DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Amendment
The amendment filed 10/30/2025 has been entered. Claims 1-5, 7 - 13, 15 - 17, 19, 21, and 22 - 24 are pending. Claims 6 and 14 have been cancelled.
Claim Rejections - 35 USC § 112
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
Claims 11 - 15 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claims contain subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventors, at the time the application was filed, had possession of the claimed invention.
In regard to claim 11, lines 7 - 9 recite, “generating personalized glucose ranges, based at least in part on inputting the glucose data into one or more machine learning mechanisms and receiving the personalized glucose ranges and a baseline as an output…” However, the specification does not specifically disclose or suggest that the baseline, defined in paragraph [0064] of the specification as the “average glucose level at rest before meals,” is output from the machine learning mechanism. The language of claim 11 implies that the baseline is in some way being generated by the machine learning mechanism, which is not supported by the specification or figures. Claims 12, 13, and 15 are rejected by virtue of dependence on claim 11.
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 11 - 15 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
In regard to claim 11, lines 7 - 9 recite, “generating personalized glucose ranges, based at least in part on inputting the glucose data into one or more machine learning mechanisms and receiving the personalized glucose ranges and a baseline as an output…” However, it is unclear if the one or more machine learning mechanisms are also generating the “baseline” by some calculation or processing step, which is implied by the one or more machine learning mechanisms outputting the baseline, or if the one or more machine learning mechanisms are receiving the value from elsewhere and outputting the baseline. Further clarification is required. Claims 12, 13, and 15 are rejected by virtue of dependence on claim 11.
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 - 5, 7 - 13, 15 - 17, 19, and 21, 23, and 24 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claims recite steps or executable instructions as part of a method or system (both of which are within a statutory category of invention) to “generate” personalized glucose ranges in the case of claims 1 & 11 and “generate” a postprandial prediction response and biomarker score in the case of claim 16, which fall in the category of a mental process. This judicial exception is not integrated into a practical application because with regard to Revised step 2A, prong 1, an exception is present as noted above, and with regard to Revised step 2A, prong 2, the claim does not recite additional elements that integrate the judicial exception into a practical application. Further, with regard to step 2B, the claim does not recite additional elements that integrate the judicial exception into practical application. In particular, as written, claim 1 only positively claims the limitations of “accessing” data, “generating ranges” based on “inputting” and “receiving” data, and “causing data… to be presented”, which merely amounts to generalized data manipulation steps in a general computer environment. The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception because while “one or more machine learning mechanisms” are discussed in claim 1, they are not positively claimed, thus only the steps of “inputting” and “receiving” data are required which does not integrate the judicial exception into practical application. Additionally, claims 2 - 10 merely discuss steps for accessing different types of data with no specific method or structure required to collect the data. Further, the memory and processor discussed in claim 11 are general structures and do not impose a meaningful limitation onto the claim scope, as the limitations do not constitute use of the exception in the context of “a particular machine” and claims 12, 13, & 15 are directed towards accessing or receiving data and presenting data to a user, which is not sufficient to integrate the judicial exception into practical application. Like claim 1, claim 16 is directed towards “accessing” data, “generating ranges” based on “inputting” and “receiving” data, and “causing data… to be presented” in addition to “determining data,” which merely amounts to generalized data manipulation steps in a general computer environment. Claims 17, 19, 21, and 23 are directed towards steps of accessing data to perform a calculation or generate a score and claim 24 is directed towards presenting a visual to a user, all of which are not sufficient to integrate the judicial exception into practical application.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1 - 5, 7 - 13, 15 - 19, 23, & 24 are rejected under 35 U.S.C. 103 as being unpatentable over Kamath (US 11389090 B2 - Previously Cited), and further in view of Simpson (US 20160328991 A1) and further in view of Arunachalam (US 20200098463 A1).
In regard to claim 1, Kamath discloses a method comprising:
accessing glucose data associated with a user, the glucose data generated during a period of time; Kamath discloses that the method includes the use of an analyte sensor system (FIG. 1, component 102) that is coupled to a host to measure an analyte, such as glucose (Column 24, lines 28 - 30). The analyte sensor system communicates via wireless communication signal (FIG. 1, component 110) with other devices such as computers, tablets, smart phone devices, and remote terminals (FIG. 1; Column 26, lines 50 - 63)
accessing context data associated with the user, the context data representing activities of the user during the period of time: Kamath discloses that context data, including activity data, user input information, location information, and sources of physiological data other than a continuous glucose monitor (CGM) are accessed by the system (FIG. 10, component 1004; Column 44, lines 47 - 58)
generating personalized glucose ranges, wherein the personalized glucose ranges are personalized for the user and include a personalized upper limit and a personalized lower limit; Kamath discloses that a range may be set by a system or by default for 70-160 mg/dL, and the range may be expanded (e.g., to 70-180 mg/dL) or contracted as is appropriate for a particular patient, which may be determined for example from a pattern of CGM or low-fidelity data, such as context data describing the activity of the user. Examiner notes expansion and contraction of glucose range as appropriate for a particular patient as indicating the range would be considered generating a personalized range. Kamath further discloses a personalized upper limit and a personalized lower limit (Figure 12A depicts upper limit of 180 mg/dL and lower limit of 80 mg/dL).
and causing data associated with the personalized glucose ranges to be presented within a user interface to the user. Kamath further discloses that the guidance, glucose concentration level, or other information may be transmitted via a network 124 and delivered to a patient or caretaker device, such as hand-held device 112 or tablet 114 (Column 27, lines 29-32; FIG. 1).
While Kamath discloses the generation of personalized ranges of glucose ranges using glucose and context data, they do not disclose that the generated ranges are based at least in part on one or more machine learning mechanisms and receiving the personalized glucose ranges as an output.
However, Simpson teaches a method for continuously monitoring a user using a continuous blood glucose monitor and other context data, such as activity data (paragraph [0013]), and outputting a graphical representation of the user’s data compared to a desired range of values over a time period where the desired range of values may be based on a modeled, ideal, or predicted glucose concentration value or range of values (paragraph [0020]). For instance, in FIG. 6, Simpson teaches a computed range of values with an upper bound (FIG. 6, component 92) and lower bound (FIG. 6, component 94) used as a comparison for the user’s glucose trace values (FIG. 6, component 88).
It would have been obvious to one of ordinary skill in the art to have modified the method disclosed by Kamath with the teaching of Simpson that a glucose range can be calculated using machine learning mechanisms, such as computation modeling methods, because it would be considered combining prior art elements according to known methods to yield the predictable result of generating personalized glucose ranges.
While Kamath discloses using glucose and context data to personalize a glucose range and Simpson further teaches the use of a machine learning mechanism to establish personalized glucose ranges, the combination does not specify that the generated ranges are based at least in part on inputting the glucose data and the context data into one or more machine learning mechanisms and receiving the personalized glucose ranges as an output.
However, Arunachalam teaches that machine learning may be utilized to determine an equation, function, or model for calculating the glucose level as a function of a subset of input variables that are correlative to or predictive of the subsequent glucose level based on a variety of inputs including historical data associated with the cluster of historical patient states and further that the operational context associated with the patient at the particular time, one or more attributes for activities or events (e.g., a bolus amount of insulin, an amount of carbohydrates consumed, an exercise duration and/or intensity, and/or the like) that are input to the glucose prediction model may be varied to determine a range of potential predicted glucose outcomes for the patient (paragraph [0043]), such as maintaining a user’s glucose levels within a target range (paragraph [0112]) as that outlined in FIG. 10, component 1006.
It would have been obvious to one of ordinary skill in the art to have modified the method disclosed by Kamath as modified by Simpson, which includes the generation of a personalized glucose range using glucose and context data and the use of machine learning mechanisms to generate a glucose range, with the teaching of Arunachalam that blood glucose data and context data can be used as inputs to a machine learning model to determine a range of potential predicted glucose outcomes for a user because Kamath already discloses the use of blood glucose and context data to determine an individualized blood glucose range for a user and Simpson further teaches that machine learning mechanisms can be utilized to generate individualized blood glucose ranges for a user, the teaching that blood glucose data and context data can be used as inputs to a machine learning model to determine a blood glucose range would be considered combining prior art elements according to known methods to yield the predictable result of generating personalized glucose ranges.
In regard to claim 2, Kamath as modified discloses the method of claim 1, further comprising receiving the glucose data from a continuous glucose monitoring (CGM) device associated with the user. Kamath discloses receiving the glucose data from a continuous glucose monitoring (CGM) device associated with the user where the first server system (FIG. 1, component 125) handles the communication of CGM data (Column 27, lines 29-30).
In regard to claim 3, Kamath as modified discloses the method of claim 1, wherein the glucose data is collected during a calibration period that lasts at least two days. Kamath discloses a calibration period where an estimated glucose concentration level is received from a continuous glucose monitoring (CGM) system for a first time period, where the first time period is a number of days such as 7 days, 10 days, or 14 days (Column 44, lines 36-43).
In regard to claim 4, Kamath as modified discloses the method of claim 1, further comprising receiving nutritional data associated with the user, wherein the nutritional data indicates at least one food item that the user consumed during a calibration period. Kamath discloses that information on user meals may optionally include information about the meal such as meal content including the content of carbohydrates, protein and fat, or food type such as “one slice of pizza and a glass of milk” (Column 49-50, lines 67-5). Kamath additionally discloses that the nutritional information is recorded during the calibration period (Column 22, lines 19-21), wherein non-CGM data is inclusive of user-input data (Column 4, lines 24-25), such as meal data (Column 3, lines 15-17).
In regard to claim 5, Kamath as modified discloses the method of claim 1, wherein the context data indicates at least one of a time of day of consumption of a previously eaten food item, an identity of the a previously eaten food item, an amount of exercise, and/or an amount of sleep associated with the user during a calibration period. Kamath discloses that context data includes non-CGM information relating to the patient may be received for the first time period, including physiologic information about the patient such as one or more of heart rate, respiration, oxygen concentration, skin tone, moisture content on the skin, activity, or activity patterns (Figure 10, steps 1002 and 1004, Col 44, lines 47-56). Kamath further discloses that context data is input by the user during the calibration period (period (Column 22, lines 19-21), wherein non-CGM data is inclusive of user-input data (Column 4, lines 24-25), such as meal, exercise, and sleep data (Column 3, lines 15-17).
In regard to claim 7, Kamath as modified discloses the method of claim 1, further comprising:
determining a time of day in which a food item was eaten by the user; Kamath discloses that different types of data including CGM data and activity data may be correlated with one or more of time to allow matching of CGM data with corresponding activity data, location, day, patient-input data, or other information (Column 23, lines 50-54). Examiner notes that patient-input data includes meal and activity data (Column 3, lines 15-17)
determining that a glucose level changes to at least one of above a first threshold value or below a second threshold value during the time of day; Kamath discloses an upper and lower target glucose level which act as a first threshold and second threshold (FIG. 12, components 1232 & 1230). Glucose levels are tracked as a function of time (FIG. 12, component 1234) and glucose level changes can be visually observed by a user as well as if the glucose value exceeds a first threshold value or goes below a second threshold value.
associating the glucose level change with the food item in response to the glucose level changing when the food item was eaten. Kamath discloses that glucose level changes are displayed with event data such as when a food item was eaten (FIG. 12A, components 1208, 1212, and 1220).
In regard to claim 8, Kamath as modified discloses the method of claim 1, further comprising:
receiving at least one input identifying at least one food item; Kamath discloses that information about user meals is collected including information about the meal including information about the food item type such as “one slice of pizza and a glass of milk” (Column 49-50, lines 67-5)
determining food data associated with the at least one food item; Kamath discloses that information about user meals is collected including information about the meal such as nutritional content (e.g., carbohydrates, protein and fat; Column 49-50, lines 67-5)
determining health data associated with the user; Kamath discloses collected sensor data may be used in combination with population-based data, such as gender, age, location, ethnicity, job type, A1C, BMI, weight, or other demographic information to determine health data associated with a user, such as glucose control estimates including estimated glucose level, range, or status (Column 24, lines 57-61).
generating a postprandial glucose response prediction associated with the at least one food item based at least in part on the food data and the health data; Kamath discloses that their method includes the generation of a predicted effect on glucose response based on received non-CGM information relating to a user (FIG. 10, component 1008), where the non-CGM information includes user input data such as meal data, exercise data, activity data, and sleep data (Column 46, lines 14 - 21), and the previously determined relationship from the calibration or first time period between the estimated glucose concentration levels, or health data, and non-CGM information (FIG. 10).
causing the postprandial glucose response prediction to be presented within a user interface to the user in relation to their personalized glucose ranges. Kamath further discloses that the system may electronically notify the patient about diabetic information (FIG. 10, component 1012). The display can include visual information (FIG. 12A) that includes upper and lower bounds of the user’s personalized target range (FIG. 12A, components 1230 & 1232) with meal indicators (FIG. 12A, components 1208, 1212, 1220) and graphical representations of estimated glucose concentration levels (FIG. 12A, component 1236).
In regard to claim 9, Kamath as modified discloses the method of claim 8, wherein the food data includes at least one of sugar content, carbohydrate content, glycemic index of the carbohydrate, fiber content, an amount of processing, or a food category. Kamath discloses that a user enters information about consumed meals, including meal content, such as information on carbohydrates, fat, and protein, or identifying a food type or category, such as pizza or a glass of milk (Column 49-50, lines 67-5)
In regard to claim 10, Kamath as modified discloses the method of claim 8, wherein health data includes at least one of a blood glucose control score, a blood fat control score, or a gut health score. Kamath discloses that collected sensor data may be used in combination with population-based data, such as gender, age, location, ethnicity, job type, A1C, BMI, weight, or other demographic information to determine a glucose control estimate such as estimated glucose level, range, or status (Column 24, lines 57-61). Examiner notes that glucose control estimate is equivalent to blood glucose control score.
In regard to claim 11, Kamath discloses a system comprising:
one or more processors (FIG. 2, components 204, 256, & 276);
and one or more non-transitory computer-readable media comprising instructions that (Column 14, lines 7 - 10), when executed by the one or more processors, cause the one or more processors to perform operations comprising:
accessing glucose data associated with a user. Kamath discloses that the method includes the use of an analyte sensor system (FIG. 1, component 102) that is coupled to a host to measure an analyte, such as glucose (Column 24, lines 28 - 30). The analyte sensor system communicates via wireless communication signal (FIG. 1, component 110) with other devices such as computers, tablets, smart phone devices, and remote terminals (FIG. 1; Column 26, lines 50 - 63).
generating personalized glucose ranges, wherein the personalized glucose ranges are personalized for the user and include a personalized upper limit and a personalized lower limit; Kamath discloses that a range may be set (e.g., by a system or by default) for 70-160 mg/dL, and the range may be expanded (e.g., to 70-180 mg/dL) or contracted as is appropriate for a particular patient, which may be determined for example from a pattern of CGM or low-fidelity data, such as context data describing the activity of the user. Examiner notes expansion and contraction of glucose range as appropriate for a particular patient as indicating the range would be considered generating a personalized range. Kamath further discloses a personalized upper limit and a personalized lower limit (Figure 12A depicts upper limit of 180 mg/dL and lower limit of 80 mg/dL).
and causing data associated with the personalized glucose ranges to be presented within a user interface to the user. Kamath further discloses that the guidance, glucose concentration level, or other information may be transmitted via a network 124 and delivered to a patient or caretaker device, such as hand-held device 112 or tablet 114 (Column 27, lines 29-32; FIG. 1).
While Kamath discloses the generation of personalized ranges of glucose ranges using glucose data, they do not disclose that generating the personalized glucose ranges and baseline is based at least in part on one or more machine learning mechanisms and receiving the personalized glucose ranges and a baseline as an output or that data associated with the personalized glucose ranges and the baseline are presented concurrently to a user.
However, Simpson teaches a method for continuously monitoring a user using a continuous blood glucose monitor and other context data, such as activity data (paragraph [0013]), and outputting a graphical representation of the user’s data compared to a desired range of values over a time period where the desired range of values may may be based on a modeled, ideal, or predicted glucose concentration value or range of values (paragraph [0020]). For instance, in FIG. 6, Simpson teaches a computed range of values with an upper bound (FIG. 6, component 92) and lower bound (FIG. 6, component 94) used as a comparison for the user’s glucose trace values (FIG. 6, component 88). Additionally, Simpson teaches that an ideal or desired glucose trace can be modeled and displayed near a graphed glucose trace (paragraph [0207]) in addition to the personalized glucose range of values (FIG. 6, components 60 & 70). Examiner notes that the modeled ideal or desired glucose trace acts as a baseline because the modeled ideal or desired glucose trace acts as a comparison point for the actual measurements taken by a user to determine how well a user is managing their glucose levels.
It would have been obvious to one of ordinary skill in the art to have modified the method disclosed by Kamath with the teaching of Simpson that a glucose range can be calculated using machine learning mechanisms, such as computation modeling methods, because it would be considered combining prior art elements according to known methods to yield the predictable result of generating personalized glucose ranges.
While Kamath discloses using glucose data to personalize a glucose range and Simpson further teaches the use of a machine learning mechanism to establish personalized glucose ranges, the combination does not specify that the generated ranges are based at least in part on inputting the glucose data into one or more machine learning mechanisms and receiving the personalized glucose ranges and baseline as an output.
However, Arunachalam teaches that machine learning may be utilized to determine an equation, function, or model for calculating the glucose level as a function of a subset of input variables that are correlative to or predictive of the subsequent glucose level based on a variety of inputs including historical data associated with the cluster of historical patient states and further that the operational context associated with the patient at the particular time, one or more attributes for activities or events (e.g., a bolus amount of insulin, an amount of carbohydrates consumed, an exercise duration and/or intensity, and/or the like) that are input to the glucose prediction model may be varied to determine a range of potential predicted glucose outcomes for the patient (paragraph [0043]), such as maintaining a user’s glucose levels within a target range (paragraph [0112]) such as that outlined in FIG. 10, component 1006. Additionally, Arunachalam teaches that recurrent neural networks can be utilized to predict an average glucose level, or baseline, of the patient (paragraph [0102]) to achieved a desired A1C. Examiner notes that in paragraph [0064] of the specification, Applicant defines the “baseline” as the “average glucose level at rest before meals”.
It would have been obvious to one of ordinary skill in the art to have modified the method disclosed by Kamath as modified by Simpson, which includes the generation of a personalized glucose range using glucose and context data and the use of machine learning mechanisms to generate a glucose range, with the teaching of Arunachalam that blood glucose data and context data can be used as inputs to a machine learning model to determine a range of potential predicted glucose outcomes for a user because Kamath already discloses the use of blood glucose and context data to determine an individualized blood glucose range for a user and Simpson further teaches that machine learning mechanisms can be utilized to generate individualized blood glucose ranges for a user, the teaching that blood glucose data and context data can be used as inputs to a machine learning model to determine a blood glucose range would be considered combining prior art elements according to known methods to yield the predictable result of generating personalized glucose ranges.
In regard to claim 12, Kamath as modified discloses the method of claim 11, wherein the personalized glucose ranges are further based at least in part on accessing at least one of personalized health score data, single meal data, single food data, or combination of meal data. Kamath discloses that context data includes non-CGM information relating to the patient may be received for the first time period, including physiologic information about the patient such as one or more of heart rate, respiration, oxygen concentration, skin tone, moisture content on the skin, activity, or activity patterns (Figure 10, steps 1002 and 1004, Col 44, lines 47-56). Kamath further discloses that context data is input by the user during the calibration period (period (Column 22, lines 19-21), wherein non-CGM data is inclusive of user-input data (Column 4, lines 24-25), such as meal, exercise, and sleep data (Column 3, lines 15-17).
In regard to claim 13, Kamath as modified discloses the system of claim 11, further comprising:
receiving nutritional data associated with the user, wherein the nutritional data indicates at least one food item that the user consumed during a calibration period; Kamath discloses that information on user meals may optionally include information about the meal such as meal content including the content of carbohydrates, protein and fat, or food type such as “one slice of pizza and a glass of milk” (Column 49-50, lines 67-5). Kamath additionally discloses that the nutritional information is recorded during the calibration period (Column 22, lines 19-21), wherein non-CGM data is inclusive of user-input data (Column 4, lines 24-25), such as meal data (Column 3, lines 15-17).
identifying a glucose change associated with the user at a time in which the at least one food item was consumed; Kamath further discloses that their system generates a predicted effect on glucose response based on received non-CGM information relating to a user (FIG. 10, component 1008), where the non-CGM information includes user input data such as meal data, exercise data, activity data, and sleep data (Column 46, lines 14 - 21), and the previously determined relationship from the calibration or first time period between the estimated glucose concentration levels, or health data, and non-CGM information (FIG. 10).
and generating a biomarker score associated with the at least one food item based at least in part on the glucose change. Kamath discloses that guidance displayed to a user is based on the correlation between glycemic health and another measurable input, such as food type. For instance, the system determines which foods cause a large glucose spike and uses that correlation to generate an alert to a user (Column 36, lines 57 - 65). Kamath further discloses that glucose levels, in this case a biomarker score, are displayed in response to meals consumed and reported by the user, such as in an “after lunch report” (FIG. 12A, components 1214 & 1222; Column 49, lines 51 - 53).
In regard to claim 15, Kamath as modified discloses the system of claim 11, wherein the data associated with the personalized glucose ranges includes a line graph indicating a glucose level via a y-axis over a period of time via an x-axis (FIG. 12A & FIG. 12B)
In regard to claim 16, Kamath discloses a method comprising:
accessing glucose data associated with a user, the glucose data being obtained at least partly from a continuous glucose monitor (CGM); Kamath discloses that the method includes the use of an analyte sensor system (FIG. 1, component 102) that is coupled to a host to measure an analyte, such as glucose (Column 24, lines 28 - 30). The analyte sensor system communicates via wireless communication signal (FIG. 1, component 110) with other devices such as computers, tablets, smart phone devices, and remote terminals (FIG. 1; Column 26, lines 50 - 63).
generating personalized glucose ranges, wherein the personalized glucose ranges are personalized for the user and include a personalized upper limit and a personalized lower limit; Kamath discloses that a range may be set (e.g., by a system or by default) for 70-160 mg/dL, and the range may be expanded (e.g., to 70-180 mg/dL) or contracted as is appropriate for a particular patient, which may be determined for example from a pattern of CGM or low-fidelity data, such as context data describing the activity of the user. Examiner notes expansion and contraction of glucose range as appropriate for a particular patient as indicating the range would be considered generating a personalized range. Kamath further discloses a personalized upper limit and a personalized lower limit (Figure 12A depicts upper limit of 180 mg/dL and lower limit of 80 mg/dL).
receiving at least one input identifying at least one food item; Kamath further discloses that their system generates a predicted effect on glucose response based on received non-CGM information relating to a user (FIG. 10, component 1008), where the non-CGM information includes user input data such as meal data (FIG. 10) and meal data includes information including identifying a food type such as “one slice of pizza and a glass of milk” (Column 49-50, lines 67-5).
determining food data associated with the at least one food item, the food data including at least one of a sugar content, a carbohydrate content, a glycemic index of the carbohydrate, a fiber content, an amount of processing, or a food category; Kamath discloses that information on user meals may optionally include information about the meal such as meal content including the content of carbohydrates, protein and fat, or food type such as “one slice of pizza and a glass of milk” (Column 49-50, lines 67-5).
determining health data associated with the user based at least in part on the personalized glucose ranges; Kamath discloses collected sensor data may be used in combination with population-based data, such as gender, age, location, ethnicity, job type, A1C, BMI, weight, or other demographic information to determine health data associated with a user, such as glucose control estimates including estimated glucose level, range, or status (Column 24, lines 57-61).
generating a postprandial response prediction associated with the at least one food item based at least in part on the food data and the health data; ; Kamath discloses that their method includes the generation of a predicted effect on glucose response based on received non-CGM information relating to a user (FIG. 10, component 1008), where the non-CGM information includes user input data such as meal data, exercise data, activity data, and sleep data (Column 46, lines 14 - 21), and the previously determined relationship from the calibration or first time period between the estimated glucose concentration levels, or health data, and non-CGM information (FIG. 10).
generating at least one biomarker score associated with the at least one food item based at least in part on the postprandial response prediction, the biomarker score including a personalized impact of the at least one food item on glucose, lipids, insulin, energy, hunger, or microbiome of the user and causing the at least one biomarker score to be presented within a user interface to the user. Kamath discloses that guidance displayed to a user is based on the correlation between glycemic health and another measurable input, such as food type. For instance, the system determines which foods cause a large glucose spike and uses that correlation to generate an alert to a user (Column 36, lines 57 - 65). Kamath further discloses that glucose levels, in this case a biomarker score, are displayed in response to meals consumed and reported by the user, such as in an “after lunch report” (FIG. 12A, components 1214 & 1222; Column 49, lines 51 - 53).
While Kamath discloses the generation of personalized ranges of glucose ranges using glucose data, they do not disclose that generating personalized glucose ranges is based at least in part on inputting the glucose data into one or more machine learning mechanisms and receiving the personalized glucose ranges as an output.
However, Simpson teaches a method for continuously monitoring a user using a continuous blood glucose monitor and other context data, such as activity data (paragraph [0013]), and outputting a graphical representation of the user’s data compared to a desired range of values over a time period where the desired range of values may may be based on a modeled, ideal, or predicted glucose concentration value or range of values (paragraph [0020]). For instance, in FIG. 6, Simpson teaches a computed range of values with an upper bound (FIG. 6, component 92) and lower bound (FIG. 6, component 94) used as a comparison for the user’s glucose trace values (FIG. 6, component 88).
It would have been obvious to one of ordinary skill in the art to have modified the method disclosed by Kamath with the teaching of Simpson that a glucose range can be calculated using machine learning mechanisms, such as computation modeling methods, because it would be considered combining prior art elements according to known methods to yield the predictable result of generating personalized glucose ranges.
While Kamath discloses using glucose data to personalize a glucose range and Simpson further teaches the use of a machine learning mechanism to establish personalized glucose ranges, the combination does not specify that the generated ranges are based at least in part on inputting the glucose data into one or more machine learning mechanisms and receiving the personalized glucose ranges and baseline as an output.
However, Arunachalam teaches that machine learning may be utilized to determine an equation, function, or model for calculating the glucose level as a function of a subset of input variables that are correlative to or predictive of the subsequent glucose level based on a variety of inputs including historical data associated with the cluster of historical patient states and further that the operational context associated with the patient at the particular time, one or more attributes for activities or events (e.g., a bolus amount of insulin, an amount of carbohydrates consumed, an exercise duration and/or intensity, and/or the like) that are input to the glucose prediction model may be varied to determine a range of potential predicted glucose outcomes for the patient (paragraph [0043]), such as maintaining a user’s glucose levels within a target range (paragraph [0112]) such as that outlined in FIG. 10, component 1006.
It would have been obvious to one of ordinary skill in the art to have modified the method disclosed by Kamath as modified by Simpson, which includes the generation of a personalized glucose range using glucose and context data and the use of machine learning mechanisms to generate a glucose range, with the teaching of Arunachalam that blood glucose data and context data can be used as inputs to a machine learning model to determine a range of potential predicted glucose outcomes for a user because Kamath already discloses the use of blood glucose and context data to determine an individualized blood glucose range for a user and Simpson further teaches that machine learning mechanisms can be utilized to generate individualized blood glucose ranges for a user, the teaching that blood glucose data and context data can be used as inputs to a machine learning model to determine a blood glucose range would be considered combining prior art elements according to known methods to yield the predictable result of generating personalized glucose ranges.
In regard to claim 17, Kamath as modified discloses the method of claim 16, further comprising applying at least one weight to at least one of the health data, wherein the at least one biomarker score is generated based at least in part on the at least one weight. Arunachalam teaches that system utilizes machine learning to determine which combination of historical sensor glucose measurement data, historical delivery data, historical auxiliary measurement data (e.g., historical acceleration measurement data, historical heart rate measurement data, and/or the like), historical event log data, historical geolocation data, and other historical or contextual data are correlated to or predictive of the occurrence of a particular event, activity, or metric for a particular patient, and then determines a corresponding equation, function, or model for calculating the value of the parameter of interest based on that set of input variables (paragraph [0096]). The relative weightings applied to the respective variables of the predictive subset are based on differing correlations between a particular input variable and the historical data for that particular patient (paragraph [0096]). The generate models can be used for calculating future glucose levels or otherwise generating recommendations in a manner that is influenced by historical data (paragraph [0097]).
In regard to claim 19, Kamath as modified discloses the method of claim 16, further comprising generating the at least one biomarker score based on accessing at least one of a family health history associated with the user, measures of blood chemistry taken in a fasting state associated with the user, or measures of blood chemistry that do not change postprandially associated with the user. Kamath discloses that the patient's condition or health history is evaluated for diabetes or diabetes risk factors, including the patient's weight, body mass index, family history, and test results such as A1C, fasting glucose level, and oral glucose tolerance (Column 37-38, lines 66-5).
In regard to claim 23, Kamath as modified discloses the method of claim 1 above, wherein generating personalized glucose ranges further uses non-glucose-factor data, the non-glucose factor data including lipid profile data, obesity data, and health risk data associated with the period of time. Kamath discloses that the personalized glucose ranges are adjusted for a user based on CGM data, low-fidelity data, or data input from a user (Column 49, lines 1 - 16). Low-fidelity data includes a risk metric for the user, or health risk data associated with the user’s past, that is determined based upon demographics, medical records, average or fasting glucose levels, or post-meal glucose excursion characteristics (Column 33, lines 45 - 66). Examiner notes that Kamath discloses that medical or health records consist of information including family history, weight, BMI, history of metabolic disease, blood result tests, or any combination thereof (Column 10, lines 20 - 27) which one of ordinary skill in the art would recognize would reasonably include obesity data (e.g. weight data) and lipid profile data (e.g. blood result tests).
In regard to claim 24, Kamath as modified discloses the system of claim 11 above, wherein the personalized glucose ranges are presented within a user interface to the user, together with a shape of glucose trace logged by the user. Kamath discloses that their system includes the presentation of the personalized glucose ranges, including the upper limit (FIG. 12A, component 1230) and lower limit (FIG. 12A, component 1232), with a shape of a glucose trace logged by the user (FIG. 12A, component 1236).
Claims 21 & 22 are rejected under 35 U.S.C. 103 as being unpatentable over Kamath (US 11389090 B2 - Previously Cited), and further in view of Simpson (US 20160328991 A1) and further in view of Arunachalam (US 20200098463 A1) as applied to claim 16 above and further in view of Hadjigeorgiou (US 20200066181 A1 - Previously Cited).
In regard to claims 21 & 22, Kamath as modified teaches the method of claim 16. Kamath teaches the method of claim 16. However, Kamath does not teach the method further comprising generating personalized triglycerides data for the user, wherein the at least one biomarker score is based at least in part on the personalized triglycerides data, and wherein generating personalized triglycerides data further comprises performing a blood test to generate blood data and determining the personalized triglycerides data based at least in part on the blood data.
However, Hadjigeorgiou teaches generating personalized triglycerides data for a user where a prediction manager (FIG. 1, component 122) can predict target values for the biomarkers associated with triglycerides (paragraph [0041], lines 1-22); and the triglyceride data may include data about triglycerides for an individual (Paragraph [0066]); Additionally, Hadjigeorgiou teaches that the at least one biomarker score is based at least in part on the personalized triglycerides data where the prediction manager utilizes a scorer (FIG. 4, component 126A) to generate a score for a target biomarker, including triglycerides (paragraph [0081]). Hadjigeorgiou further teaches that generating personalized triglycerides data comprises performing a blood test (Paragraph [0066]) to generate blood data (FIG. 3, blood data 306C) and determining the personalized triglycerides data based at least in part on the blood data, where the blood data may include blood tests relating to a variety of different biomarkers, including blood sugar, insulin, triglycerides (Paragraph [0067]). Examiner notes blood data associated with measuring triglycerides from a blood test determining data about triglycerides for an individual as determining personalized triglycerides data because it is coming from an individual’s blood.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the method of Kamath to incorporate the teachings of Hadjigeorgiou by including the additional steps to generate a biomarker score based on personalized triglyceride data and determine personalized triglyceride data from a blood data from a blood test. Doing so would allow for generation of a score for food items or meals, as recognized by Hadjigeorgiou (Paragraph [0024], lines 9-14)
Response to Arguments
Applicant’s arguments, see Remarks, filed 10/30/2025, with respect to the rejection of claims 1 - 17, 19, 21, and 22 under 35 U.S.C. 101 have been fully considered and are partially persuasive. Examiner notes that features of independent claims 1, 11, and 16 are directed towards a mental process, specifically the generation of personalized glucose ranges. While the use of a machine learning mechanism as defined in paragraph [0082] of the specification as including specific models such as decision tree learning, association rule learning, artificial neural networks (including, in examples, deep learning), inductive logic programming, support vector machines, clustering, Bayesian networks, reinforcement learning, representation learning, similarity and metric learning, sparse dictionary learning, and/or rules-based machine learning, recites significantly more than the abstract idea and could integrate the judicial exception into a practical application, the machine learning mechanism referred to in claim 1 is not positively claimed and the method only requires that data is input and received from a machine learning mechanism. The rejection of claims 1 - 5, 7 - 13, 15 - 17, 19, & 21 under 35 U.S.C. 101 have been maintained as evidenced by the updated 101 rejection above which addresses the newly presented limitations to the claims. The rejection of claim 22 under 35 U.S.C. 101 is considered persuasive and the rejection has been withdrawn because the required step of “performing a blood test to generate blood data,” provides sufficient structure to bring the judicial exception into practical application.
Applicant’s arguments, see Remarks, filed 10/30/2025, with respect to the rejections of claims 1 - 16 and 19 under 35 U.S.C. 103 in view of Kamath and in further view of Anand and claims 17, 21, and 22 under 35 U.S.C. 103 in view of Kamath and in further view of Hadjigeorgiou have been fully considered and are persuasive. Therefore, the rejection has been withdrawn. However, upon further consideration, a new grounds of rejection for claims 1-5, 7 - 13, 15 - 17 are made under 35 U.S.C. 103 in view of Kamath and in further view of Simpson and further in view of Arunachalam. Additionally, a new grounds of rejection for claims 21 - 22 are made under 35 U.S.C. 103 in view of Kamath and in further view of Simpson and further in view of Arunachalam and further in view of Hadjigeorgiou.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SIENNA CHRISTINE PYLE whose telephone number is (703)756-5798. The examiner can normally be reached 8 am - 5:30 pm M - T; Off first Fridays; 8 am - 4 pm second Fridays.
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/interviewp