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 Claims
The present application is being examined under the claims filed 08/28/2025.
Claims 1-20 are pending.
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 09/09/2025 is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Response to Amendment
This Office Action is in response to Applicant’s communication filed 08/28/2025 in response to office action mailed . The Applicant’s remarks and any amendments to the claims or specification have been considered with the results that follow.
Response to Arguments
Regarding double patenting rejections
In Remarks page 11, Argument 1
The Office Action provisionally rejects claims 1-20 on the ground of nonstatutory double patenting as allegedly being unpatentable over claims 1-19 of copending Application No. 17/368,407 in view of U.S. 2020/0218974 Al to Cheng (hereinafter, "Cheng"). Applicant respectfully requests that the nonstatutory double patenting rejection be withdrawn in light of the amendments to the claims.
Examiner’s response to Argument 1
Examiner disagrees. The amendments to the claims are minor and do not substantially change the scope of the claims. Therefore, the double patenting rejections are maintained.
Regarding 35 U.S.C. 101 rejections
In Remarks page 12, Argument 2
In line with the court's requirements for patent-eligible subject matter, the present claims do recite improvements to machine learning models. Notably, the claims recite a non conventional machine learning framework that enables accurate fall risk assessment by processing data from disparate data sources. See Specification, [0061-0062] Current methodologies "configured to predict a user's fall risk are limited, as these methodologies are unable to process data from disparate data sources to generate a dynamic fall prediction for the user". Id Thus, improvements to a machine learning model's ability to generate predictions based on data from disparate data sources, such as those recited by the claims, provide patenteligible improvements in technology.
Examiner’s response to Argument 2
Examiner disagrees. As discussed in the responses to arguments below, the claims are primarily directed to processes that could be performed mentally and the additional elements do not substantially limit the recitation of mental processes. See Examiner’s responses and rejections under 35 U.S.C. 101 below.
In Remarks page 11, Argument 3
As amended, the claims recite improved techniques for predictive data analysis and corresponding fall prediction and prevention that are based on outputs provided by a fall prediction model comprising a first recurrent neural network framework, a second RNN framework, a fully connected neural network framework, and an ensemble machine learning framework. See Specification, [0046-0048]. The improved techniques are not mental processes as defined by the Manual of Patent Examining Procedure ("MPEP"), § 2106 and, even if they are, they are directly applied to improve interpretability of machine learning technologies. A more detailed discussion regarding the eligibility of the claims is provided below with respect to claim 1, as an example, although the remarks apply similarly to the rest of the claims. Applicant respectfully submits that independent claim I is directed to patent eligible subject matter under 35 U.S.C. § 101, according to at least the Step 2A analysis set out by the MPEP.
Examiner’s response to Argument 3
Examiner disagrees. The claimed techniques are directed to a mental process and are not directed to any technical improvement. Further details are provided in the responses to arguments below and the rejections under 35 U.S.C. 101.
In Remarks page 13, Argument 4
(Examiner summarizes Applicant’s arguments) Applicant argues that the limitations of claim 1 cannot be performed in the human mind. Applicant argues that all of the limitations claimed are necessarily rooted in machine technology and because they involve complex data and operations that could not be performed in the human mind. Applicant cites to the limitations of claim 1, and also cites examples in the MPEP of mental processes and argues that claim 1 is not directed to any of the examples.
Examiner’s response to Argument 4
Examiner disagrees. MPEP 2106.04(a)(2) III. recites “The courts consider a mental process (thinking) that "can be performed in the human mind, or by a human using a pen and paper" to be an abstract idea” and “the "mental processes" abstract idea grouping is defined as concepts performed in the human mind, and examples of mental processes include observations, evaluations, judgments, and opinions.” MPEP 2106.04(a)(2) III. C. further recites “Claims can recite a mental process even if they are claimed as being performed on a computer.” Accordingly, Examiner has determined that many of the limitations of claim 1 could be performed in the human mind. For example, consider the following limitation:
generating, by the one or more processors and by processing the one or more categorical timeseries feature data fields using a second RNN framework of the fall prediction machine learning model of the fall prediction machine learning model, a categorical timeseries embedding for the user feature date object”
RNN’s are a specialized type of neural network that generates a sequence given an input. RNNs are usually composed of simple operations such as mathematical functions, addition, multiplication, matrix multiplication, and activation functions. RNNs may also include simple algorithm operations that are not necessarily directed to math, but rather to algorithmic evaluations of data. That is, generating a sequence using an RNN amounts to evaluating given data using a series of simple operations which could be performed in the human mind or by a human using pen and paper as a tool. Although the limitation is claimed as using processors, the MPEP is clear that limitations which could otherwise be performed mentally but are merely applied on a general-purpose computer are still directed to mental processes. Similar reasoning applies to other limitations identified by the examiner as being directed to a mental process. See the rejections under 35 U.S.C. 101 below for a complete analysis.
In Remarks pages 15-17, Argument 5
(Examiner Summarizes Applicant’s arguments) Applicant argues that claim 1 is directed to an improvement to technology by providing capability to process disparate data and improves fall prediction machine learning. Applicant cites to the specification which states that obtaining data from disparate data sources can address the issues faced by current methodologies, and moreover that current methodologies do not provide a way to dynamically predict a fall cause nor advise a user on corrective actions to reduce their fall likelihood. Applicant argues that the specification is sufficient such that a person having ordinary skill would recognize the improvement and that claim 1 provides this improvement (in particular that disparate data sources are input and processed to enable generation of a fall prediction data object). Applicant concludes that claim 1 achieves higher performance and usability than traditional methods for inference and that claim 1 integrates the alleged abstract ideas into a practical application.
Examiner’s response to Argument 5
Examiner disagrees. Although the claim does recite limitations related to receiving data from disparate sources and sending predictions and suggestions, these limitations are recited at a high level of generality. Both alone and in combination, these limitations amount to mere sending and receiving data over a network. These are highly generic tasks that can be performed with any ordinary computer and the specification nor the claims provide more specific details about how the collection of data itself provides any improvement in the art. Rather, a person having ordinary skill in the art would recognize that any alleged improvement would come from the processing of data, which as argued above and in the rejections under 35 U.S.C. 101 below, is directed to a mental process alone. That is, the alleged improvements to performance and usability merely arrive by processing of data which could be performed in the human mind with the mere addition of generic inputting/outputting of data. Thus the Examiner maintains that the claim does not include the components or steps necessary to provide an improvement apart from the abstract ideas alone. Therefore, the claims cannot be directed to an improvement to technology. The rejections under 35 U.S.C. 101 of the independent and all dependent claims are maintained accordingly.
Regarding 35 U.S.C. 103
In Remarks page 17-18, Argument 6
(Examiner summarizes Applicant’s arguments) Applicant argues that none of the cited references, alone or in combination teaches: "one or more sensory notifications indicating at least one corrective action, the at least one corrective action configured to prevent or lessen the severity of the predicted fall event". Applicant argues that “suggestions to prevent the detected event in the future” does not describe “sensory notifications indicating at least one corrective action”. Applicant argues that the rejections of the independent claims should be withdrawn, as with the dependent claims for containing the same deficiencies.
Examiner’s response to Argument 6
Examiner disagrees. Rakshit teaches (paragraph [0023]) “in certain embodiments, safety manager 110 can generate a text report to accompany the visual simulation that notes any predicted change with respect to the cause of the event, an appropriate notification of the detected event, and suggestions to prevent the detected event in the future. For example, safety manager 110 can recommend additional actions a user can take to increase safety (e.g. personal protective equipment, protocols, etc.).” The broadest reasonable interpretation of “one or more sensory notifications indicating at least one corrective action” includes the notifications including suggestions to prevent the detected event and actions for a user to increase safety because increasing safety and preventing an event is a corrective action. Therefore, the rejection is maintained. The rejections of the dependent claims are maintained for similar reasons.
In Remarks page 18, Argument 7
For all the reasons above, Applicant believes that the present application is now m condition for allowance and respectfully requests favorable reconsideration of the Application. Should there be any issue that impedes allowance of a claim, the Examiner is invited to telephone the undersigned attorney so that the issue can be expeditiously resolved.
In response to Argument 7
For the reasons given above and in the rejections below, the application is not in consideration for allowance and all rejections are maintained.
Double Patenting — Nonstatutory
The nonstatutory double patenting rejection is based on a judicially created doctrine grounded in public policy (a policy reflected in the statute) so as to prevent the unjustified or improper timewise extension of the “right to exclude” granted by a patent and to prevent possible harassment by multiple assignees. A nonstatutory double patenting rejection is appropriate where the conflicting claims are not identical, but at least one examined application claim is not patentably distinct from the reference claim(s) because the examined application claim is either anticipated by, or would have been obvious over, the reference claim(s). See, e.g., In re Berg, 140 F.3d 1428, 46 USPQ2d 1226 (Fed. Cir. 1998); In re Goodman, 11 F.3d 1046, 29 USPQ2d 2010 (Fed. Cir. 1993); In re Longi, 759 F.2d 887, 225 USPQ 645 (Fed. Cir. 1985); In re Van Ornum, 686 F.2d 937, 214 USPQ 761 (CCPA 1982); In re Vogel, 422 F.2d 438, 164 USPQ 619 (CCPA 1970); In re Thorington, 418 F.2d 528, 163 USPQ 644 (CCPA 1969).
A timely filed terminal disclaimer in compliance with 37 CFR 1.321(c) or 1.321(d) may be used to overcome an actual or provisional rejection based on nonstatutory double patenting provided the reference application or patent either is shown to be commonly owned with the examined application, or claims an invention made as a result of activities undertaken within the scope of a joint research agreement. See MPEP § 717.02 for applications subject to examination under the first inventor to file provisions of the AIA as explained in MPEP § 2159. See MPEP § 2146 et seq. for applications not subject to examination under the first inventor to file provisions of the AIA . A terminal disclaimer must be signed in compliance with 37 CFR 1.321(b).
The filing of a terminal disclaimer by itself is not a complete reply to a nonstatutory double patenting (NSDP) rejection. A complete reply requires that the terminal disclaimer be accompanied by a reply requesting reconsideration of the prior Office action. Even where the NSDP rejection is provisional the reply must be complete. See MPEP § 804, subsection I.B.1. For a reply to a non-final Office action, see 37 CFR 1.111(a). For a reply to final Office action, see 37 CFR 1.113(c). A request for reconsideration while not provided for in 37 CFR 1.113(c) may be filed after final for consideration. See MPEP §§ 706.07(e) and 714.13.
The USPTO Internet website contains terminal disclaimer forms which may be used. Please visit www.uspto.gov/patent/patents-forms. The actual filing date of the application in which the form is filed determines what form (e.g., PTO/SB/25, PTO/SB/26, PTO/AIA /25, or PTO/AIA /26) should be used. A web-based eTerminal Disclaimer may be filled out completely online using web-screens. An eTerminal Disclaimer that meets all requirements is auto-processed and approved immediately upon submission. For more information about eTerminal Disclaimers, refer to www.uspto.gov/patents/apply/applying-online/eterminal-disclaimer.
Regarding Claims 1
Claim 1 is provisionally rejected on the ground of nonstatutory double patenting as being unpatentable over claim 4 (dependent on claim 1) of copending Application No. 17368407 in view of PGPUB no. US 20200218974 A1, herein referred to as Cheng. Although the claims at issue are not identical, they highly interlap in scope and are directed to the same subject matter. The differences between the relevant claim limitations are shown below
17368404 claim 1 (as amended)
17368407 claim 4 and independent claim 1 (as amended)
A computer-implemented method comprising:
receiving, by one or more processors, a user feature data object comprising one or more numerical timeseries feature data fields, one or more categorical timeseries feature data fields, and one or more static feature data fields;
generating, by the one or more processors and by processing the one or more numerical timeseries feature data fields using a first recurrent neural network (RNN) framework of a fall prediction machine learning model, a numerical timeseries embedding for the user feature data object,
wherein the fall prediction machine learning model is trained based at least in part on distillation loss;
generating, by the one or more processors and by processing the one or more categorical timeseries feature data fields using a second RNN framework of the fall prediction machine learning model, a categorical timeseries embedding for the user feature data object;
generating, by the one or more processors and by processing the one or more static feature data fields using a fully connected neural network framework of the fall prediction machine learning model, a static embedding for the user feature data object;
generating, by the one or more processors and by processing the numerical timeseries embedding, the categorical timeseries embedding, and the static embedding using an ensemble machine learning framework of the fall prediction machine learning model,
a fall prediction data object comprising
(i) a fall likelihood prediction corresponding to a predicted fall event,
and
(ii) in an instance where the fall likelihood prediction is an affirmative likelihood prediction, one or more current medical conditions of a user indicated by the numerical timeseries embedding, the categorical timeseries embedding, or the static embedding predicted to cause a fall
generating, by the one or more processors and based on the fall prediction data object, one or more sensory notifications indicating at least one corrective action, the at least one corrective action configured to prevent or lessen a severity of the predicted fall event;
and providing, by the one or more processors and to a client device prior to occurrence of the predicted fall event, the one or more sensory notifications.
Claim 1:
A computer-implemented method comprising:
Claim 4:
further comprising generating a vector representation that is associated with a numerical timeseries data field, a categorical timeseries feature data field, and a static feature data field
the plurality of machine learning frameworks comprises: (i) a first recurrent neural network (RNN) framework that is configured to process the numerical timeseries feature data field to generate a numerical timeseries embedding of the vector representation
Claim 1:
generating, by one or more processors and based on a fall prediction data object generated by a prediction machine learning model
wherein […] (3) the prediction machine learning model is trained based at least in part on a distillation loss score, wherein the distillation loss score is generated by
Claim 4
, (ii) a second RNN framework that is configured to process the categorical timeseries feature data field to generate a categorical timeseries embedding of the vector representation ,
(iii) a fully connected neural network framework that is configured to process the static feature data field to generate a static embedding of the vector representation ,
and (iv) an ensemble machine learning framework that is configured to generate the inferred prediction based at least in part on the numerical timeseries embedding, the categorical timeseries embedding, and the static embedding.
Claim 1
generating, (i) an inferred likelihood indication corresponding to the user feature data object and
(ii) a plurality of inferred cause indication corresponding to the inferred likelihood indication, wherein the plurality of inferred cause indications is generated by a plurality of machine learning frameworks
generating, by one or more processors and based on a fall prediction data object generated by a prediction machine learning model, one or more sensory notifications
providing, by the one or more processors and to a client device and prior to an occurrence of the predicted future fall event, the one or more sensory notifications.
Thus claim 4 of the reference application overlaps with many of the limitations of claim 1 of the instant application.
Claim 4 of the reference does not teach:
“in an instance where the fall likelihood prediction is an affirmative likelihood prediction, one or more current medical conditions of a user indicated by the numerical timeseries embedding, the categorical timeseries embedding, or the static embedding predicted to cause a fall”
“one or more sensory notifications indicating at least one corrective action, the at least one corrective action configured to prevent or lessen a severity of the predicted fall event”
However, Rakshit teaches:
in an instance where the fall likelihood prediction is an affirmative likelihood prediction, one or more current medical conditions of a user indicated by the numerical timeseries embedding, the categorical timeseries embedding, or the static embedding predicted to cause a fall
(paragraph [0039]) “In this way, safety manager 110 can identify or otherwise predict a reason for the detected event (e.g., low blood pressure occurred during a fall[*Examiner notes: medical condition], slipping on a surface, walking down the stairs, accidents, etc.).”
one or more sensory notifications indicating at least one corrective action, the at least one corrective action configured to prevent or lessen a severity of the predicted fall event
(paragraph [0023]) “In certain embodiments, safety manager 110 can generate a text report to accompany the visual simulation that notes any predicted change with respect to the cause of the event, an appropriate notification of the detected event, and suggestions to prevent the detected event in the future. For example, safety manager 110 can recommend additional actions a user can take to increase safety (e.g., personal protective equipment, protocols, etc.).”
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the present invention to modify the machine learning model of 17368407 claim 4 with the indication of medical conditions and fall prevention action as taught by Rakshit because (Rakshit paragraph [0039]) “In this way, safety manager 110 can identify or otherwise predict a reason for the detected event (e.g., low blood pressure occurred during a fall, slipping on a surface, walking down the stairs, accidents, etc.). In this embodiment, safety manager 110 can feed the accessed information to one or more machine learning algorithms an improve its analysis through historical learning” and (Rakshit paragraph [0023]) “For example, safety manager 110 can recommend additional actions a user can take to increase safety (e.g., personal protective equipment, protocols, etc.). In certain other embodiments, safety manager 110 can, with user permission, monitor safety protocols and transmit reminders and notifications to users to follow safety protocols.”
This is a provisional nonstatutory double patenting rejection.
Regarding Claim 2:
Claim 2 is provisionally rejected on the ground of nonstatutory double patenting as being unpatentable over claim 3 of copending Application No. 17368407 (reference application). Although the claims at issue are not identical, they highly interlap in scope and are directed to the same subject matter. The differences between the relevant claim limitations are shown below
17368404 claim 2
17368407 claim 3
The computer-implemented method of claim 1, wherein the fall prediction data object further describes, in the instance where the fall likelihood prediction is the affirmative likelihood prediction, a fall timing prediction.
Claim 3
The computer-implemented method of claim 1, wherein the inferred likelihood indication comprises a timing prediction.
Claim 1
generating, by one or more processors and based on a fall prediction data object generated by a prediction machine learning model
Thus claim 3 of the reference application teaches nearly all of the limitations of claim 2 of the instant application. Claim 3 of the reference does not teach “in the instance where the fall likelihood prediction is the affirmative likelihood prediction,”.
However, Cheng teaches in the instance where the fall likelihood prediction is the affirmative likelihood prediction, (paragraph [0213]) “An imminent fall alert[*Examiner notes: mapped to fall timing prediction] may be communicated to the subject through a noticeable alert, such as an audio alarm, tactile stimulus or other output that is recognizable as urgent and simple to interpret, and providing the subject some warning that the system 610 has assessed a high likelihood of the subject falling.”
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the present invention to modify the prediction method of 17368407 claim 4 with the fall prediction and current user conditions of Cheng because (Cheng paragraph [0003]) “Similarly, careful monitoring of activity on the worksite can be important in preventing workplace-related injuries. However, even with the most precise of these available devices, distinguishing between certain types of activity can be very difficult” and (Cheng paragraph [0003]) “Herein provided are a system and method for classifying physical activity of a subject.”
This is a provisional nonstatutory double patenting rejection.
Regarding Claim 5
Claim 5 is provisionally rejected on the ground of nonstatutory double patenting as being unpatentable over claim 1 of copending Application No. 17368407 (reference application). Although the claims at issue are not identical, they highly interlap in scope and are directed to the same subject matter.
17368404 claim 5
17368407 claim 1
The computer-implemented method of claim 1, wherein the fall prediction machine learning model is trained based at least in part on a custom loss generated by a custom loss model, and the custom loss model comprises a fall likelihood component and a fall cause component.
generating, using a custom loss model, (i) a likelihood loss value based at least in part on the ground-truth likelihood indication and the inferred likelihood indication and (ii) a plurality of cause loss value based at least in part on the plurality of ground-truth cause indication and the inferred cause indication
This is a provisional nonstatutory double patenting rejection.
Regarding Claim 6
Claim 6 is provisionally rejected on the ground of nonstatutory double patenting as being unpatentable over claim 5 of copending Application No. 17368407 (reference application). Although the claims at issue are not identical, they highly interlap in scope and are directed to the same subject matter.
17368404 claim 6
17368407 claim 5
The computer-implemented method of claim 1, further comprising transmitting a fall prediction notification describing the fall prediction data object to an edge client computing entity, wherein the edge client computing entity is configured to present the one or more sensory notifications to an end user of the edge client computing entity.
The computer-implemented method of claim 1 further comprising transmitting, to an edge client computing entity, a prediction notification, wherein the edge client computing entity is configured to present a sensory notification to an end user of the edge client computing entity based at least in part on the prediction notification.
This is a provisional nonstatutory double patenting rejection.
Regarding Claim 7
Claim 7 is provisionally rejected on the ground of nonstatutory double patenting as being unpatentable over claim 6 of copending Application No. 17368407 (reference application). Although the claims at issue are not identical, they highly interlap in scope and are directed to the same subject matter.
17368404 claim 7
17368407 claim 6
The computer-implemented method of claim 1, wherein the one or more sensory notifications comprise at least one of: (i) one or more audiovisual notifications, (ii) one or more haptic notifications, and (iii) one or more electrical pulse notifications.
The computer-implemented method of claim 5, wherein the sensory notification comprises at least one of: (i) an audiovisual notification, or (ii)an electrical pulse notification.
This is a provisional nonstatutory double patenting rejection.
Regarding Claims 8 and 15
Claims 8 and 15 of the instant application are substantially the same as claim 1 of the instant application, and claims 11 and 18 of the reference application are substantially the same as claim 4 of the reference application. The only differences are that claim 8 of the instant application and claim 11 of the reference application each recite a non-transitory computer-readable medium, and claim 15 of the instant application and claim 18 of the reference application each recite a computer system with a processor and memory. See the comparisons below.
17368404 Claim 8
17368407 Claim 11 (dependent on claim 8)
A system comprising one or more processors and at least one memory storing processor- executable instructions that, when executed by any one or more of the one or more processors, cause the one or more processors to perform operations comprising:
Claim 8:
A system comprising memory and one or more processors communicatively coupled to the memory, the one or more processors configured to
17368404 Claim 15
17368407 Claim 18 (dependent on claim 15)
One or more non-transitory computer-readable storage media storing processor- executable instructions that, when executed by one or more processors, cause the one or more processors to perform operations comprising:
One or more non-transitory computer-readable storage media including instructions that, when executed by one or more processors, cause the one or more processors to
Therefore, claims 8 and 15 are provisionally rejected on the ground of non-statutory double patenting over claims 11 and 18 respectively of the reference application for the same reasons as claim 1 above.
This is a provisional nonstatutory double patenting rejection.
Regarding Claims 9 and 16
Claims 9 and 16 of the instant application are substantially the same as claim 2 of the instant application, and claims 10 and 17 of the reference application are substantially the same as claim 3 of the reference application. Therefore, claims 9 and 16 are provisionally rejected on the ground of non-statutory double patenting over claims 10 and 17 respectively for the same reasons as claim 2 above.
This is a provisional nonstatutory double patenting rejection.
Regarding Claims 12 and 19
Claims 12 and 19 of the instant application are substantially the same as claim 5 of the instant application, and claims 8 and 15 of the reference application are substantially the same as claim 1 of the reference application. Therefore, claims 12 and 19 are provisionally rejected on the ground of non-statutory double patenting over claims 8 and 15 respectively of the reference application for the same reasons as claim 5 above.
This is a provisional nonstatutory double patenting rejection.
Regarding Claim 13
Claim 13 of the instant application is substantially the same as claim 6 of the instant application, and claim 12 of the reference application is substantially the same as claim 5 of the reference application. Therefore, claim 13 is provisionally rejected on the ground of non-statutory double patenting over claim 12 of the reference application for the same reasons as claim 6 above.
This is a provisional nonstatutory double patenting rejection.
Regarding Claim 14
Claims 14 of the instant application is substantially the same as claim 7 of the instant application, and claim 13 of the reference application is substantially the same as claim 6 of the reference application. Therefore, claim 14 is provisionally rejected on the ground of non-statutory double patenting over claim 13 of the reference application for the same reasons as claim 7 above.
This is a provisional nonstatutory double patenting rejection.
Regarding Claim 20
Claim 20 is provisionally rejected on the ground of nonstatutory double patenting as being unpatentable over claim of copending Application No. 17368407 (reference application). Although the claims at issue are not identical, they highly interlap in scope and are directed to the same subject matter.
17368404 claim 20
17368407 claim 19
The one or more non-transitory computer-readable storage media of claim 15, wherein the operations further comprise transmitting a fall prediction notification describing the fall prediction data object to an edge client computing entity, wherein the edge client computing entity is configured to present the one or more sensory notifications to an end user of the edge client computing entity based at least in part on the fall prediction notification.
The one or more non-transitory computer-readable storage media of claim 15 wherein the operations further comprise transmitting, to an edge client computing entity, a prediction notification , wherein the edge client computing entity is configured to present a sensory notification to an end user of the edge client computing entity based at least in part on the prediction notification.
This is a provisional nonstatutory double patenting rejection.
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 for containing an abstract idea without significantly more.
Regarding Claim 1:
Step 1 – Is the claim to a process, machine, manufacture, or composition of matter?
Yes, the claim is to a process.
Step 2A – Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
Yes, the claim recites the abstract ideas of:
generating, […] by processing the one or more numerical timeseries feature data fields using a first recurrent neural network (RNN) framework of a fall prediction machine learning model, a numerical timeseries embedding for the user feature data object —This limitation is directed to the abstract idea of a mental process (including an observation, evaluation, judgement, opinion) which can be performed by the human mind, or by a human using pen and paper (see MPEP 2106.04(a)(2) III. C.). This limitation is directed to a mental process because it amounts to making predictions with a neural network which can be performed by, for example, evaluation of a series of matrix multiplications and activation functions (e.g. ReLU).
generating, […] by processing the one or more categorical timeseries feature data fields using a second RNN framework of the fall prediction machine learning model, a categorical timeseries embedding for the user feature data object — This limitation is directed to the abstract idea of a mental process (including an observation, evaluation, judgement, opinion) which can be performed by the human mind, or by a human using pen and paper (see MPEP 2106.04(a)(2) III. C.). This limitation is directed to a mental process because it amounts to making predictions with a neural network which can be performed by, for example, evaluation of a series of matrix multiplications and activation functions (e.g. ReLU).
generating […] by processing the one or more static feature data fields using a fully connected neural network framework of the fall prediction machine learning model, a static embedding for the user feature data object — This limitation is directed to the abstract idea of a mental process (including an observation, evaluation, judgement, opinion) which can be performed by the human mind, or by a human using pen and paper (see MPEP 2106.04(a)(2) III. C.). This limitation is directed to a mental process because it amounts to making predictions with a neural network which can be performed by, for example, evaluation of a series of matrix multiplications and activation functions (e.g. ReLU).
generating, […] by processing the numerical timeseries embedding, the categorical timeseries embedding, and the static embedding using an ensemble machine learning framework of the fall prediction machine learning model, a fall prediction data object comprising (i) a fall likelihood prediction corresponding to a predicted fall event, and (ii) in an instance where the fall likelihood prediction is an affirmative likelihood prediction, one or more current medical conditions of a user indicated by the numerical timeseries embedding, the categorical timeseries embedding, or the static embedding predicted to cause a fall — This limitation is directed to the abstract idea of a mental process (including an observation, evaluation, judgement, opinion) which can be performed by the human mind, or by a human using pen and paper (see MPEP 2106.04(a)(2) III. C.). This limitation is directed to a mental process because it amounts to making predictions with a neural network which can be performed by, for example, evaluation of a series of matrix multiplications and activation functions (e.g. ReLU).
Step 2A – Prong 2 – Does the claim recite additional elements that integrate the judicial exception into a practical application?
No, the claim does not recite additional elements that integrate the judicial exception into a practical application. The additional elements:
receiving, […], a user feature data object comprising one or more numerical timeseries feature data fields, one or more categorical timeseries feature data fields, and one or more static feature data fields — This limitation is directed to mere data gathering and outputting which has been recognized by the courts (as per Ultramercial, 772 F.3d at 715, 112 USPQ2d at 1754) as insignificant extra-solution activity (see MPEP 2106.05(g)).
by the one or more processors — This limitation is directed to merely applying an abstract idea using a generic computer as a tool (see MPEP 2106.05(f)(2), 2106.04(d)).
wherein the fall prediction machine learning model is trained based at least in part on distillation loss — This limitation is directed to mere instructions to apply a judicial exception. Using machine learning training based on distillation loss to apply a judicial exception (see MPEP 2106.05(f)) is insufficient to integrate the judicial exception into a practical application. Even if the machine learning training based on distillation loss is implemented on a generic computer (see MPEP 2106.05(f)(2), 2106.04(d)), the limitation does not integrate the judicial exception into a practical application.
generating, […] and based on the fall prediction data object, one or more sensory notifications indicating at least one corrective actions — This limitation is directed to mere data gathering and outputting which has been recognized by the courts (as per Ultramercial, 772 F.3d at 715, 112 USPQ2d at 1754) as insignificant extra-solution activity (see MPEP 2106.05(g)).
and providing, […] to a client device prior to occurrence of the predicted fall event, the one or more sensory notifications — This limitation is directed to mere data gathering and outputting which has been recognized by the courts (as per Ultramercial, 772 F.3d at 715, 112 USPQ2d at 1754) as insignificant extra-solution activity (see MPEP 2106.05(g)).
Step 2B – Does the claim recite additional elements that amount to significantly more than the abstract idea itself?
No, the claim does not recite additional elements which amount to significantly more than the abstract idea itself. The additional elements as identified in step 2A prong 2:
receiving, […], a user feature data object comprising one or more numerical timeseries feature data fields, one or more categorical timeseries feature data fields, and one or more static feature data fields — This limitation is recited at a high level of generality and amounts to mere data gathering of storing and retrieving information in memory, which is well-understood, routine, and conventional activity (see MPEP 2106.05(d) II.), which cannot amount to significantly more than the judicial exception.
by the one or more processors — Using a generic computer as a tool (see MPEP 2106.05(f)(2), 2106.05(d)) cannot amount to significantly more than the judicial exception itself.
wherein the fall prediction machine learning model is trained based at least in part on distillation loss — Mere instructions to apply a judicial exception (see MPEP 2106.05(f)) and using a generic computer as a tool (see MPEP 2106.05(f)(2), 2106.05(d)) cannot amount to significantly more than the judicial exception itself.
generating, […] and based on the fall prediction data object, one or more sensory notifications indicating at least one corrective action, the at least one corrective action configured to prevent or lessen the severity of the predicted fall event — This limitation is recited at a high level of generality and amounts to mere data gathering of transmitting and receiving data over a network, which is well-understood, routine, and conventional activity (see MPEP 2106.05(d) II.), which cannot amount to significantly more than the judicial exception.
and providing, […] to a client device prior to occurrence of the predicted fall event, the one or more sensory notifications — This limitation is recited at a high level of generality and amounts to mere data gathering of transmitting and receiving data over a network, which is well-understood, routine, and conventional activity (see MPEP 2106.05(d) II.), which cannot amount to significantly more than the judicial exception.
Regarding Claim 2
Claim 2 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claim is dependent on claim 1 which included an abstract idea (see rejection for claim 1). The claim recites the additional limitations:
Step 2A Prong 2:
wherein the fall prediction data object further describes, in the instance where the fall likelihood prediction is the affirmative likelihood prediction, a fall timing prediction — This limitation is directed to merely limiting a judicial exception to a particular field of use (see MPEP 2106.05(h)) as it merely limits the field of the fall prediction data object.
Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2.
Step 2B:
The additional elements as identified in step 2A prong 2:
wherein the fall prediction data object further describes, in the instance where the fall likelihood prediction is the affirmative likelihood prediction, a fall timing prediction — Merely limiting a judicial exception to a particular field of use (see MPEP 2106.05(h)) cannot amount to significantly more than the judicial exception.
Thus, the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B.
Regarding Claim 3
Claim 3 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claim is dependent on claim 1 which included an abstract idea (see rejection for claim 1). The claim recites the additional limitations:
Step 2A Prong 2:
wherein: the second RNN framework comprise a long short term memory RNN framework.— This limitation is directed to merely limiting a judicial exception to a particular field of use (see MPEP 2106.05(h)) as it merely limits the field of the second RNN framework.
Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2.
Step 2B:
The additional elements as identified in step 2A prong 2:
wherein: the second RNN framework comprise a long short term memory RNN framework — Merely limiting a judicial exception to a particular field of use (see MPEP 2106.05(h)) cannot amount to significantly more than the judicial exception.
Thus, the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B.
Regarding Claim 4
Claim 4 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claim is dependent on claim 1 which included an abstract idea (see rejection for claim 1). The claim recites the additional limitations:
Step 2A Prong 2:
wherein: the fall prediction machine learning model has fewer parameters as compared to a trained teacher fall prediction machine learning model — This limitation is directed to merely limiting a judicial exception to a particular field of use (see MPEP 2106.05(h)) as it merely limits the field of the fall prediction machine learning model.
and wherein the distillation loss comprises a custom loss generated by a custom loss model and a distillation loss score, and wherein the distillation loss score is based at least in part on one or more teacher outputs from a trained teacher fall prediction machine learning model, one or more inferred outputs of the fall prediction machine learning model, a ground truth fall likelihood and a ground truth fall cause — This limitation is directed to merely limiting a judicial exception to a particular field of use (see MPEP 2106.05(h)) as it merely limits the field of the distillation loss.
Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2.
Step 2B:
The additional elements as identified in step 2A prong 2:
wherein: the fall prediction machine learning model has fewer parameters as compared to a trained teacher fall prediction machine learning model — Merely limiting a judicial exception to a particular field of use (see MPEP 2106.05(h)) cannot amount to significantly more than the judicial exception.
and wherein the distillation loss comprises a custom loss generated by a custom loss model and a distillation loss score, and wherein the distillation loss score is based at least in part on one or more teacher outputs from a trained teacher fall prediction machine learning model, one or more inferred outputs of the fall prediction machine learning model, a ground truth fall likelihood and a ground truth fall cause — Merely limiting a judicial exception to a particular field of use (see MPEP 2106.05(h)) cannot amount to significantly more than the judicial exception.
Thus, the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B.
Regarding Claim 5
Claim 5 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claim is dependent on claim 1 which included an abstract idea (see rejection for claim 1). The claim recites the additional limitations:
Step 2A Prong 2:
wherein the fall prediction machine learning model is trained based at least in part on a custom loss generated by a custom loss model, and the custom loss model comprises a fall likelihood component and a fall cause component — This limitation is directed to mere instructions to apply a judicial exception. Using machine learning model training based on loss to apply a judicial exception (see MPEP 2106.05(f)) is insufficient to integrate the judicial exception into a practical application. Even if the machine learning model training based on loss is implemented on a generic computer (see MPEP 2106.05(f)(2), 2106.04(d)), the limitation does not integrate the judicial exception into a practical application.
Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2.
Step 2B:
The additional elements as identified in step 2A prong 2:
wherein the fall prediction machine learning model is trained based at least in part on a custom loss generated by a custom loss model, and the custom loss model comprises a fall likelihood component and a fall cause component — Mere instructions to apply a judicial exception (see MPEP 2106.05(f)) and using a generic computer as a tool (see MPEP 2106.05(f)(2), 2106.05(d)) cannot amount to significantly more than the judicial exception itself.
Thus, the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B.
Regarding Claim 6
Claim 6 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claim is dependent on claim 1 which included an abstract idea (see rejection for claim 1). The claim recites the additional limitations:
Step 2A Prong 2:
further comprising transmitting a fall prediction notification describing the fall prediction data object to an edge client computing entity wherein the edge client computing entity is configured to present one or more sensory notifications to an end user of the edge client computing entity — This limitation is directed to mere data gathering and outputting which has been recognized by the courts (as per Ultramercial, 772 F.3d at 715, 112 USPQ2d at 1754) as insignificant extra-solution activity (see MPEP 2106.05(g)).
Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2.
Step 2B:
The additional elements as identified in step 2A prong 2:
further comprising transmitting a fall prediction notification describing the fall prediction data object to an edge client computing entity wherein the edge client computing entity is configured to present one or more sensory notifications to an end user of the edge client computing entity — This limitation is recited at a high level of generality and amounts to mere data gathering of transmitting and receiving data over a network, which is well-understood, routine, and conventional activity (see MPEP 2106.05(d) II.), which cannot amount to significantly more than the judicial exception.
Thus, the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B.
Regarding Claim 7
Claim 7 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claim is dependent on claim 1 which included an abstract idea (see rejection for claim 1). The claim recites the additional limitations:
Step 2A Prong 2:
wherein the one or more sensory notifications comprise at least one of: (i) one or more audiovisual notifications, (ii) one or more haptic notifications, and (iii) one or more electrical pulse notifications — This limitation is directed to merely limiting a judicial exception to a particular field of use (see MPEP 2106.05(h)) as it merely limits the field of the sensory notifications.
Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2.
Step 2B:
The additional elements as identified in step 2A prong 2:
wherein the one or more sensory notifications comprise at least one of: (i) one or more audiovisual notifications, (ii) one or more haptic notifications, and (iii) one or more electrical pulse notifications — Merely limiting a judicial exception to a particular field of use (see MPEP 2106.05(h)) cannot amount to significantly more than the judicial exception.
Thus, the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B.
Regarding Claim 8
Independent claim 8 is a computer system claim corresponding to method claim 1, which was directed to an abstract idea, therefore the same rejection and rationale applies. The only difference is that claim 8 recites the following additional elements treated under step 2A prong 2 and step 2B:
Step 2A Prong 2:
A system comprising one or more processors and at least one memory storing processor-executable instructions that, when executed by any one or more of the one or more processors, cause the one or more processors to perform operations comprising — This limitation is directed to merely applying an abstract idea using a generic computer as a tool (see MPEP 2106.05(f)(2), 2106.04(d)).
Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2.
Step 2B:
A system comprising one or more processors and at least one memory storing processor-executable instructions that, when executed by any one or more of the one or more processors, cause the one or more processors to perform operations comprising — Using a generic computer as a tool (see MPEP 2106.05(f)(2), 2106.05(d)) cannot amount to significantly more than the judicial exception itself.
Thus, the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B.
Regarding Claim 9
Dependent claim 9 is a system claim corresponding to method claim 2, which was directed to an abstract idea, therefore the same rejection and rationale applies.
Regarding Claim 10
Dependent claim 10 is a system claim corresponding to method claim 3, which was directed to an abstract idea, therefore the same rejection and rationale applies.
Regarding Claim 11
Dependent claim 11 is a system claim corresponding to method claim 3, which was directed to an abstract idea, therefore the same rejection and rationale applies.
Regarding Claim 12
Dependent claim 12 is a system claim corresponding to method claim 5, which was directed to an abstract idea, therefore the same rejection and rationale applies.
Regarding Claim 13
Dependent claim 13 is a system claim corresponding to method claim 6, which was directed to an abstract idea, therefore the same rejection and rationale applies.
Regarding Claim 14
Dependent claim 14 is a system claim corresponding to method claim 7, which was directed to an abstract idea, therefore the same rejection and rationale applies.
Regarding Claim 15
Independent claim 15 is a non-transitory computer-readable medium claim corresponding to method claim 1, which was directed to an abstract idea, therefore the same rejection and rationale applies. The only difference is that claim 15 recites the following additional elements treated under step 2A prong 2 and step 2B:
Step 2A Prong 2:
One or more non-transitory computer-readable storage media storing processor- executable instructions that, when executed by one or more processors, cause the one or more processors to perform operations — This limitation is directed to merely applying an abstract idea using a generic computer as a tool (see MPEP 2106.05(f)(2), 2106.04(d)).
Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2.
Step 2B:
One or more non-transitory computer-readable storage media storing processor- executable instructions that, when executed by one or more processors, cause the one or more processors to perform operations — Using a generic computer as a tool (see MPEP 2106.05(f)(2), 2106.05(d)) cannot amount to significantly more than the judicial exception itself.
Thus, the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B.
Regarding Claim 16
Dependent claim 16 is a non-transitory computer-readable medium claim corresponding to method claim 2, which was directed to an abstract idea, therefore the same rejection and rationale applies.
Regarding Claim 17
Dependent claim 17 is a non-transitory computer-readable medium claim corresponding to method claim 3, which was directed to an abstract idea, therefore the same rejection and rationale applies.
Regarding Claim 18
Dependent claim 18 is a non-transitory computer-readable medium claim corresponding to method claim 4, which was directed to an abstract idea, therefore the same rejection and rationale applies.
Regarding Claim 19
Dependent claim 19 is a non-transitory computer-readable medium claim corresponding to method claim 5, which was directed to an abstract idea, therefore the same rejection and rationale applies.
Regarding Claim 20
Claim 20 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claim is dependent on claim 15 which included an abstract idea (see rejection for claim 15). The claim recites the additional limitations:
Step 2A Prong 2:
wherein the operations further comprise transmitting a fall prediction notification describing the fall prediction data object to an edge client computing entity, wherein the edge client computing entity is configured to present one or more sensory notifications to an end user of the edge client computing entity based at least in part on the fall prediction notification — This limitation is directed to mere data gathering and outputting which has been recognized by the courts (as per Ultramercial, 772 F.3d at 715, 112 USPQ2d at 1754) as insignificant extra-solution activity (see MPEP 2106.05(g)).
Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2.
Step 2B:
The additional elements as identified in step 2A prong 2:
wherein the operations further comprise transmitting a fall prediction notification describing the fall prediction data object to an edge client computing entity, wherein the edge client computing entity is configured to present one or more sensory notifications to an end user of the edge client computing entity based at least in part on the fall prediction notification — This limitation is recited at a high level of generality and amounts to mere data gathering of transmitting and receiving data over a network, which is well-understood, routine, and conventional activity (see MPEP 2106.05(d) II.), which cannot amount to significantly more than the judicial exception.
Thus, the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B.
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.
Claims 1-4, 6-11, and 13-18, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Cheng et al. (PGPUB no. US 20200218974 A1) herein referred to as Cheng in view of NPL reference Jang et al. “Developing neural network models for early detection of cardiac arrest in emergency department” herein referred to as Jang, Wu et al. “MULTI-TEACHER KNOWLEDGE DISTILLATION FOR COMPRESSED VIDEO ACTION RECOGNITION ON DEEP NEURAL NETWORKS” herein referred to as Wu, and Rakshit (PGPUB no. US20220392369A1) herein referred to as Rakshit.
Regarding Claim 1
Cheng teaches:
A computer-implemented method comprising:
(paragraph [0162]) “The first device 17 in the system 10 as shown in FIG. 2 is exemplified by a smartphone, but any suitable mobile or other device may be applied (e.g. smartphone, smartwatch, tablet, laptop computer, desktop computer, cloud-based server etc.).”
receiving, by one or more processors, a user feature data object
(Paragraph [0210]) "The training data 628[*Examiner notes: mapped to user feature data object] is received by the processor 618[*Examiner notes: mapped to processor] with the input data 620 and may be provided to the NN 690 to train the NN 690”
comprising […] and one or more static feature data fields
(paragraph [0006]) "Input data including pressure data is collected from sensors during activities performed by the subject."; (paragraph [0007]) "The NN may apply activity classification across a specified window of time of the input data[*Examiner notes: mapped to static feature data field]. The window of time may be the known and static width of an event identified by a user"; [*Examiner note: Although the neural network is applied across multiple timesteps, the data at each timestep is treated as static data by Cheng.]
generating, by the one or more processors and by processing the one or more static feature data fields using a fully connected neural network framework of the fall prediction machine learning model, a static embedding for the user feature data object
(paragraph [0007]) “The NN may apply activity classification across an entire data set of the input data. The NN may apply activity classification across a specified window of time of the input data. The window of time may be the known and static width of an event identified by a user[*Examiner notes: mapped to the static feature data fields], who may be the subject or a different individual, within the input data, in which the event is a distinct waveform within the input data and delineated using known event detection methods. The window of time may be of a defined time range that is progressed through the input data and the input data within the time window is classified before translating the time window along a timeline of the input data to apply the NN[*Examiner notes: mapped to fully connected neural network framework] to a subsequent portion of the input data within the time window after translation[*Examiner notes: mapped to the static feature data fields].”; (paragraph [0007]) “The input data may also be classified through an event window in which an event is detected in the input data, event limits defined on either side of the event, and then the input data within the event window may be classified. The classified activity data[*Examiner notes: mapped to static feature embedding] may be communicated to a user by display of the classified activity or by storage of the classified activity data.”; (paragraph [0009]) "The NN may be a convolutional neural network (“CNN”). The CNN includes at least one convolutional layer and at least one fully connected layer[*Examiner notes: mapped to fully connected neural network framework]."; [*Examiner notes: The output of the neural network (i.e. the classified activity) is the embedding.]
and providing, by the one or more processors and to a client device prior to occurrence of the predicted fall event, the one or more sensory notifications
(paragraph [0208]) “The communication module 642 may also communicate the classified activity data 638 to the subject through an alert output 699[*Examiner notes: mapped to generating one or more sensory notifications], which may be located on a garment 698[*Examiner notes: mapped to client device], in the event that the classified activity data 638 is indicative of an imminent fall[*Examiner notes: mapped to based on the fall prediction data object, prior to occurrence]."; figure 16 no. 698 and 699; [*Examiner notes: The sensory notification is indicative of an imminent fall, not a fall that has already happened or is happening]
PNG
media_image1.png
370
700
media_image1.png
Greyscale
Cheng does not explicitly teach:
one or more numerical timeseries feature data fields, one or more categorical timeseries feature data fields
generating, by the one or more processors and by processing the one or more numerical timeseries feature data fields using a first recurrent neural network (RNN) framework of a fall prediction machine learning model, a numerical timeseries embedding for the user feature data object,
wherein the fall prediction machine learning model is trained based at least in part on distillation loss;
generating, by the one or more processors and by processing the one or more categorical timeseries feature data fields using a second RNN framework of the fall prediction machine learning model, a categorical timeseries embedding for the user feature data object
by processing the numerical timeseries embedding, the categorical timeseries embedding, and the static embedding using an ensemble machine learning framework of the fall prediction machine learning model
(ii) in an instance where the fall likelihood prediction is an affirmative likelihood prediction, one or more current medical conditions of a user indicated by the numerical timeseries embedding, the categorical timeseries embedding, or the static embedding predicted to cause a fall
generating, by the one or more processors and based on the fall prediction data object, one or more sensory notifications indicating at least one corrective action, the at least one corrective action configured to prevent or lessen a severity of the predicted fall event
However, Jang Teaches:
one or more numerical timeseries feature data fields, one or more categorical timeseries feature data fields
(page 44, column 1, last paragraph) “The updatable variables include vital signs and level of consciousness measured within 6 h of each prediction. The ‘update’ on the current state is only made when there is a new measurement."; (page 44, column 1, second to last paragraph
) "The following variables were used to predict an event of cardiac arrest: demographic information (age and sex), chief complaints, vital signs (systolic blood pressure [SBP], diastolic blood pressure [DBP], heart rate [HR], respiratory rate [RR], and body temperature [BT])[mapped to numerical timeseries data], and level of consciousness (measured according to the AVPU scale)[mapped to categorical timeseries data]"
generating, by the one or more processors and by processing the one or more numerical timeseries feature data fields using a first recurrent neural network (RNN) framework of a fall prediction machine learning model, a numerical timeseries embedding for the user feature data object,
(page 44 column 2 paragraph 3) “The third one is ‘the hybrid’ model, in which the baseline and sequence data are processed[*Examiner notes: mapped to numerical timeseries embedding] separately first, and then fused to predict a single outcome. Specifically, baseline variables are processed within the MLP architecture while sequence data are processed within the LSTM architecture[*Examiner notes: mapped to first RNN framework]"; figure 1(c)
generating, by the one or more processors and by processing the one or more categorical timeseries feature data fields using a second RNN framework of the fall prediction machine learning model, a categorical timeseries embedding for the user feature data object
(page 44 column 2 paragraph 3) "The third one is ‘the hybrid’ model, in which the baseline and sequence data are processed separately[*Examiner notes: mapped to categorical timeseries embedding] first, and then fused to predict a single outcome. Specifically, baseline variables are processed within the MLP architecture while sequence data are processed within the LSTM architecture[*Examiner notes: mapped to second RNN framework]"; figure 1(c)
by processing the numerical timeseries embedding, the categorical timeseries embedding, and the static embedding using an ensemble machine learning framework of the fall prediction machine learning model
(page 44 column 2 paragraph 3) "The third one is ‘the hybrid’ model, in which the baseline and sequence data are processed separately first, and then fused to predict a single outcome[*Examiner notes: mapped to ensemble machine learning framework]. Specifically, baseline variables are processed within the MLP architecture while sequence data are processed within the LSTM architecture"; figure 1(c)
PNG
media_image2.png
312
752
media_image2.png
Greyscale
Cheng, Jang, and the instant application are analogous because they are all directed to neural networks applied towards adverse medical events.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the present invention to modify the fall prediction machine learning model of Cheng by incorporating the adverse health event prediction system and ensemble technique of Jang because "The ANN model improves upon MEWS and conventional machine learning algorithms in the prediction of cardiac arrest in EDs. We found that the hybrid ANN model utilizing both baseline and sequence information achieved the best performance." (Jang page 48 column 2 line 1). One could reasonably expect that techniques that are effective in predicting adverse cardiac events could also be successfully applied to predicting fall events.
And Wu teaches:
wherein the fall prediction machine learning model is trained based at least in part on distillation loss
(page 2203 paragraph above equation 5) “There are two objective functions[*Examiner notes: mapped to distillation loss] when training the student model. The first objective function L1 minimizes the cross entropy with the soft labels (qTt)i and the soft probability (qTs )i produced by the student model”; equation 7; [*Examiner notes: A distillation loss can be defined as a difference between a distribution of the teacher network’s soft labels (output from the teacher network) and the distribution of the student network’s predictions (output from the teacher network). Thus, the paragraph above teaches a distillation loss in describing exactly this concept.]
Cheng, Jang, Wu, and the present application are analogous because they are all directed to neural networks.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the present invention to modify the machine learning model of Cheng in view of Jang by incorporating the teacher-student framework as taught by Wu because, (Wu Abstract) "Experiments show that we can reach a 2.4× compression rate in a number of parameters and a 1.2× computation reduction with 1.79% loss of accuracy on the UCF-101 dataset and 0.35% loss of accuracy on the HMDB51 dataset."
And Rakshit teaches:
(ii) in an instance where the fall likelihood prediction is an affirmative likelihood prediction, one or more current medical conditions of a user indicated by the numerical timeseries embedding, the categorical timeseries embedding, or the static embedding predicted to cause a fall
(paragraph [0015]) “Application 104 can further communicate with safety manager 110 to transmit instructions to detect events, identify potential causes for the detected events[*Examiner notes: affirmative likelihood prediction], and create visual simulations of the detected events comprising one or more graphic icon overlays indicating potential causes and potential portions of the user that may be injured.”; (paragraph [0039]) “In step 306, safety manager 110 identifies potential causes for the event. In this embodiment, safety manager 110 identifies potential causes for the event by utilizing one or more machine learning algorithms and artificial intelligence algorithms[*Examiner notes: embedding]. […] In this way, safety manager 110 can identify or otherwise predict a reason for the detected event (e.g., low blood pressure occurred during a fall[*Examiner notes: one or more current medical conditions], slipping on a surface, walking down the stairs, accidents, etc.). In this embodiment, safety manager 110 can feed the accessed information to one or more machine learning algorithms an improve its analysis through historical learning.”
generating, by the one or more processors and based on the fall prediction data object, one or more sensory notifications indicating at least one corrective action, the at least one corrective action configured to prevent or lessen a severity of the predicted fall event
(paragraph [0023]) “in certain embodiments, safety manager 110 can generate a text report to accompany the visual simulation that notes any predicted change with respect to the cause of the event, an appropriate notification of the detected event, and suggestions to prevent the detected event in the future. For example, safety manager 110 can recommend additional actions a user can take to increase safety (e.g. personal protective equipment, protocols, etc.).”
Cheng, Jang, Wu, Rakshit, and the instant application are analogous because they are all directed to machine learning.
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the present invention to modify the machine learning model of Cheng in view of Jang and Wu with the indication of medical conditions and fall prevention action as taught by Rakshit because (Rakshit paragraph [0039]) “In this way, safety manager 110 can identify or otherwise predict a reason for the detected event (e.g., low blood pressure occurred during a fall, slipping on a surface, walking down the stairs, accidents, etc.). In this embodiment, safety manager 110 can feed the accessed information to one or more machine learning algorithms an improve its analysis through historical learning” and (Rakshit paragraph [0023]) “For example, safety manager 110 can recommend additional actions a user can take to increase safety (e.g., personal protective equipment, protocols, etc.). In certain other embodiments, safety manager 110 can, with user permission, monitor safety protocols and transmit reminders and notifications to users to follow safety protocols.”
Regarding Claim 2
Cheng in view of Jang, Wu, and Rakshit teaches:
The computer-implemented method of claim 1
(see rejection of claim 1)
And Cheng further teaches:
wherein the fall prediction data object further describes, in the instance where the fall likelihood prediction is the affirmative likelihood prediction, a fall timing prediction
(Paragraph [0213]) “An imminent fall alert[*Examiner notes: mapped to fall timing prediction] may be communicated to the subject through a noticeable alert, such as an audio alarm, tactile stimulus or other output that is recognizable as urgent and simple to interpret, and providing the subject some warning that the system 610 has assessed a high likelihood of the subject falling.”
Regarding Claim 3
Cheng in view of Jang, Wu, and Rakshit teaches:
The computer-implemented method of claim 1
(see rejection of claim 1)
And Jang further teaches:
wherein: the second RNN framework comprise a long short term memory RNN framework.
(page 44 column 2 paragraph 3) "The third one is ‘the hybrid’ model, in which the baseline and sequence data are processed separately first, and then fused to predict a single outcome. Specifically, baseline variables are processed within the MLP architecture while sequence data are processed within the LSTM architecture[*Examiner notes: mapped to the second RNN framework, which comprises a long short term memory RNN]"; figure 1(c)
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the present invention to combine Cheng, Wu, and Rakshit with Jang for the same reasons given in claim 1 above.
Regarding Claim 4
Cheng in view of Jang, Wu, and Rakshit teaches:
The computer-implemented method of claim 1
(see rejection of claim 1)
Cheng further teaches:
[…] and a ground truth fall cause
(paragraph [0214]) "training based on the resulting classified activity data 638 and the training confirmation 645, the loss function is calculated 658. The loss function is based on the difference between the classified activity data 638 of the training data 628 and the training confirmation 645[mapped to distillation loss being based on a ground truth fall cause]."; figure 17 box 658
And Wu further teaches:
wherein: the fall prediction machine learning model has fewer parameters as compared to a trained teacher fall prediction machine learning model
(page 2205 column 2 conclusion) "In this study, we compressed the model, which is currently the most efficient method for action recognition, and improved the overall speed by using knowledge distillation technology to transfer its knowledge to a small model. The small model[*Examiner notes: mapped to fall prediction machine learning model] has richer knowledge than the "vanilla" small model, yet has fewer parameters and less complexity than the original cumbersome models[*Examiner notes: trained teacher fall prediction machine learning model]."; (page 2204 colum 1 conclusion) “Our models were pre-trained on the ILSVRC2012-CLS dataset”
and wherein the distillation loss comprises a custom loss generated by a custom loss model and a distillation loss score
(Page 2203 paragraph above equation 5) “There are two objective functions[mapped to custom loss model] when training the student model.”; equations 7 and 8; [*Examiner note: The model is custom as it is chosen by the authors of the paper. Equation 7 defines a distillation loss score]
PNG
media_image3.png
135
732
media_image3.png
Greyscale
and wherein the distillation loss score is based at least in part on one or more teacher outputs from the trained teacher fall prediction machine learning model, one or more inferred outputs of the fall prediction machine learning model, a ground truth fall likelihood
(page 2203, column 1 just above equation 5) “The first objective function L1 minimizes the cross entropy with the soft labels (qTt)i and the soft probability (qTs)i produced by the student model. (qTs)i"; equation 7; [*Examiner note: qTs is the output from the student model, qTt is the output from the teacher model. Equation 7 includes reference to both of these variables.]
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the present invention to combine Cheng, Jang, and Rakshit with Wu for the same reasons given in claim 1 above.
Regarding Claim 6
Cheng in view of Jang, Wu, and Rakshit teaches:
The computer-implemented method of claim 1
(see rejection of claim 1)
And Cheng further teaches:
further comprising transmitting a fall prediction notification describing the fall prediction data object to an edge client computing entity
(paragraph [0208]) "The communication module 642 may also communicate the classified activity data 638 to the subject through an alert output 699, which may be located on a garment 698[*Examiner notes: mapped to edge client computing entity], in the event that the classified activity data 638 is indicative of an imminent fall[*Examiner notes: mapped to fall prediction notification]."; figure 16 no. 698 and 699
wherein the edge client computing entity is configured to present the one or more sensory notifications to an end user of the edge client computing entity.(paragraph [0213]) "An imminent fall alert may be communicated to the subject through a noticeable alert[*Examiner notes: mapped to one or more sensory notifications], such as an audio alarm, tactile stimulus or other output that is recognizable as urgent and simple to interpret, and providing the subject some warning that the system 610 has assessed a high likelihood of the subject falling."
Regarding Claim 7
Cheng in view of Jang, Wu, and Rakshit teaches:
The computer-implemented method of claim 1
(see rejection of claim 1)
And Cheng further teaches:
wherein the one or more sensory notifications comprise at least one of: (i) one or more audiovisual notifications, (ii) one or more haptic notifications, and (iii) one or more electrical pulse notifications
(paragraph [0213]) "An imminent fall alert may be communicated to the subject through a noticeable alert, such as an audio alarm, tactile stimulus or other output that is recognizable as urgent and simple to interpret, and providing the subject some warning that the system 610 has assessed a high likelihood of the subject falling."
Regarding Claim 8
Claim 8 is a system claim corresponding to method claim 1. The only difference is that claim 8 recites a system comprising one or more processors and at least one memory.
Cheng teaches:
A system comprising one or more processors and at least one memory storing processor- executable instructions that, when executed by any one or more of the one or more processors, cause the one or more processors to perform operations
(paragraph [0224]) “Embodiments of the disclosure can be represented as a computer program product stored in a machine-readable medium[*Examiner notes: mapped to memory] (also referred to as a computer-readable medium, a processor-readable medium, or a computer usable medium having a computer-readable program code embodied therein). […] The instructions stored on the machine-readable medium can be executed by a processor or other suitable processing device, and can interface with circuitry to perform the described tasks.”
The remaining limitations of the claim are taught by the rejection of claim 1.
Regarding Claim 9
Claim 9 is a system claim corresponding to method claim 2. The only difference is that claim 9 recites a system comprising one or more processors and at least one memory, taught in the rejection of claim 8 above. The remaining limitations of the claim are taught by the rejection of claim 2.
Regarding Claim 10
Claim 10 is a system claim corresponding to method claim 3. The only difference is that claim 10 recites a system comprising one or more processors and at least one memory, taught in the rejection of claim 8 above. The remaining limitations of the claim are taught by the rejection of claim 3.
Regarding Claim 11
Claim 11 is a system claim corresponding to method claim 4. The only difference is that claim 11 recites a system comprising one or more processors and at least one memory, taught in the rejection of claim 8 above. The remaining limitations of the claim are taught by the rejection of claim 4.
Regarding Claim 13
Claim 13 is a system claim corresponding to method claim 6. The only difference is that claim 13 recites a system comprising one or more processors and at least one memory, taught in the rejection of claim 8 above. The remaining limitations of the claim are taught by the rejection of claim 6.
Regarding Claim 14
Claim 14 is a system claim corresponding to method claim 7. The only difference is that claim 14 recites a system comprising one or more processors and at least one memory, taught in the rejection of claim 8 above. The remaining limitations of the claim are taught by the rejection of claim 7.
Regarding Claim 15:
Claim 15 is a non-transitory computer-readable storage media claim corresponding to method claim 1. The only difference is that claim 15 recites one or more non-transitory computer-readable storage media.
Cheng teaches:
One or more non-transitory computer-readable storage media storing processor- executable instructions that, when executed by one or more processors, cause the one or more processors to perform operations
(paragraph [0224]) “Embodiments of the disclosure can be represented as a computer program product stored in a machine-readable medium (also referred to as a computer-readable medium, a processor-readable medium, or a computer usable medium having a computer-readable program code embodied therein). The machine-readable medium can be any suitable tangible, non-transitory medium […] The instructions stored on the machine-readable medium can be executed by a processor or other suitable processing device, and can interface with circuitry to perform the described tasks.”
The remaining limitations of the claim are taught by the rejection of claim 1.
Regarding Claim 16
Claim 16 is a non-transitory computer-readable storage media claim corresponding to method claim 2. The only difference is that claim 16 recites one or more non-transitory computer-readable storage media, taught by the rejection of claim 15 above. The remaining limitations of the claim are taught by the rejection of claim 2.
Regarding Claim 17
Claim 17 is a non-transitory computer-readable storage media claim corresponding to method claim 3. The only difference is that claim 17 recites one or more non-transitory computer-readable storage media, taught by the rejection of claim 15 above. The remaining limitations of the claim are taught by the rejection of claim 3.
Regarding Claim 18
Claim 18 is a non-transitory computer-readable storage media claim corresponding to method claim 4. The only difference is that claim 18 recites one or more non-transitory computer-readable storage media, taught by the rejection of claim 15 above. The remaining limitations of the claim are taught by the rejection of claim 4.
Regarding Claim 20:
Cheng in view of Jang in view of Wu teaches:
The one or more non-transitory computer-readable storage media of claim 15
(see rejection of claim 15)
And Cheng further teaches:
wherein the operations further comprise transmitting a fall prediction notification describing the fall prediction data object to an edge client computing entity
(paragraph [0208]) "The communication module 642 may also communicate the classified activity data 638 to the subject through an alert output 699, which may be located on a garment 698[*Examiner notes: mapped to edge client computing entity], in the event that the classified activity data 638 is indicative of an imminent fall[*Examiner notes: mapped to fall prediction notification]."; figure 16 no. 698 and 699
wherein the edge client computing entity is configured to present one or more sensory notifications to an end user of the edge client computing entity
(paragraph [0213]) "An imminent fall alert may be communicated to the subject through a noticeable alert[*Examiner notes: mapped to one or more sensory notifications], such as an audio alarm, tactile stimulus or other output that is recognizable as urgent and simple to interpret, and providing the subject some warning that the system 610 has assessed a high likelihood of the subject falling."
based at least in part on the fall prediction notification
(paragraph [0213]) "An imminent fall alert may be communicated to the subject through a noticeable alert[*Examiner notes: mapped to one or more sensory notifications], such as an audio alarm, tactile stimulus or other output that is recognizable as urgent and simple to interpret, and providing the subject some warning that the system 610 has assessed a high likelihood of the subject falling[*Examiner notes: based at least in part on the fall prediction notification].”
Claims 5, 12, and 19 are rejected under 35 U.S.C. 103 as being unpatentable over Cheng in view of Jang, Wu, Rakshit, and further in view of NPL reference Sultana et al. “Classification of Indoor Human Fall Events Using Deep Learning” herein referred to as Sultana.
Regarding Claim 5
Cheng in view of Jang, Wu, and Rakshit teaches:
The computer-implemented method of claim 1
(see rejection of claim 1)
And Cheng further teaches:
wherein the fall prediction machine learning model is trained based at least in part on a custom loss generated by a custom loss model
(paragraph [0214]) “Where the system 610 is applying the training data 628 as the input data 620 and is training based on the resulting classified activity data 638 and the training confirmation 645, the loss function is calculated 658[*Examiner notes: mapped to custom loss]. The loss function is based on the difference between the classified activity data 638 of the training data[*Examiner notes: mapped to custom loss model] 628 and the training confirmation 645. The loss function may be applied to setting weights and biases 659. After setting the weights and biases 659, additional input data 620 may be received and applied to the neural network 690[*Examiner notes: mapped to fall prediction machine learning model].”
Cheng in view of Jang, Wu, and Rakshit does not teach:
and the custom loss model comprises a fall likelihood component
However, Sultana teaches:
and the custom loss model comprises a fall likelihood component
(page 9 above equation 5) “Finally, sigmoid activation function [31] is applied for binary classification of fall and non-fall[*Examiner notes: mapped to fall likelihood] action through the following equations.”; (page 9 above equation 6) “Here following binary cross-entropy loss function [32] equation is used to compare the predicted class with the actual class.”; [*Examiner notes: The loss model of Sultana includes a fall likelihood component because the predicted and actual classes represent a likelihood of a fall (as shown in the preceding citation)]
Cheng, Jang, Wu, Rakshit, Sultana and the present application are analogous because they are all directed towards neural networks.
It would have been obvious to a person having ordinary skill in the art to modify the neural network of Cheng in view of Jang in view of Wu to include the ground truth fall likelihood, and thereby incorporating comparisons of the fall likelihood component into the loss function as taught by Sultana because (Sultana page 10 section 4.3 line 1) "To implement this scratch model, we have made several experiments on the percentage of training and validation datasets to achieve better accuracy and we have got a mean accuracy of 99% where 99.8% and 98% accuracy for the UR fall detection dataset and Multiple cameras fall dataset, respectively". That is, since Cheng includes making a determination about falls/non-falls, there would be a benefit of implementing a sophisticated and accurate model to perform this step.
Regarding Claim 12
Claim 12 is a system claim corresponding to method claim 5. The only difference is that claim 12 recites a system comprising one or more processors and at least one memory, taught in the rejection of claim 8 above. The remaining limitations of the claim are taught by the rejection of claim 5.
Regarding Claim 19
Claim 19 is a non-transitory computer-readable storage media claim corresponding to method claim 5. The only difference is that claim 19 recites one or more non-transitory computer-readable storage media, taught by the rejection of claim 15 above. The remaining limitations of the claim are taught by the rejection of claim 5.
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 Ezra J Baker whose telephone number is (703)756-1087. The examiner can normally be reached Monday - Friday 10:00 am - 8:00 pm 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, David Yi can be reached at (571) 270-7519. 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.
/E.J.B./Examiner, Art Unit 2126
/DAVID YI/Supervisory Patent Examiner, Art Unit 2126
Prosecution Insights