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 .
Priority
Priority FOR EP 23161140.1 03/10/2023 claims are acknowledge.
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 08/20/2025. The submission is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Claim Rejections - 35 USC § 112(b)
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.
Claim 8-17 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.
Claim 8 receives movement/position measurements and a risk score, but later requires detection based on undefined “physiological data.” Generating a antecedent basic effect. Claim 8 is therefore indefinite under §112(b), and claims 9–17 inherit the defect.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1–20 are rejected under 35 U.S.C. 101 because they recite the mental process identified below without additional elements that integrate the exception into a practical application or provide an inventive concept amounting to significantly more.
Step 1: Statutory Categories
Claims 1–6, 8–17, 19, and 20 require component-defined apparatuses or systems and therefore fall within the machine category. Claims 7 and 18 require ordered acts and therefore fall within the process category. Each claim under rejection satisfies Step 1 under MPEP § 2106.03.
Step 2A, Prong One
Independent Claims Analysis:
Prong One asks whether the claims recite a judicial exception.
Representative:
Claim 19:
A system for detecting the onset of a somnambulism event of a user during a sleep period of the user, the system comprising a first apparatus for determining a somnambulism event risk level score of a user, the first apparatus comprising:
interface circuitry configured to receive, before the sleep period, physiological data of the user generated before the sleep period and
processing circuitry configured to determine a risk level score representing a likelihood that a somnambulism event of the user will occur during the sleep period based on the physiological data; and
a second apparatus for detecting the onset of a somnambulism event of the user during the sleep period, the second apparatus comprising:
interface circuitry configured to receive movement and/or position measurements of the user generated during the sleep period and to receive the risk level score from the first apparatus; and
processing circuitry configured to detect the onset of a somnambulism event of the user based on the movement and/or position measurements and the risk level score.
Non-bold language identifies the judicial exception; bold language identifies the additional elements preserved for later evaluation.
Claim 19 recite non-bold above limitations 1-6 mental process. Limitations 1–6 collectively recite observing or recording specified information and evaluating it to estimate somnambulism risk or identify event onset. In limitations 2 and 5, the non-bold receipt language states only the informational result obtaining pre-sleep physiological information, sleep-period movement or position information, and a risk score without requiring a particular signal-acquisition or reception mechanism. In limitations 3 and 6, determine a risk level score . . . based on the physiological data and detect the onset . . . based on the movement and/or position measurements and the risk level score are evaluations and judgments. The limiting onset functions in limitations 1 and 4 state the same mental-process result. For example a trained clinician or sleep technician using written sensor readings on a patient chart could review pre-sleep physiological readings and assign a risk category; during the specified sleep period, the technician could review movement or position readings, consult the risk category, and judge whether onset has occurred. Such high-level collection and evaluation of information falls within the mental-process grouping observations, evaluations, and judgments.
Claims 1 and 7 recite only the risk-assessment portion; claim 7 additionally requires a trained machine-learning model. Claims 8 and 18 recite only the onset-detection portion; claim 18 states it as a method without circuitry. Claim 20 reverses claim 19’s order by first identifying onset and then using awake and labeled prior-sleep information to assess future risk. These differences change the claim form, input information, or additional elements, but not the judicial exception: each claim still recites receiving information and making a result-level risk or onset judgment without requiring evaluation mechanism that a person could not practically perform.
Accordingly, claims 1, 7, 8, 18, 19, and 20 recite the same mental-process abstract idea. Their distinct additional elements-circuitry, machine-learning implementation, or inter-apparatus data relationships-remain preserved for separate evaluation at Prong Two.
Dependent Claims Analysis:
Claims 3, 4, 9–11, 13, and 14 further specify the information or evaluation. The adjustable observation period, pattern comparison, customized thresholds, sleep-stage classification, event-stage classification, and use of further pre-sleep or sleep-period data are observations, comparisons, or judgments practically performed using written information. These claims narrow but retain the mental-process exception.
Claims 2 and 15 add wearable-device connectivity as an additional element; their listed physiological-data types remain informational content. Claims 5, 6, and 17 add machine-learning models and training or updating operations; their physiological data, scores, event notifications, sleep datasets, and event labels remain informational content used in the evaluation. Claim 16 adds memory, saving, export, and an external-device endpoint; the sleep-dataset content remains part of the exception.
Claims 1–20 therefore recite a mental-process abstract idea and proceed to Prong Two.
Step 2A, Prong Two
At Step 2A Prong Two, the additional elements are considered individually and as an ordered combination to determine whether they integrate the judicial exception into a practical application.
Additional elements evaluation:
Claim 1, an apparatus and interface and processing circuitry; for claim 7, a trained machine-learning model; for claim 8, an apparatus and interface and processing circuitry; for claim 18, no technical element beyond method form; for claim 19, a system, first and second apparatuses, interface and processing circuitry, and the first-apparatus risk-score source relationship; and for claim 20, the corresponding system, circuitry, apparatuses, and first-apparatus labeled-data source relationship
Independent-claim implementation. The apparatuses, interface circuitry, processing circuitry, and trained-model implementation merely supply the tools and operating environment used to perform the recited observations and evaluations. The claims specify no sensor structure, communication protocol, processor architecture, model operation, data transformation, or other technical rule that changes how the hardware or software operates. They therefore amount to applying the mental process through functionally recited computer and monitoring components rather than improving those components. MPEP §§ 2106.05(a), (f), and (h).
Claims 19 and 20 additionally recite machine-source relationships. Claim 19 requires interface circuitry “to receive the risk level score from the first apparatus,” while claim 20 requires interface circuitry to receive “from the first apparatus, labeled sleeping physiological data.” These limitations identify only the source and destination of information used in the next evaluative operation. Neither claim specifies a communication protocol, data-transmission technique, modified interface operation, or transformation of the transferred information. Accordingly, the relationships merely route information required as input to the subsequent evaluation and constitute insignificant data-transfer activity under MPEP § 2106.05(g), rather than a technological application that meaningfully limits the recited abstract idea.
Wearable devices-claims 2 and 15. The wearable device and communicative connection identify the source of the physiological observations. Neither claim changes operation of the wearable, improves a sensor, or recites a particular measurement technique. The connection therefore performs necessary data gathering and does not impose a meaningful technological limit on the inherited evaluation.
Machine learning-claims 5–7 and 17. Claim 7 uses a trained model to determine risk; claim 5 periodically trains and updates a model; claim 6 updates training with another physiological-data, score, and event-label iteration; and claim 17 trains from labeled sleep datasets. Under their full BRI, the claims state model use or a desired training result rather than the components or steps that improve model operation.
Claim 12 recites the desired result—auditory or vibratory feedback when onset is detected or improvement in detector or feedback-device operation. The additional elements therefore merely instruct application of the exception and do not integrate it into a practical application. MPEP §§ 2106.05(a), (f).
Memory and export-claim 16. The memory, saving, export, and external-device endpoint preserve or forward acquired information. Claim 16 recites no storage structure, retrieval operation, or technical data organization that changes memory or computer operation. These limitations therefore constitute insignificant post-solution data handling under MPEP § 2106.05(g).
Considered as ordered combinations, none claims a technical interaction that changes operation of the recited hardware, software, sensing, communication, treatment, or storage technology. Accordingly, claims 1–20 do not integrate the mental-process exception into a practical application.
Step 2B
Step 2B asks whether the claim as a whole amounts to “significantly more” than the exception itself.
Independent hardware and transfers. The specification permits the processing circuitry to be a dedicated or shared processor, multiple processors, DSP, ASIC, or FPGA coupled to ordinary ROM or RAM, and permits the detection circuitry to use the same alternatives. Spec., pp. 7 and 18. The current MPEP § 2106.05(d)(II) identifies high-level receiving and transmitting data, repetitive computer calculations, electronic recordkeeping, and storing or retrieving information as well-understood, routine, and conventional computer functions, citing, inter alia, TLI Communications, Electric Power Group, buySAFE, Intellectual Ventures v. Symantec, Flook, Bancorp, Alice, and Versata. Used only to apply the identified judicial exception in prong one.
Claims 1 and 8 use the circuitry only for those recited receipt and processing functions. Claims 19 and 20 add no transmission protocol or changed communications operation to the risk-score and labeled-data transfers. Their two-apparatus placement merely follows the logical sequence of the underlying evaluations: score then detect in claim 19, and detect and label then evaluate later risk in claim 20. Claim 18 has no additional element. These elements therefore do not add significantly more, individually or in their claimed relationships.
Machine learning-claims 5–7 and 17. The specification presents RNNs, CNNs, temporal learning, attention mechanisms, continual or online learning, transfer learning, and standard optimizers as selectable implementations. Spec., pp. 12–16 and 24–25. The claims select none of the operations that allegedly improve model training or operation. Claim 7 requires result-level inference, claim 5 periodic updating, claim 6 another labeled update iteration, and claim 17 training using labeled sleep datasets. Applying machine learning to perform an abstract evaluation, without claiming an improvement to machine-learning technology, supplies no inventive concept. Recentive Analytics, Inc. v. Fox Corp., 134 F.4th 1205, 1212–16 (Fed. Cir. 2025).
Wearable devices-claims 2 and 15. The specification identifies the wearable source through existing alternatives such as a smartwatch, fingertip sensor, photoplethysmogram, accelerometer, and gyroscope. Spec., pp. 9–10. Claims 2 and 15 require no new sensor, signal-acquisition technique, or wearable operation. The wearable and connection therefore perform only the ordinary acquisition and receipt functions recognized in Electric Power Group and TLI Communications and do not supply an inventive concept.
Claim 12. The same result-only elements add no inventive concept because they specify what notification occurs, but no technological means or improved operation for producing it. As in FairWarning, where event-triggered processor notification supplied no “something more,” the elements individually and in their one-way arrangement do not add significantly more. MPEP §§ 2106.05(a), (f).
Memory and export-claim 16. Saving information in memory and exporting it to an external device are high-level electronic recordkeeping, storage, and transmission functions recognized in MPEP § 2106.05(d)(II), Alice, Versata, TLI Communications, and buySAFE. Because claim 16 adds no storage structure, retrieval mechanism, or technical data organization, it does not supply an inventive concept.
Claims 3, 4, 9–11, 13, and 14 add only the mental-process content identified at Prong One and therefore contain no new additional element capable of providing significantly more. Under BASCOM, the additional elements have also been considered in their ordered combinations. Here they are placed only in the sequence demanded by the informational process and provide no nonconventional component placement, relationship, or technical interaction. Accordingly, claims 1–20 do not amount to significantly more than the recited judicial exception.
Claims 1–20 are therefore rejected under 35 U.S.C. § 101 as directed to a judicial exception without additional elements amounting to significantly more.
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.
Claims 1, 2, 4, and 7–19 are rejected under 35 U.S.C. 103 as being unpatentable over Sathaye et al. (US 2023/0238112 A1) in view of Arrington et al. (US 2019/0223781 A1).
Claim 19.
Sathaye teaches,
A system for detecting
period of the user, the system comprising a first apparatus for determining a : (Sathaye discloses predictive data analysis system 101 and computing entity 106 performing pre-sleep and in-sleep parasomnia-likelihood functions, with each functionality performable by any number of computing entities. Sathaye [0064]-[0065], [0091]; Figs. 1, 2, and 4.)
interface circuitry configured to receive, before the sleep period, physiological data of the user generated before the sleep period (Sathaye discloses data provided by sensor devices, communications interface 220 for communicating data, and acquisition of pre-sleep individual monitoring data for a defined pre-sleep window, including ECG and/or EEG data. Sathaye [0065], [0073], [0081], [0092]; Figs. 1, 2, and 4.)
processing circuitry configured to determine a risk level score representing a likelihood that a ; (Sathaye discloses a processor-and-memory apparatus processing pre-sleep monitoring data with a machine-learning model to generate a pre-sleep parasomnia-episode likelihood score representing the likelihood of one or more parasomnia episodes during an upcoming sleep window. Sathaye [0005], [0030], [0093], [0124]-[0125]; Figs. 4-5.) and
: (Sathaye discloses a functionally distinct in-sleep prediction role, implemented by computing entity 106 and performable by any number of computing entities, for an ongoing sleep window. Sathaye [0031], [0091], [0128], [0150]; Figs. 4 and 7.)
interface circuitry configured to receive movement and/or position measurements of the user generated during the sleep period and to receive the risk level score ; (Sathaye discloses communications interface 220; sequential generation of a pre-sleep score and acquisition of in-sleep monitoring data; in-sleep movement measurement sequence 714; an in-sleep static vector that may contain the pre-sleep score; and merger of the static vector and movement representation into the in-sleep model input. Sathaye [0073], [0081], [0128]-[0129], [0139]-[0140], [0149]; Figs. 2, 4, and 7.)
and
processing circuitry configured to detect the . (Sathaye discloses processing that combines the in-sleep static vector, ECG, EEG, movement, audio, facial, and thermal representations to generate a likelihood that the monitored individual is currently experiencing a parasomnia episode during the ongoing sleep window. Sathaye [0031], [0149]-[0150]; Fig. 7.)
Primary Sathaye discloses a pre-sleep likelihood vector having a value for each parasomnia episode type for an upcoming sleep window, but does not disclose somnambulism as the selected type (Sathaye [0030]). Secondary Arrington expressly defines “sleepwalking” as “somnambulism” and identifies sleepwalking as a target of parasomnia detection and intervention (Arrington [abstract],[0002], [0007], [0015]). A POSITA would have configured one of Sathaye’s disclosed type-specific outputs for sleepwalking because Arrington identifies that condition for detection and Sathaye seeks faster real-time detection and intervention to reduce harm (Sathaye [0023]). This preserves Sathaye’s pre-sleep model and inputs. Because Sathaye already produces separate type-indexed values, the modification predictably would produce a somnambulism-specific risk score, and a POSITA reasonably would have expected implementation success.
Primary Sathaye’s in-sleep model combines ongoing monitoring, including movement, with its pre-sleep score to determine a current parasomnia likelihood, but does not disclose the transition constituting onset (Sathaye [0031], [0139]-[0140], [0149]-[0150]; Fig. 7). Secondary Arrington expressly states that sensors may identify parasomnia onset and discloses repeated collection and threshold-based episode determinations (Arrington [0002], [0044]-[0045], [0052]-[0053], [0056]; Fig. 1). A POSITA would have modified the output handling of the sleepwalking-configured Sathaye model to mark its first negative-to-positive determination as onset because Sathaye treats real-time harm reduction as time-critical and Arrington identifies the lack of immediate assistance during an episode as a problem (Sathaye [0023]; Arrington [0013]). The change preserves Sathaye’s movement and pre-sleep-score inputs. Repeated threshold determinations make transition marking predictable, and success reasonably would have been expected without replacing the model.
Primary Sathaye permits its functions to be distributed among multiple computing entities and its in-sleep model uses a pre-sleep score, but does not disclose the claimed first-apparatus-to-second-apparatus allocation and score receipt (Sathaye [0031], [0091], [0149]-[0150]). Secondary Arrington discloses wearable sleep-management devices with movement sensing, networked clients and servers, and remote-computing results delivered to a resource-limited client (Arrington [0025], [0047], [0063], [0070], [0101]-[0104]; Fig. 9). A POSITA would have modified Sathaye by placing its pre-sleep predictor on a server as the first apparatus, retaining the score-using in-sleep detector on a sensor-local wearable as the second apparatus, and transmitting the numerical score over Arrington’s network because this uses remote resources while keeping time-critical sensing local, consistent with Sathaye [0023]. Sathaye’s multi-entity option, existing score input, and Arrington’s result-delivery channel make the allocation and transfer predictable without changing either model’s function.
Claim 1.
Sathaye teaches,
An apparatus for determining a (Sathaye discloses predictive data analysis system 101 and computing entity 106 performing pre-sleep prediction of an upcoming-window parasomnia likelihood score. Sathaye [0030], [0064]-[0065], [0091]-[0093]; Figs. 1, 2, 4, and 5.)
interface circuitry configured to receive, before a sleep period of the user, physiological data of the user generated before the sleep period, and (Sathaye discloses communications interface 220 and acquisition of pre-sleep individual monitoring data during a defined period before an expected sleep window. The data includes pre-sleep ECG and EEG sequences supplied by sensors. Sathaye [0065], [0073], [0092]; Figs. 1, 2, and 4.)
processing circuitry configured to determine a risk level score representing a likelihood that a somnambulism event of the user will occur during the sleep period based on the physiological data generated before the sleep period. (Sathaye discloses processing the pre-sleep physiological monitoring data to produce a score representing the likelihood of a parasomnia episode during the following sleep window. Its output may contain a separate value for each parasomnia-episode type. Sathaye [0030], [0093], [0124]-[0125]; Fig. 5.)
Sathaye supplies the apparatus, claimed pre-sleep physiological input, and future-sleep risk score, but Sathaye does not expressly identify somnambulism as the selected parasomnia type. Secondary Arrington expressly identifies sleepwalking as somnambulism and teaches monitoring physiological indicators and movement to identify sleepwalking and other parasomnia episodes. Arrington [0002], [0007], [0015], [0044], [0069]-[0070].
A person of ordinary skill in the art would have added a sleepwalking-designated value to Sathaye's disclosed type-indexed output because Arrington identifies sleepwalking as a known parasomnia suitable for sensor-based detection and intervention. The skilled artisan would have used Sathaye's disclosed training-and-deployment techniques with class-specific examples for the known sleepwalking type. Sathaye [0030], [0172]-[0174]; Arrington [0044]-[0047], [0052]-[0053], [0069]-[0070]. Sathaye already provides the pre-sleep inputs and a type-indexed likelihood architecture, so generating a sleepwalking-designated numerical output would have been a predictable configuration.
Claim 7.
Sathaye teaches,
A method for determining a (Sathaye discloses predictive data analysis for parasomnia-episode management, including pre-sleep prediction of an upcoming-window parasomnia likelihood score. Sathaye [0030], [0064]-[0065], [0091]-[0093]; Figs. 1, 4, and 5.)
receiving, before a sleep period of the user, physiological data of the user generated before the sleep period, and (Sathaye discloses obtaining pre-sleep individual monitoring data during a defined period before an expected sleep window. The data includes pre-sleep ECG and EEG sequences and may be supplied by sensors through communications interface 220. Sathaye [0065], [0073], [0092]; Figs. 1, 2, and 4.)
determining a risk level score representing a likelihood that a (Sathaye discloses processing the pre-sleep monitoring data with a pre-sleep parasomnia-episode likelihood prediction machine-learning model, including a dense or fully connected neural network, to produce a score representing the likelihood of a parasomnia episode during the following sleep window. Its output may contain a separate value for each parasomnia-episode type. Sathaye expressly describes trained parasomnia-episode likelihood models and states that its training-and-deployment techniques may be used for a pre-sleep likelihood model. Sathaye [0030], [0093], [0124]-[0125], [0172]-[0174]; Figs. 5 and 9.)
Primary Sathaye supplies the claimed pre-sleep physiological input, trained model, and future-sleep risk score, but does not expressly identify somnambulism as the selected parasomnia type. Secondary Arrington expressly identifies sleepwalking as somnambulism and teaches monitoring physiological indicators to identify sleepwalking and other parasomnia episodes. Arrington [0002], [0007], [0015], [0044], [0069]-[0070].
A person of ordinary skill in the art would have added a sleepwalking-designated value to Sathaye's disclosed type-indexed output because Arrington identifies sleepwalking as a known parasomnia suitable for sensor-based detection and intervention. The skilled artisan would have used Sathaye's disclosed training-and-deployment techniques with class-specific examples for that known parasomnia type. Sathaye [0030], [0172]-[0174]; Arrington [0044]-[0047], [0052]-[0053], [0069]-[0070]. Sathaye already provides the pre-sleep physiological inputs, trained-model framework, and type-indexed likelihood output, so generating a sleepwalking-designated numerical output would have been a predictable configuration.
Claim 8.
Sathaye teaches,
An apparatus for detecting the onset of a (Sathaye discloses a processor-and-memory apparatus and predictive data analysis computing entity 106 configured to perform in-sleep parasomnia prediction during an ongoing sleep window. Sathaye [0005], [0031], [0064]-[0065], [0091]; Figs. 1, 2, 4, and 7.)
interface circuitry configured to receive movement and/or position measurements of the user generated during the sleep period of the user, and a risk level score representing a likelihood that a (Sathaye discloses communications interface 220, in-sleep movement measurement sequence 714 comprising body-movement measurements during the ongoing sleep window, and an in-sleep static data vector 711 that may include the pre-sleep parasomnia likelihood score for that ongoing sleep window. Sathaye combines the movement-based representation and static vector into the in-sleep model input and permits the disclosed functions to be performed by any number of computing entities. Sathaye [0031], [0073], [0091], [0128]-[0129], [0139]-[0140], [0149]; Figs. 2, 4, and 7.)
processing circuitry configured to detect (Under the construction above, Sathaye discloses processing the movement-based representation together with the static vector containing the pre-sleep score to produce an in-sleep likelihood that the monitored individual is currently experiencing a parasomnia episode. Sathaye [0031], [0129], [0149]-[0150]; Fig. 7.)
Sathaye combines ongoing movement data with the pre-sleep score to determine a current parasomnia likelihood and permits type-indexed outputs, but does not expressly select somnambulism, label the transition into the current episode as onset, or state that the same interface receives both inputs. Secondary Arrington expressly identifies sleepwalking as somnambulism, teaches sensors that identify the onset of a parasomnia episode, continuously repeats data collection and episode determinations, and discloses network interfaces through which distributed or cloud functions and their results are exchanged. Arrington [0002], [0007], [0015], [0044]-[0045], [0052]-[0053], [0056], [0069]-[0070], [0101]-[0104]; Figs. 1 and 9.
A person of ordinary skill in the art would have implemented Sathaye's permitted multi-entity arrangement by using communications interface 220 of the in-sleep apparatus to receive both sensor movement data and the pre-sleep score generated by another computing entity. Arrington supplies a concrete distributed implementation in which a network interface exchanges data and delivers results of remote functions to a client. Arrington [0101]-[0104]. The reason would have been to use remote resources for the pre-sleep computation while retaining time-sensitive in-sleep sensing and detection at the receiving apparatus, consistent with Sathaye's real-time integration objective. Sathaye [0023], [0073], [0091].
Claim 18.
Sathaye teaches,
A method for detecting (Sathaye discloses an in-sleep predictive-data-analysis process for determining whether a monitored individual is currently experiencing a parasomnia episode during an ongoing sleep window. Sathaye [0031], [0091], [0128]-[0150]; Figs. 4 and 7.)
receiving movement and/or position measurements of the user generated during the sleep period of the user; (Sathaye discloses obtaining in-sleep individual monitoring data that includes movement measurement sequence 714, comprising body-movement measurements during the ongoing sleep window. Sathaye [0128], [0139]-[0140]; Fig. 7.)
receiving a risk level score representing a likelihood that a (Sathaye discloses a pre-sleep likelihood score for the sleep window following the pre-sleep window and includes that score in in-sleep static data vector 711 for the associated ongoing sleep window. Sathaye [0030]-[0031], [0125], [0129]; Figs. 5 and 7.)
detecting (Sathaye combines the in-sleep movement-based representation with the static vector containing the pre-sleep score and processes the merged input to determine a likelihood that the monitored individual is currently experiencing a parasomnia episode. Sathaye [0030-0031], [0149]-[0150]; Fig. 7.)
Sathaye therefore supplies the claimed inputs and their joint use in a current-episode determination, but does not expressly select somnambulism or identify the transition constituting onset. Arrington supplies those missing teachings by identifying sleepwalking as somnambulism, expressly targeting identification of parasomnia onset, continuously gathering physiological and movement input, and applying episode thresholds. Arrington [0002], [0007], [0015], [0044]-[0045], [0052]-[0053], [0056], [0069]-[0070]; Fig. 1.
A person of ordinary skill in the art would have trained a sleepwalking-designated value in Sathaye's type-indexed current-episode output because Arrington identifies sleepwalking as somnambulism and a sensor-detectable target. The skilled artisan then would have operated Sathaye's combined movement-and-pre-sleep-score model in Arrington's continuous cycle, applied an episode threshold to the resulting current-episode likelihood, and marked the first transition from a non-episode state to a threshold-satisfying state as onset. Sathaye [0031], [0149]-[0150], [0172]-[0174]; Arrington [0007], [0044]-[0045], [0052]-[0053], [0056], [0069]-[0070].
Claim 2.
Sathaye and Arrington teaches,
The apparatus of claim 1, wherein the interface circuitry is communicatively connectable to a wearable device and is configured to receive the physiological data from the wearable device, wherein the physiological data comprises one or more of cardiac, respiratory, vascular, inertia measurement unit, and localization data of the user.(Sathaye, par. 0028 describe a , 0033, 0035, fig. 1, par. 0046, 0066-0067)
Sathaye discloses body-attached wristbands, headbands, facial sensors, and epidermal patches collecting real-time physiological metrics including ECG, heart rate, pulse, blood oxygen, skin conductance, and motion measurements.
Claim 4.
Sathaye and Arrington teaches,
The apparatus of claim 1, wherein the processing circuitry is configured to match patterns between the physiological data of the user and previous physiological data corresponding to previous somnambulism event occurrences of the user and/or one or more different users. (Sathaye combines current pre-sleep ECG/EEG representations with a historical representation of a preceding period in the same model input. Sathaye [0041], [0124]–[0125]. It also accumulates physiological training entries with targets derived from feedback indicating whether a parasomnia occurred. Sathaye [0172], [0174], [0177]. Although the detailed training embodiment concerns an ongoing sleep window, Sathaye expressly states that the same deployment technique applies to its pre-sleep likelihood model. Sathaye [0173].)
Claim 9.
Sathaye and Arrington teaches,
The apparatus of claim 8, wherein the processing circuitry is configured to detect the onset of a somnambulism event of the user based on one or more thresholds within the movement and/or position measurements, wherein the one or more thresholds are customized according to the user and/or the risk level score.
Sathaye reasonably supplies the movement data, risk-score input, combined detection processing, and a downstream likelihood-score threshold. Refer to par. 0139, 0043, 0099, 0149-0150
It does not supply the added claim-9 relationship: a movement/position threshold customized according to the user or risk-level score.
Primary Sathaye teaches a movement sequence and combines its representation with a static vector that can include the pre-sleep likelihood score to produce an in-sleep score ([0043], [0139], [0149]–[0150]), but thresholds only the resulting score ([0099]); it lacks a user-customized threshold applied to movement measurements.
Secondary Arrington identifies episode “onset” ([0044]), continuously collects user data ([0056]), compares “user movements” with “a known threshold value” to determine an episode ([0052]), and permits “custom parameters” to be “set or adjusted by the user” ([0053]).
A POSITA would have inserted Arrington’s user-adjustable movement comparison into Sathaye’s movement-processing path because Arrington explains that input values fluctuate between “normal sleeping patterns” and “a sleeping disorder episode” ([0052]); the threshold separates episode-indicative movement from ordinary sleep movement.
Claim 10.
Sathaye and Arrington teaches,
The apparatus of claim 8, wherein the processing circuitry is configured to determine a sleep stage of the user during the sleep period, including light sleep, deep sleep, and/or rapid eye movement (REM) sleep, and configured to detect the onset of a somnambulism event of the user further on the condition that user has transitioned to the sleep stage. (Sathaye “determines a detected sleep stage,” including “a rapid eye movement (REM) sleep stage” and “a deep sleep stage,” and calculates the current score “based at least in part on … the detected sleep stage.” Sathaye [0096]–[0097].)
Sathaye determines REM, non-REM, and deep sleep and uses the detected stage in its current parasomnia likelihood. It does not expressly make sleepwalking-onset detection contingent on entry into the stage. Arrington states that parasomnias occur “in association with specific stages of sleep” and identifies “sleepwalking (i.e., somnambulism)” as NREM parasomnia. Arrington [0007].
A POSITA implementing Arrington’s sleepwalking-onset function in Sathaye would enable that decision when Sathaye’s existing stage detector first reports the associated NREM/deep stage. Arrington supplies the reason: sleepwalking is stage-associated; Sathaye already supplies the required stage signal to its real-time model. The modification preserves the sensors, score, and classifier and predictably suppresses onset decisions outside the relevant stage. Although Arrington teaches an association rather than a mandatory gate, using the already-available stage output as that gate directly implements the teaching without changing either system’s operating principle.
Claim 11.
Sathaye and Arrington teaches,
The apparatus of claim 8, wherein the processing circuitry is configured to further identify transitions between somnambulism event stages, including not an event, event onset, sleepwalking event, and/or end of the event.( Sathaye provides likelihoods that the user is “experiencing a parasomnia episode” or “not experiencing” one, and tests whether a later score “falls below” the threshold. Sathaye [0150], [0104].)
Sathaye produces successive event/no-event likelihoods, begins intervention when a threshold is satisfied, and later checks whether a new score falls below the threshold. It does not characterize those changes as transitions between not-event, onset, sleepwalking, or event-end states.
Arrington expressly determines thresholds “to identify when a user transitions from experiencing normal sleeping patterns to experiencing sleep disorder episodes.” Arrington [0058]; see also [0044], [0055]–[0056].
A POSITA would retain Sathaye’s preceding thresholded state and compare it with the current state, identifying the first negative-to-positive change as not-event→event-onset. This converts successive scores into a control event distinguishing when intervention begins from an ongoing positive state. Sathaye states that “time is of the essence” for harm reduction, while Arrington identifies the lack of “immediate assistance” as a problem (Sathaye [0023]; Arrington [0013]). Because both systems already perform repeated threshold determinations, the modification would predictably identify the claimed transition without altering Sathaye’s classifier.
Claim 12.
Sathaye and Arrington teaches,
The apparatus of claim 8, wherein an output of the processing circuitry is communicatively connectable to a somnambulism feedback device and the processing circuitry is configured to cause the somnambulism feed- back device to provide vibratory and/or auditory feedback to the user if the onset of a somnambulism event is detected. (Sathaye can “send instructions/signals to stimulation devices,” “set the vibration,” “broadcast audio data,” and cause audio or tactile generators to perform the intervention. Sathaye [0065], [0068], [0103].)
Sathaye directly supplies the communication and feedback hardware but triggers it from a threshold-satisfying parasomnia likelihood.
Arrington identifies parasomnia onset and expressly provides “an auditory alert” or “a haptic alert” during an episode. Arrington [0021], [0044]–[0045].
A POSITA would route the sleepwalking-onset output supplied by the incorporated Arrington modification to Sathaye’s existing stimulation-device command. Both references use immediate sensory intervention during a detected parasomnia episode, and Sathaye already provides the communication path and auditory/tactile actuators. Substituting the onset state for Sathaye’s generic threshold trigger predictably causes vibration or audio when onset is detected, without hardware redesign.
Claim 13.
Sathaye and Arrington teaches,
The apparatus of claim 8, wherein
the interface circuitry is configured to receive physiological data of the user generated during the sleep period of the user, and wherein the processing circuitry is configured to detect the onset of a somnambulism event of the user further based on the physiological data generated during the sleep period. (Sathaye [0095], [0128]–[0150].)
Sathaye obtains ongoing-sleep ECG/pulse, EEG, EOG, EMG, oxygen, and skin-conductance data, converts the data into representations, merges those representations with movement information, and uses the merged input to generate the current parasomnia likelihood.
Claim 14.
Sathaye and Arrington teaches,
The apparatus of claim 13, wherein the interface circuitry is configured to receive physiological data generated before the sleep period of the user that was used to determine the risk level score, and wherein the processing circuitry is configured to detect the onset of a somnambulism event of the user based on the physiological data generated both before and during the sleep period. (Sathaye [0092]–[0093], [0043], [0129], [0149]–[0150].)
Sathaye obtains pre-sleep ECG/EEG data, generates the pre-sleep likelihood score from that data, obtains ongoing-sleep physiology, and generates the current score from both. It additionally places pre-sleep feature or model-input data directly in the in-sleep static vector.
Claim 15.
Sathaye and Arrington teaches,
The apparatus of claim 13, wherein the interface circuitry is communicatively connectable to a wearable device and is configured to receive the physiological data from the wearable device, wherein the physiological data further comprises cardiac, respiratory, vascular, inertia measurement unit, and/or localization data of the user generated during the sleep period.( Sathaye uses “data provided by one or more sensor devices” and describes “ECG/pulse measurements for an ongoing sleep window” from a sensor “connected to a wrist band of the monitored individual.” Sathaye [0065]–[0066]; see receiving-capable communications interface 220 at [0073].)
Sathaye uses sensor-provided data and expressly obtains ongoing-sleep ECG/pulse measurements from a sensor connected to the user’s wristband. Its processing entity includes communications interfaces capable of receiving data. These passages collectively supply the wearable source, communication path, sleep timing, and cardiac-data category.
Claim 16.
Sathaye and Arrington teaches,
The apparatus of claim 13, the apparatus further comprising a memory, and wherein the processing circuitry is configured to save in the memory and/or export to an external
device at least part of the physiological data of the user generated during the sleep period as part of a sleep-dataset. (Sathaye states that storage subsystem 108 “store[s] input data,” that model inputs include ECG, EEG, and movement representations of the ongoing sleep window, and that a “new training entry” is “added to the training entry set.” Sathaye [0071], [0174], [0176]–[0177].)
Claim 17.
Sathaye and Arrington teaches,
The apparatus of claim 16, wherein
the processing circuitry is based on a machine learning model trained by ground truth information, the ground truth information comprising respective sleep-datasets of the user and/or one or more different users, each sleep-dataset having associated therewith a corresponding label of whether or not a somnambulism event occurred during the sleep period. (Sathaye [0028], [0056-0057], [0172], [0174], [0176]–[0177].)
Sathaye expressly teaches that user feedback for particular sleep windows aggregates “training data entries” used “to train and deploy” a parasomnia likelihood model (Sathaye [0172]). Each entry is associated with “a training model input and a target model output” [0174]. The input contains ECG, EEG, movement, audio, thermal, and emotion representations of the same ongoing sleep window [0176]–[0177], reasonably constituting the claimed sleep-dataset because claim 16 permits “at least part” of the sleep-period physiology. The corresponding target output is determined from feedback by someone aware of “whether the ongoing sleep window includes parasomnia episodes” [0177]. That actual event/no-event outcome used as the target is ground-truth labeling even though Sathaye does not use the words “ground truth.” Because “the user and/or one or more different users” is alternative language, entries from the monitored user satisfy the limitation; different-user datasets are unnecessary. Thus Sathaye directly describes the model-training, sleep-dataset, per-dataset association, and event/no-event-label substance.
Claim 3 is rejected under 35 U.S.C. 103 as being unpatentable over Sathaye et al. in view of Arrington et al., and further in view of Kirollos (US 2018/0071471 A1).
Claim 3.
Sathaye and Arrington teaches,
The apparatus of claim 1, wherein the physiological data is measured during a pre-specified time period before the sleep period, wherein the pre-specified time period is adjustable based on a user input and/or one or more thresholds within the physiological data of the user. ( Sathaye, par. 0027 recite “pre-sleep window” may refer to a data construct that describes a defined-length period of time prior to an expected/scheduled/detected sleep window of the monitored individual, such as a 12 hour period; ,0028, 0009, 0042)
Sathaye [0027], [0092]. It therefore directly supplies a pre-specified measurement period. Sathaye also provides a user-input interface, but does not connect that interface or a physiological threshold—to adjustment of the pre-sleep window’s duration or boundary. Sathaye [0087].
Kirollos describes a user-selected pre-sleep limit that can be dynamically adjusted and adapted to the user’s changing sleep timing. Kirollos [0183]–[0184], [0187]–[0188], [0192], [0199]. Applying that user-adjustable temporal-boundary technique to Sathaye’s existing acquisition window would align the pre-sleep measurements with the user’s actual sleep timing.
Claims 5, 6, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Sathaye et al. in view of Arrington et al., and further in view of Gu et al. (US 2020/0397363 A1).
Claim 20.
Sathaye teaches,
A system for determining a : (Sathaye discloses predictive data analysis system 101 and computing entity 106 performing pre-sleep and in-sleep parasomnia-likelihood functions, including an upcoming-window likelihood score and a current ongoing-window likelihood score, with each functionality performable by any number of computing entities. Sathaye [0030]-[0031], [0064]-[0065], [0091]-[0093]; Figs. 1, 4, 5, and 7.)
interface circuitry configured to receive sleeping physiological data of the user generated while the user was sleeping, the sleeping physiological data including movement and/or position measurements; and(Sathaye discloses communications interface 220 and sensor-data acquisition for an ongoing sleep window. Figure 7 identifies in-sleep ECG sequence 712, EEG sequence 713, movement measurement sequence 714, bedside-audio sequence 715, facial-feature sequence 716, and thermal-camera sequence 717. The movement sequence may contain body-movement and mattress pressure/weight measurements. Sathaye [0065], [0073], [0081], [0095], [0128], [0139]-[0140]; Figs. 1, 2, 4, and 7.)
processing circuitry configured to detect (Sathaye discloses processing that merges ongoing-sleep ECG, EEG, movement, audio, facial, thermal, and static representations and generates a likelihood that the monitored individual is currently experiencing a parasomnia episode. Sathaye [0031], [0128], [0149]-[0150]; Fig. 7.)
(Sathaye discloses a pre-sleep prediction role that processes pre-sleep monitoring data and generates a parasomnia-episode likelihood score for an upcoming sleep window, with the functionality performable by one or more computing entities. Sathaye [0030], [0091]-[0093], [0124]-[0125]; Figs. 4-5.)
interface circuitry configured to receive physiological data of the user generated while the user was , (Sathaye discloses communications interface 220 and acquisition of ECG and/or EEG data for a defined pre-sleep window, which the pre-sleep model uses as input. Sathaye [0027], [0073], [0081], [0092]-[0093], [0124]; Figs. 2, 4, and 5.)
(Sathaye discloses a training entry whose model input is based on ECG, EEG, and movement representations of an ongoing sleep window and whose target output is based on feedback indicating whether that ongoing sleep window included a parasomnia episode.. Sathaye [0008]-[0009], [0176]-[0177]; Fig. 9, step 912, Fig. 1, and Fig. 3.)
processing circuitry configured to determine a risk level score representing a likelihood that a Sathaye discloses a pre-sleep model that merges pre-sleep static, ECG, EEG, medication, historical, and substance-intake representations and generates a parasomnia likelihood score for an upcoming sleep window. The historical representation may use preceding-night ECG and EEG data. Sathaye also states that its dynamic-deployment techniques may be used with a pre-sleep model and separately discloses feedback-labeled training entries for a parasomnia-likelihood model. Sathaye [0041], [0092]-[0093], [0124]-[0125], [0173], [0176]-[0177]; Figs. 4, 5, and 9.)
Primary Sathaye discloses type-indexed current and future parasomnia-likelihood outputs, but does not disclose somnambulism as the type used by either the in-sleep detector or future-risk predictor (Sathaye [0030]-[0031], [0125], [0150]). Secondary Arrington expressly equates sleepwalking with somnambulism and identifies sleepwalking as a condition addressed by parasomnia detection and intervention (Arrington [0002], [0007], [0015]). A POSITA would have configured Sathaye’s relevant in-sleep and pre-sleep type channels for sleepwalking because Arrington identifies that condition for detection and Sathaye seeks timely intervention to reduce harm (Sathaye [0023]). The change preserves Sathaye’s detector/predictor roles and their respective inputs. Because both models already provide values by episode type, the modification predictably would yield somnambulism-specific outputs, and a POSITA reasonably would have expected implementation success.
Primary Sathaye determines from ongoing physiology and movement whether a parasomnia episode is occurring, but does not disclose the episode’s onset (Sathaye [0031], [0128], [0139]-[0140], [0149]-[0150]; Fig. 7). Secondary Arrington expressly teaches identifying parasomnia onset and separately discloses repeated monitoring, threshold episode decisions, and accelerometer-based sleepwalking detection (Arrington [0007], [0015], [0044], [0047], [0052]-[0053], [0056], [0070]; Fig. 1). A POSITA would have modified the output handling of Sathaye’s sleepwalking-configured in-sleep model to designate the first negative-to-positive sleepwalking determination as onset because Arrington identifies the lack of immediate assistance during an episode as a problem and Sathaye treats real-time harm reduction as time-critical (Arrington [0013]; Sathaye [0023]). The modification retains Sathaye’s existing physiology and movement inputs. Repeated classifier outputs make transition marking predictable, and success reasonably would have been expected without replacing the model.
Primary Sathaye creates training entries pairing ongoing-sleep features with feedback indicating whether a parasomnia occurred, but does not disclose sending a somnambulism-labeled dataset from its detector role to a distinct predictor role (Sathaye [0176]-[0177]; Fig. 9). Secondary Gu teaches an implant sending physiology and prediction information to an external monitor; the monitor separately obtains confirmation or negation, associates the records, and supplies them for patient-specific machine learning (Gu [0030]-[0033], [0058]-[0060], [0070], [0072], [0074]-[0077]; Figs. 8-11). A POSITA would have implemented Sathaye’s detector and predictor on communicating apparatuses because Sathaye expressly permits each function to be performed by any number of computing entities. Gu’s multi-device data flow solves the resulting coordination need by transferring labeled sleep outcomes to the separated predictor for individualized updating and improved prediction. See at least: Sathaye [0091], [0172]-[0173], [0177]; Gu [0004]-[0005], [0058]-[0063].
Primary Sathaye’s pre-sleep predictor role obtains pre-sleep ECG and/or EEG and uses those measurements as model input, but does not disclose that the measurements were generated while the user was awake (Sathaye [0091]-[0093], [0124]-[0125]; Figs. 4-5). Secondary Arrington collects physiological signals to determine awake or asleep state, teaches that the signal values vary by state, permits the method to operate throughout the day, and discloses a wearable used during the day and while sleeping (Arrington [0020], [0047]-[0049], [0063]; Fig. 1). A POSITA would have added Arrington’s awake/asleep determination so an all-day wearable could identify awake-generated readings and supply them to Sathaye’s pre-sleep predictor to generate a likelihood score for an upcoming sleep window with more complete data.This prevents sleep-period data from being selected as pre-sleep input while preserving continuous collection the state-separation benefit Arrington uses to prevent normal awake activity from being treated as a sleep episode. Sathaye [0027], [0092]-[0095], [0124]-[0125]; Arrington [0020], [0047]-[0049], [0063].
Claim 5.
Sathaye and Arrington teaches,
The apparatus of claim 1, wherein the processing circuitry comprises a machine learning model, the machine learning model trained and periodically updated by ground truth information comprising previous physiological data of the user and/or one or more different users corresponding to a previous somnambulism event occurrence of the respective user.
Sathaye aggregates labeled entries and trains and deploys a model after an entry-count threshold is met, but it does not disclose recurring updates to the trained model; [0178] optionally updates the score rather than the model. Gu teaches a fixed-interval pipeline that transfers physiological and prediction data, associates that data with confirmation or negation of the actual event, integrates the associated data, retrains an updated patient-specific prediction algorithm, verifies it, and uploads it. Gu [0033], [0060]–[0061], [0065]–[0067], [0070], [0072], [0074]–[0078]. A skilled artisan would have applied Gu’s recurring pipeline to Sathaye’s accumulated event-labeled physiological entries because Sathaye identifies deployment-environment reliability concerns and Gu updates patient-specific prediction algorithms to improve prediction performance. The resulting system would periodically refresh Sathaye’s pre-sleep somnambulism model without changing its operating principle.
Claim 6.
Sathaye and Arrington and Gu teaches,
The apparatus of claim 5, wherein
the interface circuitry is configured to receive a notification after the sleep period of whether or not a somnambulism event occurred during the sleep period, and wherein the processing circuitry is configured to update the training of the machine learning model with a new iteration of ground truth information comprising the physiological data of the user generated before the sleep period, the corresponding risk level score, and a corresponding label of whether or not a somnambulism event occurred during the sleep period.
Sathaye uses “user feedback for the particular time windows” to create training entries whose target identifies “whether the ongoing sleep window includes parasomnia episodes,” but does not expressly state that its interface receives this feedback after the sleep period. Sathaye [0172], [0177]. Gu teaches an external monitoring device that “receive[s] feedback from the user which indicates confirmation or negation” of an event and permits processing when “a predetermined period of time elapses.” Gu [0062], [0070]. A POSITA would have set that period to expire when Sathaye’s sleep window ends because the final outcome for the entire window is then available; an earlier negative label could represent only an incomplete window. The modification predictably provides post-sleep notification of whether somnambulism occurred.
Sathaye supplies pre-sleep physiological input, its corresponding “pre-sleep parasomnia episode likelihood score,” and event feedback used to aggregate training entries. Sathaye [0030], [0172]. However, Sathaye does not expressly disclose updating its pre-sleep model with one new ground-truth instance associating that physiological input and score with the corresponding post-sleep event label. Gu supplies this relationship and update process by associating physiological and prediction information with external data indicating whether the event “occurs … or not,” integrating those data, and using them to generate an updated algorithm. Gu [0065]–[0067], [0074]–[0075]. A POSITA would have applied Gu’s process to Sathaye’s model because Gu teaches that such patient-specific updating provides “an increased rate of successful prediction.” Gu [0060]. Receiving the completed-window outcome avoids incomplete labeling and allows the model to compare its original prediction with the actual result. Because the references use compatible physiological, prediction, feedback, and machine-learning data, the modification predictably produces the claimed new ground-truth training iteration.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JOSHUA DAMIAN RUIZ whose telephone number is (571)272-0409. The examiner can normally be reached 0800-1800.
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, Shahid Merchant can be reached at (571) 270-1360. 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.
/J.D.R./Examiner, Art Unit 3684
/Shahid Merchant/Supervisory Patent Examiner, Art Unit 3684
Prosecution Insights