engaged in some activity and is not sedentary. In a typical implementation, a motion sensor output shows that the user or a user's limb is moving at a reasonable rate. Regardless of how the signal processing logic (e.g., block 1717) determines that the user is participating in the first activity, it begins processing of the buffered output data via the first activity channel. In this channel, the logic assumes that there is a motion artifact that should be removed or reduced before calculating the user's heart rate. The first activity channel may employ an adaptive filter to remove or reduce the motion artifacts from the time domain signal. See block 1721 in the embodiments of FIG. 17. In certain implementations, the adaptive filter attempts to predict the heartbeat waveform sensor output data from the motion sensor output data. This determines the motion artifact in the heartbeat waveform sensor output because the motion artifact is the only component common to the two output signals. Stated another way, the adaptive filter subtracts the motion artifact from the heart rate output signal to provide "cleaned" heart rate output data. Examples of adaptive filtering that may be employed include least mean square filtering and recursive least squares filtering. Optionally, the processing logic passes the cleaned heartbeat waveform output signal through the adaptive filter a second time, as explained above. The examiner notes that the processor determines the data integrity based on the motion of the user. When the processor determines that the user is not in stationary mode, then the PPG waveform has poor data integrity and needs modification. The modification is done by filtering out the motion artifact in the waveform.); and the assessment processor is configured to generate a plurality of representations of the modified plurality of PPG waveforms (para. 0180; Identify a heart rate track in the spectral representation of the filtered output data from the heartbeat waveform sensor.).
However, Venkatraman fails to teach creating a spline.
Unser, in the same field of endeavor, teaches create a spline for each of the plurality of PPG waveforms (pages 23-24; determining the B-spline model of a given input signal s(k). Splines are piecewise polynomials with pieces that are smoothly connected together. The joining points of the polynomials are called knots. For a spline of degree n, each segment is a polynomial of degree n, which would suggest that we need n+1 coefficients to describe each piece.).
It would have been obvious to one in the ordinary skill in the art before the effective filling date of the claimed invention to have modified the processing steps of Venkatraman in the view of Ripoll to incorporate the teachings of Unser to provide a spline of the PPG waveforms. This modification will result in a smooth, continuous time representation that suppresses noise and facilitate precise computation of derivative and integral features as disclosed in Unser in pages 24 and 32. Additionally, creating a spline of the modified plurality of PPG waveforms will suppress noise while preserving waveform morphology, enabling reliable computation of derivatives and integrals that maps physiologically meaningful features. By combining spline modeling with motion compensated PPG waveforms, it will generate multiple consistent representations, thereby improving the accuracy and robustness of the blood pressure estimation and other physiological metrics.
Claim 5, Venkatraman teaches the wearable device of claim 4, wherein the assessment processor is further configured to create the modified plurality of PPG waveforms (paras. 0136, 0146-0149, and 0161-0183; Initially, the first activity mode is determined based on the user's motion. Many criteria may be employed to determine the user's activity mode. At a minimum, the devices should receive sensor output suggesting that the user is engaged in some activity and is not sedentary. In a typical implementation, a motion sensor output shows that the user or a user's limb is moving at a reasonable rate. Regardless of how the signal processing logic (e.g., block 1717) determines that the user is participating in the first activity, it begins processing of the buffered output data via the first activity channel. In this channel, the logic assumes that there is a motion artifact that should be removed or reduced before calculating the user's heart rate. The first activity channel may employ an adaptive filter to remove or reduce the motion artifacts from the time domain signal. See block 1721 in the embodiments of FIG. 17. In certain implementations, the adaptive filter attempts to predict the heartbeat waveform sensor output data from the motion sensor output data. This determines the motion artifact in the heartbeat waveform sensor output because the motion artifact is the only component common to the two output signals. Stated another way, the adaptive filter subtracts the motion artifact from the heart rate output signal to provide "cleaned" heart rate output data. Examples of adaptive filtering that may be employed include least mean square filtering and recursive least squares filtering. Optionally, the processing logic passes the cleaned heartbeat waveform output signal through the adaptive filter a second time, as explained above. The examiner notes that the processor determines the data integrity based on the motion of the user. When the processor determines that the user is not in stationary mode, then the PPG waveform has poor data integrity and needs modification. The modification is done by filtering out the motion artifact in the waveform.); and wherein to generate a plurality of representations of the modified plurality of PPG waveforms (para. 0180; Identify a heart rate track in the spectral representation of the filtered output data from the heartbeat waveform sensor.).
However, Venkatraman fails to teach creating a spline and computing a derivative of at least one of the splines.
Unser, in the same field of endeavor, teaches create a spline for each of the plurality of PPG waveforms and computing a derivative of at least one of the splines (pages 23-24; determining the B-spline model of a given input signal s(k). Splines are piecewise polynomials with pieces that are smoothly connected together. The joining points of the polynomials are called knots. For a spline of degree n, each segment is a polynomial of degree n, which would suggest that we need n+1 coefficients to describe each piece.).
It would have been obvious to one in the ordinary skill in the art before the effective filling date of the claimed invention to have modified the processing steps of Venkatraman in the view of Ripoll to incorporate the teachings of Unser to provide a spline of the PPG waveforms and a derivative of at least one of the splines. This modification will result in accurate approximation and precise convergence rates as disclosed in Unser in page 35. Additionally, computing the derivatives of the spline provide slope and acceleration based views of the waveform, thereby yielding physiologically meaningful features such as velocity and acceleration plethysmograms which will improve the accuracy of blood pressure assessment.
Claim 6, Venkatraman teaches the wearable device of claim 4, wherein the assessment processor is further configured to create the modified plurality of PPG waveforms (paras. 0136, 0146-0149, and 0161-0183; Initially, the first activity mode is determined based on the user's motion. Many criteria may be employed to determine the user's activity mode. At a minimum, the devices should receive sensor output suggesting that the user is engaged in some activity and is not sedentary. In a typical implementation, a motion sensor output shows that the user or a user's limb is moving at a reasonable rate. Regardless of how the signal processing logic (e.g., block 1717) determines that the user is participating in the first activity, it begins processing of the buffered output data via the first activity channel. In this channel, the logic assumes that there is a motion artifact that should be removed or reduced before calculating the user's heart rate. The first activity channel may employ an adaptive filter to remove or reduce the motion artifacts from the time domain signal. See block 1721 in the embodiments of FIG. 17. In certain implementations, the adaptive filter attempts to predict the heartbeat waveform sensor output data from the motion sensor output data. This determines the motion artifact in the heartbeat waveform sensor output because the motion artifact is the only component common to the two output signals. Stated another way, the adaptive filter subtracts the motion artifact from the heart rate output signal to provide "cleaned" heart rate output data. Examples of adaptive filtering that may be employed include least mean square filtering and recursive least squares filtering. Optionally, the processing logic passes the cleaned heartbeat waveform output signal through the adaptive filter a second time, as explained above. The examiner notes that the processor determines the data integrity based on the motion of the user. When the processor determines that the user is not in stationary mode, then the PPG waveform has poor data integrity and needs modification. The modification is done by filtering out the motion artifact in the waveform.); and wherein to generate a plurality of representations of the modified plurality of PPG waveforms (para. 0180; Identify a heart rate track in the spectral representation of the filtered output data from the heartbeat waveform sensor.).
However, Venkatraman fails to teach creating a spline and computing an integral of at least one of the splines.
Unser, in the same field of endeavor, teaches create a spline for each of the plurality of PPG waveforms and computing an integral of at least one of the splines (pages 23-24; determining the B-spline model of a given input signal s(k). Splines are piecewise polynomials with pieces that are smoothly connected together. The joining points of the polynomials are called knots. For a spline of degree n, each segment is a polynomial of degree n, which would suggest that we need n+1 coefficients to describe each piece.).
It would have been obvious to one in the ordinary skill in the art before the effective filling date of the claimed invention to have modified the processing steps of Venkatraman in the view of Ripoll to incorporate the teachings of Unser to provide a spline of the PPG waveforms and an integral of at least one of the splines. This modification will provide mathematically exact, stable, and efficient way to obtain cumulative or area based information. Thus, integrals of spline based PPG waveforms provide physiologically meaningful area features, such as systolic to diastolic ratios or total pulse volume, which correlate with blood pressure, thereby enabling robust and accurate blood pressure estimation.
Claim 13, Venkatraman teaches the method of claim 12, further comprising: creating, by the assessment processor the filtered plurality of PPG waveforms (paras. 0136, 0146-0149, and 0161-0183; Initially, the first activity mode is determined based on the user's motion. Many criteria may be employed to determine the user's activity mode. At a minimum, the devices should receive sensor output suggesting that the user is engaged in some activity and is not sedentary. In a typical implementation, a motion sensor output shows that the user or a user's limb is moving at a reasonable rate. Regardless of how the signal processing logic (e.g., block 1717) determines that the user is participating in the first activity, it begins processing of the buffered output data via the first activity channel. In this channel, the logic assumes that there is a motion artifact that should be removed or reduced before calculating the user's heart rate. The first activity channel may employ an adaptive filter to remove or reduce the motion artifacts from the time domain signal. See block 1721 in the embodiments of FIG. 17. In certain implementations, the adaptive filter attempts to predict the heartbeat waveform sensor output data from the motion sensor output data. This determines the motion artifact in the heartbeat waveform sensor output because the motion artifact is the only component common to the two output signals. Stated another way, the adaptive filter subtracts the motion artifact from the heart rate output signal to provide "cleaned" heart rate output data. Examples of adaptive filtering that may be employed include least mean square filtering and recursive least squares filtering. Optionally, the processing logic passes the cleaned heartbeat waveform output signal through the adaptive filter a second time, as explained above. The examiner notes that the processor determines the data integrity based on the motion of the user. When the processor determines that the user is not in stationary mode, then the PPG waveform has poor data integrity and needs modification. The modification is done by filtering out the motion artifact in the waveform.); and generating, by the assessment processor, a plurality of representations of the filtered plurality of PPG waveforms (para. 0180; Identify a heart rate track in the spectral representation of the filtered output data from the heartbeat waveform sensor.).
However, Venkatraman fails to teach creating a spline.
Unser, in the same field of endeavor, teaches create a spline for each of the plurality of PPG waveforms (pages 23-24; determining the B-spline model of a given input signal s(k). Splines are piecewise polynomials with pieces that are smoothly connected together. The joining points of the polynomials are called knots. For a spline of degree n, each segment is a polynomial of degree n, which would suggest that we need n+1 coefficients to describe each piece.).
It would have been obvious to one in the ordinary skill in the art before the effective filling date of the claimed invention to have modified the processing steps of Venkatraman in the view of Ripoll to incorporate the teachings of Unser to provide a spline of the PPG waveforms. This modification will result in a smooth, continuous time representation that suppresses noise and facilitate precise computation of derivative and integral features as disclosed in Unser in pages 24 and 32. Additionally, creating a spline of the modified plurality of PPG waveforms will suppress noise while preserving waveform morphology, enabling reliable computation of derivatives and integrals that maps physiologically meaningful features. By combining spline modeling with motion compensated PPG waveforms, it will generate multiple consistent representations, thereby improving the accuracy and robustness of the blood pressure estimation and other physiological metrics.
Claim 15, Venkatraman teaches the method of claim 14, further comprising creating, by the assessment processor the filtered plurality of PPG waveforms (paras. 0136, 0146-0149, and 0161-0183; Initially, the first activity mode is determined based on the user's motion. Many criteria may be employed to determine the user's activity mode. At a minimum, the devices should receive sensor output suggesting that the user is engaged in some activity and is not sedentary. In a typical implementation, a motion sensor output shows that the user or a user's limb is moving at a reasonable rate. Regardless of how the signal processing logic (e.g., block 1717) determines that the user is participating in the first activity, it begins processing of the buffered output data via the first activity channel. In this channel, the logic assumes that there is a motion artifact that should be removed or reduced before calculating the user's heart rate. The first activity channel may employ an adaptive filter to remove or reduce the motion artifacts from the time domain signal. See block 1721 in the embodiments of FIG. 17. In certain implementations, the adaptive filter attempts to predict the heartbeat waveform sensor output data from the motion sensor output data. This determines the motion artifact in the heartbeat waveform sensor output because the motion artifact is the only component common to the two output signals. Stated another way, the adaptive filter subtracts the motion artifact from the heart rate output signal to provide "cleaned" heart rate output data. Examples of adaptive filtering that may be employed include least mean square filtering and recursive least squares filtering. Optionally, the processing logic passes the cleaned heartbeat waveform output signal through the adaptive filter a second time, as explained above. The examiner notes that the processor determines the data integrity based on the motion of the user. When the processor determines that the user is not in stationary mode, then the PPG waveform has poor data integrity and needs modification. The modification is done by filtering out the motion artifact in the waveform.); and wherein to generate a plurality of representations of the filtered plurality of PPG waveforms (para. 0180; Identify a heart rate track in the spectral representation of the filtered output data from the heartbeat waveform sensor.).
However, Venkatraman fails to teach creating a spline and computing a derivative of at least one of the splines.
Unser, in the same field of endeavor, teaches create a spline for each of the plurality of PPG waveforms and computing a derivative of at least one of the splines (pages 23-24; determining the B-spline model of a given input signal s(k). Splines are piecewise polynomials with pieces that are smoothly connected together. The joining points of the polynomials are called knots. For a spline of degree n, each segment is a polynomial of degree n, which would suggest that we need n+1 coefficients to describe each piece.).
It would have been obvious to one in the ordinary skill in the art before the effective filling date of the claimed invention to have modified the processing steps of Venkatraman in the view of Ripoll to incorporate the teachings of Unser to provide a spline of the PPG waveforms and a derivative of at least one of the splines. This modification will result in accurate approximation and precise convergence rates as disclosed in Unser in page 35. Additionally, computing the derivatives of the spline provide slope and acceleration based views of the waveform, thereby yielding physiologically meaningful features such as velocity and acceleration plethysmograms which will improve the accuracy of blood pressure assessment.
Claim 16, Venkatraman teaches the method of claim 14, further comprising creating, by the assessment processor the filtered plurality of PPG waveforms (paras. 0136, 0146-0149, and 0161-0183; Initially, the first activity mode is determined based on the user's motion. Many criteria may be employed to determine the user's activity mode. At a minimum, the devices should receive sensor output suggesting that the user is engaged in some activity and is not sedentary. In a typical implementation, a motion sensor output shows that the user or a user's limb is moving at a reasonable rate. Regardless of how the signal processing logic (e.g., block 1717) determines that the user is participating in the first activity, it begins processing of the buffered output data via the first activity channel. In this channel, the logic assumes that there is a motion artifact that should be removed or reduced before calculating the user's heart rate. The first activity channel may employ an adaptive filter to remove or reduce the motion artifacts from the time domain signal. See block 1721 in the embodiments of FIG. 17. In certain implementations, the adaptive filter attempts to predict the heartbeat waveform sensor output data from the motion sensor output data. This determines the motion artifact in the heartbeat waveform sensor output because the motion artifact is the only component common to the two output signals. Stated another way, the adaptive filter subtracts the motion artifact from the heart rate output signal to provide "cleaned" heart rate output data. Examples of adaptive filtering that may be employed include least mean square filtering and recursive least squares filtering. Optionally, the processing logic passes the cleaned heartbeat waveform output signal through the adaptive filter a second time, as explained above. The examiner notes that the processor determines the data integrity based on the motion of the user. When the processor determines that the user is not in stationary mode, then the PPG waveform has poor data integrity and needs modification. The modification is done by filtering out the motion artifact in the waveform.); and wherein to generate a plurality of representations of the filtered plurality of PPG waveforms (para. 0180; Identify a heart rate track in the spectral representation of the filtered output data from the heartbeat waveform sensor.).
However, Venkatraman fails to teach creating a spline and computing an integral of at least one of the splines.
Unser, in the same field of endeavor, teaches create a spline for each of the plurality of PPG waveforms and computing an integral of at least one of the splines (pages 23-24; determining the B-spline model of a given input signal s(k). Splines are piecewise polynomials with pieces that are smoothly connected together. The joining points of the polynomials are called knots. For a spline of degree n, each segment is a polynomial of degree n, which would suggest that we need n+1 coefficients to describe each piece.).
It would have been obvious to one in the ordinary skill in the art before the effective filling date of the claimed invention to have modified the processing steps of Venkatraman in the view of Ripoll to incorporate the teachings of Unser to provide a spline of the PPG waveforms and an integral of at least one of the splines. This modification will provide mathematically exact, stable, and efficient way to obtain cumulative or area based information. Thus, integrals of spline based PPG waveforms provide physiologically meaningful area features, such as systolic to diastolic ratios or total pulse volume, which correlate with blood pressure, thereby enabling robust and accurate blood pressure estimation.
Response to Arguments
Applicant's arguments filed 12/29/2025 have been fully considered but they are not persuasive. The applicant argues that Venkatraman fails to teach “determine a data integrity of the plurality of PPG waveforms output by the PPG sensor representative of subject cadence; modify at least one of the plurality of PPG waveforms output by the PPG sensor based on the determined data integrity representative of subject cadence exceeding at least one threshold”. The examiner respectfully disagrees. Venkatraman disclose that the PPG data confidence/integrity is determined based on whether the data comprise a motion artifact or not based motion data received from motion sensors. The motion data represents the patient cadence. The processor starts by receiving data from the PPG sensor and the motion sensor and apply an activity discrimination logic which discriminate, based on data from the sensors, if the data comprises motion artifact or not. If the data comprise motion artifact (cadence of the patient is greater than a threshold) then it assigns it activity channel 1 and if the data has no motion artifact, good integrity, then it assigns it to stationary channel. The data that it is determined to have poor integrity (has motion artifact) goes through filtering process to compensate for the motion artifact present in the data. Thus, the system generates modified/filtered data that has high data integrity for further calculation of heart rates [see figure 17, paras. 0132 and 0161-0183]. Thus, Venkatraman teaches determine a data integrity of the plurality of PPG waveforms output by the PPG sensor representative of subject cadence; modify at least one of the plurality of PPG waveforms output by the PPG sensor based on the determined data integrity representative of subject cadence exceeding at least one threshold.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ZAINAB M ALDARRAJI whose telephone number is (571)272-8726. The examiner can normally be reached Monday-Thursday7AM-5PM EST.
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, Carey Michael can be reached at (571) 270-7235. 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.
/ZAINAB MOHAMMED ALDARRAJI/Patent Examiner, Art Unit 3797
/MICHAEL J CAREY/Supervisory Patent Examiner, Art Unit 3795