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 .
Status of Application, Amendments and Claims
Claims 1-8 and 10-18 are pending and claims 1, 4, 11, 14 and 16-18 are amended.
Response to Arguments
Applicant's arguments filed 20 August 2026 with respect to claims 1, 4 and 8 have been fully considered but are not persuasive.
The applicant argues with respect to claim 1:
(1) The claims require the active reinforcement algorithm to be trained to maximize detection of oxygen-desaturation events while minimizing the time for which the pulse oximetry sensor is active. This results in an adaptive physiological control loop that is not taught Lee. The examiner respectfully disagrees with this rationale as the rejection is based on the combination of Lee, Chiu, Sayadi and Sundstrom. The combination of references renders obvious the use of active reinforcement machine learning algorithm to control the operation of a pulse oximeter in response to measure respiration parameters. This combination reads on the broadest reasonable interpretation of the claims. Especially when considering the claims and disclosure do not include any specific training steps for implementing the active reinforcement machine learning algorithm.
(2) The applicant argues Chiu activates its PPG sensor when a specific even is detected according to vital signs data but not according to breathing characteristics indicative of the time since a most recent snore. The examiner does not find this line of argument persuasive. Chiu is not relied upon for using breathing characteristics for detection of snore timing. The primary reference Lee discloses this aspect of the claim. Rather, Chiu is relied upon for teaching the automatic control of the pulse oximetry sensor based on predetermined times since the detection of a primary and secondary snoring in order to conserve energy by only activating the pulse oximetry sensor when necessary.
(3) The applicant similarly argues Sayadi and Sundstrom do not cure the deficiencies of Lee and Chiu. However, as explained above Lee already discloses the use of breathing characteristics to detect a most recent snore. Sayadi is applied to teach the use of user-specific training data for a machine learning algorithm so that snore categorization may be tailored to specific users and more accurately detect and categorize snore events by the user. Sundstrom is applied to teach the known benefits of using an active reinforcement machine learning algorithm.
As acknowledged by the applicant, Lee relies on the detection of snoring to control monitoring. Each of Chiu, Sayadi and Sundstrom suggest In response to applicant's arguments against the references individually, one cannot show nonobviousness by attacking references individually where the rejections are based on combinations of references. See In re Keller, 642 F.2d 413, 208 USPQ 871 (CCPA 1981); In re Merck & Co., 800 F.2d 1091, 231 USPQ 375 (Fed. Cir. 1986).
The applicant argues with respect to claim 4:
The applicant argues that Ziegel does not teach using an active reinforcement algorithm to identify pre-apnea snoring from breathing characteristic and controlling a pulse oximeter according to the time since a most recent snore. Similar to the arguments presented with respect to claim 1 above, the rejection of claim 4 is based on a combination of references. Ziegel is used only to teach diagnosing obstructive sleep apnea comprising determining if snores are a group of snores if a time delay between temporally adjacent snores is less than or equal to a predetermined time. The active reinforcement algorithm is rendered obvious by the previously applied teaching of Sundstrom.
The applicant argues with respect to claim 8:
The applicant argues that Beckman discloses a generic reason to use a reward function to reduce cost in a robot-control context and does not motivate a reward function in an oxygen-desaturation-detection accuracy and minimization of pulse oximetry sensor activation time context. Similar to the arguments presented with respect to claim 1 above, the rejection of claim 8 is based on a combination of references. The references applied to claims 1 and 6 provide the pulse oximetry context. The applicant’s disclosure does not go into any specific detail about the reward function. Thus the applicant’s disclosure is directed to a generic reward function.
Applicant’s arguments, see Applicant’s Remarks, filed 20 August 2026, with respect to claim 10 have been fully considered and are persuasive. The rejection of claim 10 under 35 U.S.C. 103 has been withdrawn.
Claim Objections
Claim 10 is objected to because of the following informalities: “the predetermined threshold” recited in line 10 should be “the predetermined running variance threshold.” Appropriate correction is required.
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 1-8 and 10-18 are 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 applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claims 1 and 14 recites the limitation "a second predetermined time period" in the second to last line of the claims. The claims do not introduce a first predetermined time period. It is unclear if the claim should include a first predetermined time period and how that first predetermined time period would relate to the second predetermined time period.
All remaining claims are rejected for depending from an indefinite claim.
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 text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action.
Claims 1-3, 5-7 and 11-15 are rejected under 35 U.S.C. 103 as being unpatentable over Lee (US 2017/0087050) in view of Chiu et al. (US 2023/0051939) in view of Sayadi et al. (US 2019/0201268) and further in view of Sundstrom et al. (US 2020/0401944).
[Claims 1, 13-15] Lee discloses a system, a processing system (processing circuit assembly, #15), a method and computer program product, for controlling a pulse oximetry sensor (composite sensor, #12, including photodiodes, #31, of a pulse oximeter) [pars. 0033, 0035], the processing system being configured to:
obtain, from a detector (vibration sensor, #32) different to the pulse oximetry sensor, breathing data responsive to a subject's breathing (vibration of an airway inlet) [pars. 0006, 0018];
process the breathing data to determine one or more breathing characteristics (frequencies are derived from the measured vibration) [par. 0018]; and
control the pulse oximetry sensor responsive to the one or more breathing characteristics by:
processing the one or more breathing characteristics to determine whether the subject is snoring (thresholds of measured vibrations indicate primary or secondary snoring) [par. 0018]; and, in response to a determination that the subject is snoring:
processing the one or more breathing characteristics to identify a type of snoring (primary or secondary snoring) [par. 0018]; and
controlling the pulse oximetry sensor according to the identified type of snoring (means to turn on either only the vibration sensor for subjects with primary snoring only, or both the vibration sensor and the pulse oximeter for subjects having primary snoring and lung disease or secondary snoring) [pars. 0022-0023].
Lee implies the processing system controls the pulse oximetry sensor but the “means” for selectively turning on the pulse oximeter is not explicitly identified as the processing system. Lee further does not disclose processing the one or more breathing characteristics to determine a time since a most recent snore; and in response to a determination that the time since the most recent snore exceeds a second predetermined time period, controlling the pulse oximetry sensor to stop acquiring pulse oximetry data.
Chiu discloses an analogous diagnostic processing system for controlling a pulse oximetry sensor (PPG sensor, #11) wherein an PPG controller (#14) activates (turns “ON”) the pulse oximetry sensor in response to abnormal events detected during sleep (e.g. apnea) and deactivates (turns “OFF”) the pulse oximetry sensor when the abnormal event is no longer detected for a period of time in order to save power [pars. 0027-0028; Fig. 5].
PNG
media_image1.png
536
762
media_image1.png
Greyscale
It would have been obvious to one having ordinary skill in the art before the effective filing date to configured the processing system of Lee to automatically control the pulse oximetry sensor based on predetermined times since the detection of abnormal sleeping events (such a primary and secondary snoring) as suggested by Chiu in order to conserve energy by only activating the pulse oximetry sensor when necessary [Lee: par. 0023].
Lee in view of Chiu does not explicitly disclose the processing system is configured to identify a type of snoring based on an active reinforcement machine learning algorithm.
Sayadi discloses a snore detection system using user-specific training data for a machine learning algorithm so that snore categorization may be tailored to specific users and more accurately detect and categorize snore events by the user [par. 0007].
It would have been obvious to one having ordinary skill in the art before the effective filing date to configure the processing system rendered obvious by Lee in view of Chiu to use user-specific training data from a specific user and a machine learning algorithm in order to more accurately detect and categorize snore events by the user.
Lee in view of Chiu and Sayadi discloses using machine learning but does not specify an active reinforcement learning algorithm.
Sundstrom discloses active reinforcement learning is a known concept in machine learning to optimize decision making of the machine learning [par. 0059].
It would have been obvious to one having ordinary skill in the art before the effective filing date to use an active reinforcement learning algorithm in order to improve decision making and further optimize the automated control of the sensor based on detect snoring.
[Claim 2] Lee discloses the detector is a vibration sensor and Chiu specifically discloses an accelerometer which measures vibration.
[Claim 3] Lee discloses the processing system is configured to, in response to the identified type of snoring being a first type of snoring that occurs during hypopnea: control the pulse oximetry sensor to start acquiring pulse oximetry data (the pulse oximetry data is collected at the time snoring starts which would include snoring caused by hypopnea) [par. 0022].
[Claim 5] Lee discloses the processing system is configured to, in response to the identified type of snoring being a third type of snoring that occurs after an obstruction is resolved: control the pulse oximetry sensor to start acquiring pulse oximetry data (the pulse oximetry data is collected at the time snoring starts which would include snoring of a third type that occurs after obstruction is resolved) [par. 0022].
[Claim 6] Lee as modified does not explicitly disclose the processing system is configured to identity a type of snoring based on subject-specific data obtained during at least one previous sleep session for the subject.
Sayadi discloses a snore detection system using user-specific training data for a machine learning algorithm so that snore categorization may be tailored to specific users and more accurately detect and categorize snore events by the user [par. 0007].
It would have been obvious to one having ordinary skill in the art before the effective filing date to configure the processing system rendered obvious by Lee in view of Chiu to use user-specific training data from a specific user and a machine learning algorithm in order to more accurately detect and categorize snore events by the user.
[Claim 7] Based on the combination of references presented in claim 6, it would have been obvious to one having ordinary skill in the art before the effective filing date to ensure the subject-specific data comprises wherein the subject-specific data comprises features of one or more breathing characteristics for the subject for each of a plurality of types of snoring, wherein the plurality of types of snoring comprise one or more of: a first type of snoring that the subject exhibits during hypopnea events; a second type of snoring that the subject exhibits before an obstructive apnea event; a third type of snoring that the subject exhibits after an obstruction is resolved; and/or a fourth type of snoring that the subject exhibits during regular breathing. Doing so would ensure the most accurate detection and categorization of all snoring events.
[Claim 11] Lee as modified renders obvious wherein the first predetermined time period is determined based on the subject-specific data obtained during at least on previous sleep session for the subject.
Sayadi discloses a snore detection system using user-specific training data for a machine learning algorithm so that snore categorization may be tailored to specific users and more accurately detect and categorize snore events by the user [par. 0007].
Chiu discloses a vital signs detector (#13) for estimating the respiration rate of a user wherein the respiration rate is sent to the PPG controller (#14) and when the respiration rate is not within normal range, the PPG controller detects a specific event is happening and generates a control signal to activate the PPG sensor [par. 0035].
It would have been obvious to one having ordinary skill in the art before the effective filing date to configure the processing system rendered obvious by modified Lee to use user-specific training data from a specific user and a machine learning algorithm in order to more accurately detect and categorize snore events by the user, as taught by Sayadi, in combination with detected respiration rate, as taught by Chiu, in order to more accurately determine the timing of the occurrence of snoring events in order to determine the first predetermined time period.
[Claim 12] Lee discloses the processing system is configured to control the pulse oximetry sensor by: obtaining pulse oximetry data from the pulse oximetry sensor; processing the pulse oximetry data to determine a measure of oxygen saturation; and controlling the pulse oximetry sensor to stop acquiring pulse oximetry data in response to a determination that the measure of oxygen saturation exceeds an oxygen saturation threshold throughout a third predetermined time period (the reflectance pulse oximeter is configured to start checking on the tissue oxygen saturation at the time snoring starts and then to continue checking at a regular interval, for an example, every five minutes, until the tissue oxygen saturation improves above 92%) [par. 0022].
Claims 4, 16 and 17-18 are rejected under 35 U.S.C. 103 as being unpatentable over Lee (US 2017/0087050) in view of Chiu et al. (US 2023/0051939), Sayadi et al. (US 2019/0201268) and Sundstrom et al. (US 2020/0401944), as applied to claims 1 and 14 respectively above, and further in view of Zigel et al. (US 2013/0184601).
[Claim 4, 16, 18] Lee in view of Chiu renders obvious controlling the pulse oximeter based on breathing date indicative of snoring type but does not disclose in response to the identified type of snoring being a second type of snoring that occurs before an obstructive apnea event: process the one or more breathing characteristics to determine a time since a most recent snore; and control the pulse oximetry sensor to start acquiring pulse oximetry data in response to a determination that the time since a most recent snore exceeds a first predetermined time period.
Zigel discloses a device and method for diagnosing obstructive sleep apnea [abstract] comprising determining if snores are a group of snores if a time delay between temporally adjacent snores is less than or equal to a predetermined time [par. 0013; claims 1-3].
It would have been obvious to one having ordinary skill in the art before the effective filing date to configure the processing system to use a time delay threshold to group snores as taught by Zigel in order to determine if there is a gap in time large enough to constitute a cessation in breathing since obstructive sleep apnea is determined based on periods of cessation in breathing. After using the first predetermined time period to identify the presence of sleep apnea, the sconed predetermined time would be used to determine a time since the last apnea snore. The first predetermined time period and the second predetermined time period are different as they measure different properties – time since a last snore and a time delay threshold to group snores.
[Claim 17] Lee as modified renders obvious wherein the first predetermined time period is determined based on the subject-specific data obtained during at least on previous sleep session for the subject.
Sayadi discloses a snore detection system using user-specific training data for a machine learning algorithm so that snore categorization may be tailored to specific users and more accurately detect and categorize snore events by the user [par. 0007].
Chiu discloses a vital signs detector (#13) for estimating the respiration rate of a user wherein the respiration rate is sent to the PPG controller (#14) and when the respiration rate is not within normal range, the PPG controller detects a specific event is happening and generates a control signal to activate the PPG sensor [par. 0035].
It would have been obvious to one having ordinary skill in the art before the effective filing date to configure the processing system rendered obvious by modified Lee to use user-specific training data from a specific user and a machine learning algorithm in order to more accurately detect and categorize snore events by the user, as taught by Sayadi, in combination with detected respiration rate, as taught by Chiu, in order to more accurately determine the timing of the occurrence of snoring events in order to determine the first predetermined time period.
Claim 8 is rejected under 35 U.S.C. 103 as being unpatentable over Lee (US 2017/0087050) in view of Chiu et al. (US 2023/0051939), Sayadi et al. (US 2019/0201268) and Sundstrom et al. (US 2020/0401944), as applied to claim 1 above, and further in view of Beckman et al. (US 10,800,040).
[Claim 8] The previously applied references render obvious the use of active reinforcement learning algorithms to measure oxygen desaturation events (snoring) to minimize the on-time of pulse oximetry sensors but do not indicate that the active reinforcement learning algorithm utilizes a reward function based on both the accuracy of oxygen desaturation event detection and the minimization of pulse oximetry sensor activation time.
Beckman discloses reinforcement learning is an area of machine learning that seeks to learn how to make decisions in order to maximize rewards or minimize costs over a period of time. A reinforcement learning system can be used to model the reward function of the task, which can be considered as a model of the goal of the tasks and may be expressed as weighted factors that influence success at task performance [col 7, lines 47-53].
It would have been obvious to one having ordinary skill in the art before the effective filing date to incorporate a reward function into the active reinforcement learning algorithm as taught by Beckman in order to minimize costs (i.e. sensor power use) over a period of measurement time.
Allowable Subject Matter
Claim 10 would be allowable if rewritten to overcome the rejection(s) under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), 2nd paragraph, set forth in this Office action and to include all of the limitations of the base claim and any intervening claims. Lee discloses monitoring the frequency of vibration of the airway inlet to determine primary and secondary snoring in order to active the pulse oximeter but does not disclose measuring a running variance as argued on page 12 of Applicant’s Remarked filed 20 August 2026.
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 NATHAN J JENNESS whose telephone number is (571)270-5055. The examiner can normally be reached M-F 8:00-5:00 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, Jonathan Moffat can be reached at 571-272-4390. 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.
/NATHAN J JENNESS/Supervisory Patent Examiner, Art Unit 3733 9 October 2026
