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 the Claims
The status of the claims as of the response filed 5/6/2026 is as follows: Claims 1-3, 5, 8, 11, 13, and 16-19 are currently amended. Claims 4, 6-7, 9-10, 12, 14-15, and 20 are original. Claims 1-20 are currently pending in the application and have been considered below.
Response to Amendment
Objection to Claims
Claims 5 and 19 have been sufficiently amended to correct the minor informalities objected to in the previous office action, and thus the corresponding objections are withdrawn.
Double Patenting Rejection
The claims have been amended such that they sufficiently diverge in scope from the presently pending claims of parent application 19/191,141, and the corresponding provisional statutory double patenting rejections are withdrawn.
Rejection Under 35 USC 112(b)
Claims 11, 13, and 18 have been amended to sufficiently clarify the indefinite elements and limitations such that the corresponding 35 USC 112(b) rejections are withdrawn.
Rejection Under 35 USC 101
The claims have been amended but the 35 USC 101 rejections are upheld.
Rejection Under 35 USC 103
The amendments made to the claims introduce limitations that are not fully addressed in the previous office action, and thus the corresponding 35 USC 103 rejections are withdrawn. However, Examiner will consider the amended claims in light of an updated prior art search and address their patentability with respect to prior art below.
Response to Arguments
Rejection Under 35 USC 101
On pages 10-11 of the response filed 5/6/2026 Applicant argues that the claimed “continuous, real-time, automated, machine-learning-driven processing of streaming multi-sensor data is not the kind of activity that a clinician (or any human) could perform mentally, with pen-and-paper, or with conventional medical expertise.” Applicant’s arguments are fully considered, but are not persuasive. Examiner maintains that a human actor managing their personal behavior and/or interactions with other people (e.g. a monitored subject) would be capable of observing sensor readouts from at least one type of sensor placed in a subject’s home, performing data processing and mathematical techniques to tag the sensor readouts with a room label based on a floorplan or map, segment the data into windows of a given time duration, determine mathematical correlations among data of different types, determine patterns of usual behavior, and classify anomalies in behavior by looking for deviations in new data as compared to the established usual patterns, as well as use their expertise to assign severities to detected anomalies and select an appropriate recipient for routing a patient care instruction based on the assigned severities. The fact that such abstract functions are being performed by computerized and machine learning means in an automated manner does not preclude them from being abstract, and these computerized features are instead evaluated as additional elements in Steps 2A – Prong 2 and 2B. In the instant case, such high-level computerized features amount to mere instructions to “apply” the abstract idea because they merely digitize and/or automate the otherwise-abstract functions such that they occur in an automated electronic environment. Examiner notes that this analysis is supported by at least para. [0023] of Applicant’s specification, where an effect of the invention is described as “automating the process of monitoring patient activities and generating care instructions.”
On pages 11-12 Applicant argues that “the amended claims recite a particular ordered combination of technical operations that effects concrete technical improvements tied to technical problems identified in the specification,” including at least (i) use of room-tagged passive monitoring sensor devices rather than conventional wearable monitors; (ii) use of a segmentation, feature extraction, and classification pipeline that “is more than a generic ‘use a neural network’ recitation;” (iii) use of a weighted loss training mechanism that prioritizes high-severity anomalies to provide “a concrete technical improvement to the underlying machine-learning framework itself;” and (iv) employment of a severity-based delivery channel routing operation that “is necessarily rooted in computer networking and is not the kind of activity a clinician could perform mentally or with pen and paper.” Applicant’s arguments are fully considered, but are not persuasive.
Regarding (i), Examiner notes that the only purpose of the sensor network as presently recited in the claims is as a source of the data used for the main data analysis and comparison steps of the invention, such that they amount to insignificant extra-solution activity in the form of a means of necessary data gathering (see MPEP 2106.05(g)). Further, Applicant has not invented a new sensor or other monitoring device, provided any technical improvements to the underlying sensing technology or hardware of the sensors themselves, nor utilized a novel or unconventional combination of sensor devices. The independent claims merely recite “receiving, by a processor, in real-time from a sensor network, a continuous stream of sensor data comprising at least one of motion data, occupancy data, and environmental data” such that the configuration of the sensor network appears to include any combination of at least one type of motion sensor, occupancy sensor, and/or environmental sensor. Such passive/ unobtrusive sensor types distributed throughout a monitored home in a location-aware manner are well-understood, routine, and conventional, as evidenced by at least Nudd et al. (US 11633103 B1) Fig. 19; Donegan (US 20230017059 A1) [0004] & [0180]; and Wang et al. (Reference V on the PTO-892 mailed 2/9/2026) section 3 on Pgs 5-7.
Regarding (ii), Examiner notes that the functions of segmenting the enriched sensor data, computing a feature vector including statistical measures and correlation measures for each time window, and classifying each feature to produce classification data comprising a class label and a probability score are part of the abstract idea itself, because they describe mathematical calculations and correlations as well as data processing and analysis steps that a clinician could perform when evaluating sensor data to classify or label activities and behavioral patterns in the data. The fact that such abstract functions are performed “by the processor” and “by applying a neural network classifier trained on historical sensor data and medical data” does not provide any improvements to the structure or technical operation of the processor or neural network, and instead invoke these high-level computing components as tools with which to digitize and/or automate the otherwise-abstract data processing pipeline.
Regarding (iii), Examiner first notes that the training of the neural network classifier is not positively recited as a step of the claimed method, and instead included in “wherein” clauses that merely describe how the neural network classifier has previously been trained before being utilized in the method, such that the training steps are not positively recited additional elements that need be considered in Steps 2A – Prong 2 and 2B (see MPEP 2111.04(I)). However, even if this subject matter were positively recited, Examiner notes that utilizing differently weighted outcomes when training/fitting a classifier model using a loss function is a mathematical concept and would thus be considered part of the abstract idea itself. Examiner respectfully disagrees that such a feature improves the underlying machine-learning framework itself the way that the claims found eligible in Desjardins did; in Desjardins, the claims reflected an improvement to how a machine learning model itself is trained and operates to address the technical problem of ‘catastrophic forgetting’ encountered in continual learning systems, which was identified and explained as a technical problem in the specification. In contrast, the instant specification does not outline a specific technical problem in machine learning technology whose solution is reflected in the claims. Rather, paras. [0112] & [0151] of the specification note that anomaly weights are incorporated into known examples of a loss function (e.g. weighted cross-entropy loss) so that the model can learn to give more importance to critical/severe anomalies compared to less significant deviations, which is not a technical problem in the field of machine learning and instead reflects the ordinary tailoring/training of a machine learning model to fit a specific desired outcome, e.g. assignment of severity scores to detected anomalies.
Regarding (iv), Examiner respectfully disagrees that severity-based care instruction destination routing could not be performed by a human actor such as a clinician. The underlying features of this limitation are first assigning a severity score to a detected anomaly and then determining an appropriate delivery channel from a plurality of delivery channels based on the severity score, which Examiner maintains are steps that a human actor managing their personal behavior could achieve. For example, a clinician could identify an anomaly based on sensor readouts (e.g. determine that a patient has not moved from one room to another in 24 hours), assign a high priority to the anomaly (e.g. designate that the lack of movement over such a long timeframe is an emergency situation), and select an appropriate alert or message destination based on the severity (e.g. determine that emergency services should be contacted rather than merely notifying a relative or the patient themself). Accordingly, Examiner submits that this underlying severity-based routing feature is not “necessarily rooted in computer networking” as Applicant asserts, and is instead part of the abstract idea itself. The fact that such operations are performed by the processor and specifically delivered to electronic device destinations like a patient interface device, a caregiver mobile application, or an automated medication dispensing system then amount to mere instructions to “apply” the abstract idea such that the severity-based routing and delivery of care instructions to the selected entity is digitized and/or automated.
On page 12 Applicant argues that the combination of digital-map-based automatic sensor data tagging, cross-modal feature vector with sensor-medical correlation, severity-weighted loss function, and severity score-based care instruction routing amount to significantly more than any alleged abstract idea. Applicant specifically submits that “none of the reference cited by the Examiner in the Step 2B analysis… discloses or suggests this ordered combination as well-understood, routine, or conventional.” Examiner first notes that issues of patentability over the prior art are a separate consideration to the question of eligibility under 35 USC 101; MPEP 2106.05(I) states that:
Although the courts often evaluate considerations such as the conventionality of an additional element in the eligibility analysis, the search for an inventive concept should not be confused with a novelty or non-obviousness determination. See Mayo, 566 U.S. at 91, 101 USPQ2d at 1973 (rejecting "the Government’s invitation to substitute §§ 102, 103, and 112 inquiries for the better established inquiry under § 101 "). As made clear by the courts, the "‘novelty’ of any element or steps in a process, or even of the process itself, is of no relevance in determining whether the subject matter of a claim falls within the § 101 categories of possibly patentable subject matter." Intellectual Ventures I v. Symantec Corp., 838 F.3d 1307, 1315, 120 USPQ2d 1353, 1358 (Fed. Cir. 2016) (quoting Diamond v. Diehr, 450 U.S. at 188–89, 209 USPQ at 9).
Accordingly, whether the claims are found to be novel and/or non-obvious over the prior art has no bearing on analysis of patent eligibility under 35 USC 101. Additionally, Examiner notes that many of the argued features (e.g. digital-map-based sensor data tagging, computation of cross-modal feature vectors comprising various correlations, use of a severity-weighted loss function, and severity-score-based care instruction destination selection) are part of the abstract idea itself (as explained elsewhere). Because these functions are part of the abstract idea itself, they cannot provide “significantly more” than the abstract idea and thus do not confer eligibility (see MPEP 2106.05(a): “It is important to note, the judicial exception alone cannot provide the improvement. The improvement can be provided by one or more additional elements.” See also 2106.05(a)(II): “it is important to keep in mind that an improvement in the abstract idea itself… is not an improvement in technology”). Examiner maintains that the recited additional elements (e.g. a processor, digital/automatic infrastructure, neural network, and electronic delivery channels), when considered both individually and in combination, do not amount to “significantly more” than the abstract idea itself, as explained in more detail in Step 2B of the updated 35 USC 101 rejections below.
Rejection Under 35 USC 103
On pages 12-14 Applicant argues that the cited prior art fails to teach or suggest many aspects of the amended independent claims. Applicant’s arguments are fully considered, and Examiner agrees, as explained in more detail in the “Subject Matter Free from Prior Art” section below.
Priority
This application’s status as a continuation of 19/191,141 and corresponding claim of priority to provisional patent application 63/639,041 is acknowledged. However, Applicant has not complied with one or more conditions for receiving the benefit of an earlier filing date under 35 U.S.C. 119(e) as follows:
The later-filed application must be an application for a patent for an invention which is also disclosed in the prior application (the parent or original nonprovisional application or provisional application). The disclosure of the invention in the parent application and in the later-filed application must be sufficient to comply with the requirements of 35 U.S.C. 112(a) or the first paragraph of pre-AIA 35 U.S.C. 112, except for the best mode requirement. See Transco Products, Inc. v. Performance Contracting, Inc., 38 F.3d 551, 32 USPQ2d 1077 (Fed. Cir. 1994).
The disclosure of the prior-filed application, provisional Application No. 63/639,041, fails to provide adequate support or enablement in the manner provided by 35 U.S.C. 112(a) or pre-AIA 35 U.S.C. 112, first paragraph for one or more claims of this application. For example, the ‘041 provisional application does not provide adequate support for the following subject matter from the following claims:
wherein each piece of sensor data is associated with a time stamp in claims 1 and 16; paras. [0011]-[0012], [0025], [0036], [0045], & [0058], of ‘041 disclose storing time at which a patient care instruction is delivered to a user device and paras. [0024], [0042], & [0056] disclose that the system can determine if a user is not taking medications on time, but none of these disclosures show that each piece of incoming sensor data is associated with a time stamp.
enriching the sensor data by incorporating room-specific information wherein incorporating the room-specific information comprises (i) accessing a digital map of a living space of the patient in which each sensor of the sensor network is assigned to a respective room, and (ii) automatically tagging each piece of sensor data with the respective room of the sensor from which the piece of sensor data was received in claims 1 and 16; [0028] discloses that sensors may be placed in various rooms of a residence of a patient, but there is no disclosure of a positively implemented data enrichment process that incorporates room-specific information into the received sensor data by accessing a digital map of a living space and automatically tagging each piece of sensor data.
generating activity pattern data by processing the enriched sensor data and medical data of the patient… wherein generating the activity pattern data comprises: segmenting the enriched sensor data into time-windowed data blocks of a predefined duration, computing, for each time-windowed data block, a feature vector that includes statistical measures of the sensor data and correlation measures between the sensor data and the medical data, and classifying each feature vector by applying a neural network classifier trained on historical sensor data and medical data to produce classification data comprising a class label and a probability score in claims 1 and 16; though paras. [0040]-[0041] broadly disclose that pattern recognition may be accomplished by an artificial neural network such as a deep convolutional neural network, there is no discussion of segmenting enriched sensor data into time-windowed blocks of a predefined duration, computing a feature vector including statistical measures and correlation measures for each time-windowed data block, or producing classification data.
detecting an anomaly indicating a potential health risk, wherein the anomaly is detected by comparing the activity pattern data with a baseline activity pattern data of the patient in claims 1 and 16; though paras. [0024], [0042], [0049], & [0056] broadly disclose predicting abnormal (i.e. anomalous) conditions of a patient, there is no disclosure of detecting an anomaly specifically by comparing activity pattern data with a baseline activity pattern data of the patient.
wherein the neural network classifier is trained using a weighted loss function in which a severity weight is assigned to each anomaly type such that high-severity anomalies including prolonged inactivity and a missed medication dose are weighted higher than minor routine variations in claims 1 and 16; paras. [0011]-[0012], [0024], [0040]-[0042] broadly mention training and updating a prediction model, but make no mention of using a weighted loss function or severity weights of specific anomaly types in the training.
assigning a severity score to the anomaly; and determining a delivery channel from a plurality of delivery channels based on the severity score in claims 1 and 16; there is no mention of assigning severity scores to detected anomalies, nor of selecting a specific patient care instruction delivery channel based on the severity score.
wherein the plurality of delivery channel comprises… an automated medication dispensing system; delivering the patient care instructions through the determined delivery channel in claims 1 and 16; paras. [0011]-[0012], [0045], [0048]-[0049], & [0058]-[0059] disclose delivering patient care instructions to various example user devices, while paras. [0012] & [0026] disclose that medication adherence data may be collected from a medication dispenser. However, none of these disclosures support the patient care instruction being delivered to the specific delivery channel of an automated medication dispensing system.
creating a natural language summary by processing the activity pattern data, wherein the natural language summary is generated by a fine-tuned Large Language Model that has been fine-tuned based on historical sensor data and medical data in claims 5 and 19; though paras. [0025] & [0057] broadly disclose that a LLM or generative AI may be used to generate the patient care instructions, there is no disclosure of specifically creating a natural language summary using a fine-tuned LLM that has been fine-tuned based on historical sensor data and medical data.
generating patient care instructions based on the natural language summary in claim 6; because there is no disclosure of creating a natural language summary as explained above, there is also no disclosure of generating patient care instructions based on the natural language summary.
wherein the patient care instructions are generated to specify a recommended course of action based on the natural language summary in claim 7; because there is no disclosure of creating a natural language summary as explained above, there is also no disclosure of the patient care instructions specifying a recommended course of action based on the natural language summary.
door sensors comprising magnetic-contact sensors mounted on doors and configured to record entry and exit times from different rooms in claim 8; though [0026] provides an example sensor as being “a sensor for detecting opening and closing of objects placed in the residence of the patient,” and [0027] discloses “a motion and position sensor such as… magnetometer,” there is no disclosure of the specific magnetic-contact sensors mounted on doors, nor of the sensor specifically recording entry and exit times from different rooms.
seating pressure sensors integrated into chairs configured to monitor sitting duration and frequency in claim 8; though [0026] provides an example sensor as being “a pressure sensor for sensing a chair occupancy… of a patient,” there is no disclosure of the sensor being specifically integrated into chairs and being configured to specifically monitor sitting duration and frequency.
bed mat sensors comprising at least one of pressure sensors and capacitive sensors configured to track sleep patterns, including sleep duration, restlessness and optionally heart rate data in claim 8; though [0026]-[0027] provide an example sensor as being “a pressure sensor for sensing a… bed occupancy of a patient” and pressure sensors including a capacitive sensor, and paras. [0012], [0024], [0039]-[0042], [0048], [0056], & [0061] broadly disclose the system as being able to determine sleep patterns and conditions such as oversleeping, there is no disclosure of the sensor being specifically configured to track sleep duration, restlessness, or heart rate data.
the health trends including disease progression, symptom flare-ups in claim 9; though ‘041 broadly discusses determining patterns or trends in user data, there is no discussion of specific trends including disease progression or symptom flare-ups.
wherein the sensor data is received from a set of sensors in raw format… wherein the sensor data from the set of sensors is mapped and synchronized into a standardized format, with each piece of standardized sensor data further including a time stamp associated with the corresponding sensor data in claims 14 and 20; though [0038] briefly discusses storing sensor data and medical data in a predetermined or a standardized data template, it does not specifically disclose receiving sensor data in a raw format and actively mapping and synchronizing the raw data into a standardized format which includes a time stamp.
Claims 1-2, 5-9, 14, 16-17, and 19-20, (as well as claims 3-4, 10-13, 15, and 18 by virtue of dependency) are therefore not entitled to the priority date of the provisional application and instead receive the effective filing date of parent application 19/191,141: 4/28/2025.
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 the claimed invention is directed to an abstract idea without significantly more.
Step 1
In the instant case, claims 1-15 are directed to a method (i.e. a process) and claims 16-20 are directed to a system (i.e. a machine). Thus, each of the claims falls within one of the four statutory categories. Nevertheless, the claims fall within the judicial exception of an abstract idea.
Step 2A – Prong 1
Independent claims 1 and 16 recite steps that, under their broadest reasonable interpretations, cover mathematical concepts as well as certain methods of organizing human activity, e.g. managing personal behavior, relationships, or interactions between people. Specifically, claim 1 (as representative) recites:
receiving, by a processor, in real-time from a sensor network, a continuous stream of sensor data comprising at least one of motion data, occupancy data, and environmental data, wherein each piece of sensor data is associated with a time stamp;
enriching, by the processor, the sensor data by incorporating room-specific information wherein incorporating the room-specific information comprises (i) accessing a digital map of a living space of the patient in which each sensor of the sensor network is assigned to a respective room, and (ii) automatically tagging each piece of sensor data with the respective room of the sensor from which the piece of sensor data was received;
generating, by the processor, in real time, activity pattern data by processing the enriched sensor data and medical data of the patient by using a pattern recognition model, wherein generating the activity pattern data comprises:
segmenting the enriched sensor data into time-windowed data blocks of a predefined duration,
computing, for each time-windowed data block, a feature vector that includes statistical measures of the sensor data and correlation measures between the sensor data and the medical data, and
classifying each feature vector by applying a neural network classifier trained on historical sensor data and medical data to produce classification data comprising a class label and a probability score, wherein the activity pattern data comprises the activity data of the patient being indicative at least one of mobility, sleep patterns, and medication adherence, statistical measures of sensor data, temporal patterns, and correlation between the sensor data and the medical data;
detecting, by the processor, an anomaly indicating a potential health risk, wherein the anomaly is detected by comparing the activity pattern data with a baseline activity pattern data of the patient, wherein the neural network classifier is trained using a weighted loss function in which a severity weight is assigned to each anomaly type such that high-severity anomalies including prolonged inactivity and a missed medication dose are weighted higher than minor routine variations;
predicting, by the processor, health of the patient based on the anomaly and the activity pattern data by utilizing a prediction model, wherein the prediction model is trained on data from other patients, medical data associated with the patient, to recognize patterns that correlate with health conditions;
generating, by the processor, patient care instructions based on the predicted health of the patient, and assigning a severity score to the anomaly; and
determining, by the processor, a delivery channel from a plurality of delivery channels based on the severity score, wherein the plurality of delivery channel comprises a patient interface device, a caregiver mobile application, and an automated medication dispensing system;
delivering, by the processor, the patient care instructions through the determined delivery channel.
But for the recitation of generic computer components like a processor, digital/automatic features, a neural network, etc., the italicized functions, when considered as a whole, describe a patient monitoring, anomaly detection, and clinical recommendation operation that could be achieved via mathematical operations and by a human actor such as a clinician or other medical professional managing their personal behavior and/or interactions with others. For example, a clinician could obtain various types of timestamped sensor data from sensor readouts and enrich the data by adding room labels corresponding to the location where each data stream was obtained within a dwelling based on referencing a map. The clinician could utilize mathematical operations to segment the enriched data into time-windowed blocks, compute a feature vector including statistical measures and correlation measures among the data, and classify each vector by applying a previously trained/fitted classifier to the data to identify activity patterns related to mobility, sleep, medication adherence, temporal patterns, etc. The clinician could then compare the identified patterns to known baselines activity patterns for that patient to determine if any anomalies or deviations are present that might indicate a potential health risk (e.g. the patient normally wakes up at 7:00 AM and walks from the bedroom to the bathroom, but no movement between rooms has been detected on a given morning at that time after the patient wakes up). The clinician could then use their medical expertise and a prediction model (e.g. a statistical model, a decision tree, a list of rules, etc.) that has been fitted with data from other past patients to predict a health state of the patient based on the detected activity patterns and anomalies (e.g. predicting that the patient in the above scenario may have fallen and is too injured to get up based on known patterns of falls learned from previous patients who have fallen) as well as assign a severity to the anomaly (e.g. an high severity if the patient is believed to be incapacitated and need external assistance). The clinician could finally use their medical expertise to formulate a suggested patient care action based on the predicted health state (e.g. that someone should check on a patient suspected to have fallen), select an appropriate recipient for the suggested action based on the severity of the anomaly (e.g. a caregiver with clinical expertise), and deliver the action to the selected recipient (e.g. via verbal communication, a written report, etc.). Accordingly, claim 1 recites an abstract idea in the form of a certain method of organizing human activity. Claim 16 recites substantially similar subject matter as claim 1 and is also found to recite an abstract idea under the same analysis.
Dependent claims 2-15 and 17-20 inherit the limitations that recite an abstract idea from their dependence on claims 1 and 16, respectively, and thus these claims also recite an abstract idea under the Step 2A – Prong 1 analysis. In addition, claims 2-7, 9, 11-15, and 17-20 recite additional limitations that further describe the abstract idea identified in the independent claims.
Specifically, claims 2 and 17 specify that the predefined duration is a 30-minute time window and that the feature vector further includes a Pearson correlation between motion detections and environmental temperature fluctuations and a binary indicator of medication timing alignment. These limitations lend further detail to the abstract mathematical operations and data representations, e.g. by specifying a specific duration for the segmenting operation, specifying a specific type of correlation captured in the feature vector, and describing a binary indicator in the feature vector, such that they merely further describe the abstract idea itself.
Claims 3 and 18 recite capturing temporal dependencies in activity data of a patient across a sequence of time-windowed data blocks, which merely further describes mathematical analysis of the correlations between different types of temporal data and thus merely further describes the mathematical analysis concepts utilized in the independent claims.
Claim 4 specifies that the room specific information provides insight into activity of the patient in different areas of a home, which is a type of labeling that a clinician would be capable of adding to sensor data in line with its location of origin within a home.
Claims 5 and 19 creating a natural language summary by processing the pattern data to produce actionable insights related to patient care, which a clinician could achieve by evaluating the data and using their medical expertise to write up a summary or report understandable to a lay person.
Claims 6-7 recite generating patient care instructions based on the natural language summary, the instructions specifying one or more actions of various types to be performed to assist the patient, which a clinician could achieve by using their medical expertise to come up with recommended care actions (e.g. situational advice, treatment plan modifications, medication reminders, or alerts) for the patient based on their summary of the patient’s situation.
Claim 9 recites that the patient care instructions include recommendations based on historical health trends derived from medical data, the health trends including disease progression, symptom flare-ups, or treatment adherence, which a clinician could achieve by observing trends in the specific patient’s disease progression, symptom exacerbations, and/or treatment adherence and/or known trends in these domains from experience with previous patients and considering such trends when making care recommendations.
Claim 11 recites storing the activity pattern data and the natural language summary in a database for future use and refinement of the prediction model, which a clinician could achieve by storing such determinations in a patient’s chart, medical records, or other aggregated files with the intention to consider them in the future for making patient care determinations and adjusting predictive models.
Claim 12 recites that the instructions are delivered in real-time based on a dynamic analysis of the patient’s current status as reflected in the most recent sensor and medical data, which a clinician could achieve by making dynamic evaluations about the patient’s status based on the most up-to-date data available and communicating care recommendations to the patient and/or another caregiver in real-time (e.g. by speaking to them).
Claim 13 recites authenticating a caregiver before delivery of the instruction data, which a clinician could achieve by visually identifying a caregiver before communicating care recommendations to them, e.g. by looking at the face of someone claiming to be a patient’s wife and comparing it to a picture of the wife on file before providing information about the patient’s care.
Claims 14 and 20 recite mapping and synchronizing sensor data into a standardized format including a time stamp, which a clinician could achieve by temporally aligning received sensor readings and performing any necessary conversions such as unit standardization.
Claim 15 specifies that the medical data is prestored and associated with the patient and includes at least one of medical conditions, treatment history, medication data, and vital sign data. A clinician would be capable of storing these types of data (e.g. in a patient’s chart or medical records) and accessing the prestored data as needed when performing activity pattern analysis.
However, recitation of an abstract idea is not the end of the analysis. Each of the claims must be analyzed for additional elements that indicate the abstract idea is integrated into a practical application to determine whether the claim is considered to be “directed to” an abstract idea.
Step 2A – Prong 2
The judicial exception is not integrated into a practical application. In particular, independent claims 1 and 16 do not include additional elements that integrate the abstract idea into a practical application. The additional elements of claims 1 and 16 include a processor to perform the various steps, a sensor network from which the sensor data is received, specifying that the map is digital and that the sensor data is tagged automatically, use of a trained neural network classifier to produce the classification data, and presumably electronic delivery of the patient care instructions to a delivery channel comprising a patient interface device, a caregiver mobile application, or an automated medication dispensing system. The use of a processor to perform the receiving, enriching, accessing, tagging, generating, segmenting, computing, classifying, detecting, predicting, assigning, determining, delivering, etc. steps in a digital environment and automated manner amounts to instructions to “apply” the judicial exception because these otherwise-abstract functions are merely being automated and/or digitized with a high-level computing component (see MPEP 2106.05(f)). Similarly, the use of a neural network classifier to produce the classification data amounts to instructions to “apply” the judicial exception because the otherwise-abstract function of analyzing data to label it with a class output and calculate a probability score are merely being digitized and/or automated with a high-level neural network. Receiving the sensor data specifically from a sensor network amounts to insignificant extra-solution activity in the form of data gathering, because the sensor network merely serves as a means of obtaining sensor data for the main analysis steps of the invention (see MPEP 2106.05(g)). Delivering the patient care instructions through an electronic delivery channel to at least one of a patient interface device, a caregiver mobile application, or an automated medication dispensing system also amounts to mere instructions to “apply” the judicial exception because the otherwise-abstract function of communicating patient care instructions to a selected recipient is being digitized and/or automated such that it occurs in an electronic environment with high-level computing components like an interface device or mobile application. Accordingly, claims 1 and 16 as a whole are each directed to an abstract idea without integration into a practical application.
The judicial exception recited in dependent claims 2-15 and 17-20 is also not integrated into a practical application under a similar analysis as above. The functions of claims 2, 4, 6-7, 9, 12, 14-15, 17, and 20 are performed with the same additional elements introduced in the independent claims, without introducing any new additional elements of their own, and accordingly also do not integrate the judicial exception into a practical application.
Claims 3 and 18 specify that the neural network classifier is a deep learning model comprising at least one of a CNN and LSTM RNN. These additional elements again amount to mere instructions to “apply” the judicial exception, because the otherwise-abstract functions of classifying a feature vector to produce classification data and recognizing patterns indicative of health conditions are merely being digitized and/or automated with high-level neural network model types with no details about how the CNN or LSTM RNN actually achieve the functionally claimed outcome. See the Step 2A – Prong 2 analysis for a similar artificial neural network element in claim 2 of Example 47 of the AI-based subject matter eligibility examples released in July 2024.
Claims 5 and 19 recite a fine-tuned Large Language Model that is used to generate the natural language summary. This additional element similarly amounts to instructions to “apply” the judicial exception, because the otherwise-abstract function of generating a summary of data is merely being digitized and/or automated with a high-level LLM with no details about how the LLM actually achieves the functionally claimed outcome.
Claim 8 specifies that the sensor network comprises motion sensors configured to detect patient movement, door sensors comprising magnetic-contact sensors mounted on doors and configured to record entry and exit times from different rooms, seating pressure sensors integrated into chairs configured to monitor sitting duration and frequency, and bed mat sensors comprising at least one of pressure sensors and capacitive sensors configured to track sleep patterns. These types of sensors are still only recited as a means of obtaining the data necessary for the main analysis steps of the independent claims, such that they still amount to insignificant extra-solution activity in the form of mere data gathering means.
Claim 10 specifies that the delivery of the instructions is performed using at least one of a voice assistant and a mobile device. Use of either of these two elements to delivery patient care instructions amounts to instructions to “apply” the judicial exception because the otherwise-abstract function of communicating patient care instructions to a user is merely being digitized and/or automated with high-level computing components.
Claim 11 recites storing data in a database, which Examiner assumes is intended to be an electronic database for purposes of examination. Storing data in an electronic database amounts to mere instructions to apply the judicial exception because the otherwise-abstract function of storing data is merely being digitized such that it occurs digitally. This step may also be considered insignificant extra-solution activity because it nominally stores data electronically with a nominal intended result of future use of the stored data.
Claim 13 specifies that authenticating a caregiver is performed using electronic visit verification that captured biometric authentication data. Use of EVV to perform the authentication amounts to mere instructions to “apply” the judicial exception because the otherwise-abstract function of authenticating the identity of a user prior to divulging medical information is merely being digitized and/or automated such that it occurs with high-level electronic infrastructure.
Accordingly, the additional elements of claims 1-20 do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea. Claims 1-20 are directed to an abstract idea.
Step 2B
The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to integration of the abstract idea into a practical application, the additional elements of a processor, digital/automatic infrastructure, and delivery channels of a patient interface device, caregiver mobile application, or automated medication dispensing system amount to mere instructions to “apply” the judicial exception using generic computing components. As evidence of the generic nature of the above recited additional elements, Examiner notes that Applicant’s specification is silent as to the particulars of the processor and delivery devices; paras. [0025] & [0184]-[0185], briefly discuss these additional elements in the same level of detail as the claims, leaving one of ordinary skill in the art to understand that any generic processor-based device may be utilized as the processor and that any known type of user device (e.g. smart phones, PDAs, tablets, laptops, or other user interface devices) capable of executing software applications may be utilized as means for delivering the patient care instructions.
The sensor network of claims 1 and 16, whose architecture is further detailed in claim 8, amounts to insignificant extra-solution activity in the form of a means for necessary data gathering (as explained above). Examiner further notes that it is well-understood, routine, and conventional to receive sensor data from a sensor network including motion sensors, door sensors, seating pressure sensors, and bed mat sensors for the purpose of patient pattern analysis and monitoring, as evidenced by at least Ruth et al. (US 20250325238 A1) [0044]; Wang et al. (Reference V on the PTO-892 mailed 2/9/2026) section 3.1 & Table 7; and Jefferson et al. (US 20180122209 A1) Figs. 1A & 2, [0023]-[0024] & [0053].
The use of a deep neural network model as in claims 1, 3, 16, and 18 as well as an LLM as in claims 5 and 19 to perform various otherwise-abstract functions also amount to instructions to “apply” the judicial exception using high-level computing components, as explained above. Examiner notes that Applicant’s specification provides no specific details about how the DNN or LLM are trained or utilized in a manner that would provide improvements to the underlying machine learning technology of these models, and instead appears to merely apply known machine learning modelling techniques to the use case of patient activity pattern identification and summarization. Further, it is well-understood, routine, and conventional to utilize trained deep neural networks like CNN and LSTM RNN for activity classification, as evidenced by at least Wang Table 7 & section 5.3 on Pgs 17-18; Lu (Reference W on the PTO-892 mailed 2/9/2026) abstract & section 2.3 on Pgs 15-18; and Ankalaki et al. (Reference X on the PTO-892 mailed 2/9/2026) section VIII on Pgs 93477-93478 & section XII on Pg 93481; as well as to utilize LLMs to process/summarize health-related data, as evidenced by at least Ruth [0032], [0055], & [0112]; Baronov (US 20250325237 A1) [0269]-[0286]; Matsui (US 20250325209 A1) [0081]; and Vohra (US 20240062877 A1) [0045].
The use of a voice assistant or a mobile device for delivery of the patient care instructions as in claim 10 also amounts to mere instructions to “apply” the judicial exception, as explained above. Examiner notes that no details about a mobile device or voice assistant are disclosed in the specification, with para. [0188] even providing an example of a known voice assistant such as Amazon’s Alexa. This disclosure leaves one of ordinary skill in the art to understand that any known mobile device and/or voice assistant technology may be utilized with the invention to deliver information to a user.
The use of an electronic database to store data as in claim 11 amounts to mere instructions to “apply” the judicial exception and/or insignificant extra-solution activity in the form of nominal data storage, as explained above. Applicant’s specification provides no specific details about the architecture of the database, leaving one of ordinary skill in the art to understand that any known type of electronic database could be utilized to store the data. Further, Examiner notes that it is well-understood, routine, and conventional to store and retrieve information in memory, as outlined in MPEP 2106.05(d)(II).
The use of electronic visit verification to authenticate a caregiver as in claim 13 amounts to mere instructions to “apply” the judicial exception, as explained above. Examiner notes that the specification does not provide any specific details about the electronic visit verification process beyond that it may include capturing biometric data such as fingerprint or facial recognition data as in [0010], leaving one of ordinary skill in the art to understand that any known means of authenticating user identity via biometrics may be utilized. Further, Examiner notes that it is well-understood, routine, and conventional to utilize biometric-based user authentication techniques in a health data environment, as evidenced by at least Ruth [0105] & [0110]; Saavedra et al. (US 20230326318 A1) [0092]-[0093]; and Francois (US 20170262604 A1) [0085] & [0389].
Further, the combination of these additional elements is not expanded upon in the specification as a unique arrangement and as such relies on the knowledge of one of ordinary skill in the art to understand a combination of various known computing components appropriate for implementing the claimed system. Additionally, the combination of a processor analyzing data received from a sensor network using machine learning predictive techniques like deep neural networks to identify patterns and anomalies in patient data which may be stored in a database and corresponding information may be delivered to user devices with biometric authentication means is well-understood, routine, and conventional, as evidenced by at least Ruth abstract, [0036], [0040], [0044], [0105], & [0136]; and Crabtree et al. (US 20250349399 A1) abstract, [0022]-[0023], [0157], [0176]-[0177], & [0206].
Analyzing these additional elements as an ordered combination adds nothing that is not already present when considering the elements individually; the overall effect of the processor, digital/automatic infrastructure, neural network and other machine learning-based models, sensor network, user devices, etc. in combination is to digitize and/or automate a patient monitoring, anomaly detection, and clinical recommendation operation that could otherwise be achieved via mathematical concepts and as a certain method of organizing human activity. Thus, when considered as a whole and in combination, claims 1-20 are not patent eligible.
Subject Matter Free from Prior Art
The following is a statement of reasons for the indication of subject matter free from prior art:
The prior art of record fails to expressly teach or suggest, either alone or in combination, each and every feature of the independent claims. In particular, the prior art fails to teach all of enriching the sensor data via a digital map in the manner claimed; segmenting, vectorizing, and classifying the enriched sensor data in the manner claimed; using a weighted loss function to train a neural network classifier to prioritize weighting of specific types of high-severity anomalies in the manner claimed; and assigning a severity score to a detected anomaly and determining a delivery channel from a plurality of delivery channels based on the severity score in the manner claimed, in combination with all of the other patient monitoring and anomaly detection features of the independent claims. Upon completion of an updated prior art search, Examiner submits that the closest related art includes:
- Ruth et al. (US 20250325238 A1), disclosing a patient monitoring system that collects time-stamped sensor data from sensors deployed in different rooms of a subject’s living space, uses trained AI models to determine baseline activity patterns of the subject and detect anomalies or deviations from the patterns, and determine and output appropriate patient care messages to various delivery channels; but failing to explicitly disclose the elements of digital map-based enrichment, segmentation, weighted loss function, and delivery channel selection based on anomaly severity scores.
- Jefferson et al. (US 20180122209 A1), Saavedra et al. (US 20230326318 A1), Ben-David (US 20240404708 A1), and Williams et al. (US 10825318 B1), disclosing similar AI-based patient monitoring systems that detect patterns and anomalies in patient sensor data from sensors deployed around a subject’s living space and deliver appropriate patient care instructions or recommendations, but failing to explicitly disclose the combination of digital map-based sensor data enrichment, segmentation and vectorization, weighted loss function, and delivery channel selection based on anomaly severity scores as recited in the instant claims.
- Donegan (US 20230017059 A1, para. [0180]) and Cope (US 11816600 B1, Col8 L12-18 & Col16 L35-67), describing sensor-based monitoring systems that include associating sensor data with certain rooms based on a digital map; but failing to explicitly disclose each and every feature of the independent claims.
- Chen (Reference U on the PTO-892 mailed 2/9/2026; chapter 5) and Wang (Reference V on the PTO-892 mailed 2/9/2026; section 3 on Pg 5, Fig. 2 on Pg 6, sections 4.1-4.2 on Pgs 10-12), describing methods of temporal segmentation and feature extraction of sensor data in activity monitoring systems for the purpose of neural network-based classification; but failing to explicitly disclose each and every feature of the independent claims.
- Donegan (US 20230017059 A1, paras. [0007], [0028], [0070]-[0072]) and Sprigg (CN 104334075 A, para. [0084]), describing assigning severity to detected events and transmitting alerts to appropriate destinations based on the severity of the events, but failing to explicitly disclose each and every feature of the independent claims.
- Firoozye (US 20260018281 A1), Kim et al. (US 20190090786 A1), and Otto et al. (US 8868616 B1), disclosing patient monitoring systems that use machine learning methods to learn and adjust weights or severities of different detected patient events/anomalies, but failing to explicitly disclose each and every feature of the independent claims.
Though many aspects of the independent claims are disclosed in the prior art, it would not have been obvious to one of ordinary skill in the art to combine the disparate features into the invention of the instant claims. Accordingly, the prior art, either alone or in combination, does not disclose or render obvious all the features of the independent claims and they are found to recite allowable subject matter, as are the claims depending therefrom.
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 KAREN A HRANEK whose telephone number is (571)272-1679. The examiner can normally be reached M-F 8:00-4:00 ET.
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.
/KAREN A HRANEK/ Primary Examiner, Art Unit 3684
Prosecution Insights