Description:FORM 2
THE PATENTS ACT, 1970
(39 of 1970)
&
THE PATENT RULES, 2003

COMPLETE SPECIFICATION
(See Section 10 and Rule 13)

Title of invention:
A SYSTEM AND METHOD FOR PREDICTING PERSONALIZED GLYCEMIC RESPONSE OF A USER

Applicant:
FITTERFLY HEALTHTECH PRIVATE LIMITED
An Indian Entity having address as:
503, Akshar Blue Chip Corporate Park
Turbhe MIDC, Navi Mumbai 400705

The following specification particularly describes the invention and the manner in which it is to be performed.
CROSS-REFERENCE TO RELATED APPLICATIONS AND PRIORITY
The present application does not claim priority from any other patent application.

FIELD OF INVENTION
The present subject matter described herein, in general, relates to a personalized glycaemic response. More particularly, the present invention relates to a glycemic response prediction for metabolic health management.

BACKGROUND OF THE INVENTION
The subject matter discussed in the background section should not be assumed to be prior art merely as a result of its mention in the background section. Similarly, a problem mentioned in the background section or associated with the subject matter of the background section should not be assumed to have been previously recognized in the prior art. The subject matter in the background section merely represents different approaches, which in and of themselves may also correspond to implementations of the claimed technology.
With the advent of increasing global prevalence of diabetes and as the number of people living with diabetes continues to rise, effective glucose management strategies have become crucial. Glycemic glucose response prediction aims to forecast an individual's blood glucose levels in response to various factors, such as diet, exercise, medication, and other contextual variables. Managing blood glucose levels can be challenging, as they are influenced by a multitude of interrelated factors. The dynamic nature of these variables makes it difficult for individuals with diabetes to maintain stable blood glucose levels, which can lead to health complications. Personalized medicine has emerged as a critical concept in diabetes management, recognizing that each individual's metabolic response is unique. More particularly, existing solutions don't take into account the individual differences in glycemic response to food, exercise, and medication.
Existing self-monitoring of blood glucose (SMBG) tools, such as traditional fingerstick testing, encounter several substantial challenges. These methods are invasive and uncomfortable, involving painful skin pricks, which often lead to infrequent testing and suboptimal diabetes management. Furthermore, these approaches provide only momentary snapshots of glucose data at specific times, making it challenging to discern glucose level trends. Consequently, they may overlook crucial spikes and dips in blood sugar between tests. The repetitive nature of skin pricking can also elevate the risk of infection and discourage regular monitoring practices.
Moreover, these conventional methods fail to account for various influencing factors, such as stress and illness, which can significantly affect glucose levels. Consequently, fingerstick testing alone frequently falls short in providing a comprehensive and holistic understanding of an individual's glucose management. Further, it includes accuracy issues and the need for regular calibration, which can pose hurdles in achieving precise and reliable glucose data.
Additionally, the traditional systems solely depend on the continuous glucose monitoring (CGM) sensors for identifying the user’s glycemic response in view of the meal consumed by the user. In the absence of CGM sensor, the traditional systems do not offer personalized nutritional advice tied to real-time blood sugar levels, resulting in people guessing how a particular meal will affect them. Further, traditional methods may not offer real-time data or predictive capabilities, making it challenging for individuals to manage glycemic levels effectively.
Artificial intelligence and machine learning play a pivotal role in this field, enabling the analysis of extensive datasets to make accurate predictions. Predictive models take into account various factors, including insulin levels, carbohydrate intake, physical activity, sleep, and stress, to estimate future blood glucose levels. These models have significant clinical implications, allowing healthcare providers to tailor treatment plans and empower individuals with diabetes to make informed decisions about their daily lives.
Despite its potential, glycemic response prediction faces challenges, such as the need for high-quality data, model interpretability, and ethical considerations surrounding privacy and data usage. Nonetheless, the future holds promise as technology continues to advance, leading to more accurate and user-friendly predictive models.
Thus, there exists a need for an innovative approach to glucose management that integrates machine learning models with personalized management. Furthermore, it represents a transformative approach to diabetes management, offering a personalized and data-driven means of achieving better blood glucose control and overall health for individuals with diabetes.

SUMMARY OF THE INVENTION
Before the present system and device and its components are described, it is to be understood that this disclosure is not limited to the system and its arrangement as described, as there can be multiple possible embodiments which are not expressly illustrated in the present disclosure. It is also to be understood that the terminology used in the description is for the purpose of describing the versions or embodiments only and is not intended to limit the scope of the present application. This summary is not intended to identify essential features of the claimed subject matter nor is it intended for use in detecting or limiting the scope of the claimed subject matter.
In one embodiment of the present disclosure, a system for predicting Personalized Glycemic Response (PGR) of a user is disclosed. The PGR may comprise a graph visualizing user’s glucose level based on user’s food intake, user’s physical activity and medication taken by the user. The system may comprise a user interface (UI) which may be configured to receive one or more user inputs. Further, the one or more user inputs may correspond to at least one of user’s food log, user’s activity details, information on medication taken by the user, and a combination thereof. Further, the system may comprise a glucose monitoring sensor which may be configured to monitor user’s glucose data. Further, the system may comprise a memory and a processor coupled with memory. The processor may be configured to execute programmed instructions stored in the memory. In an embodiment, the processor may be configured to execute the programmed instructions for generating a PGR graph based on the user’s glucose data and the one or more user inputs. The user’s glucose data, by the glucose monitoring sensor and the one or more user inputs (food, exercise, and medication) may be captured for a predetermined time-interval. Furthermore, the processor may be configured to execute programmed instructions for training a machine learning model that may be based on the PGR graph, the user’s glucose data and the one or more user inputs. Furthermore, the processor may derive one or more parameters from the PGR graph. The one or more parameters may correspond to user’s glucose pattern with respect to the one or more user’s inputs. Further, the processor may be configured to execute programmed instructions for determining a glycemic response score that may be based on the one or more derived parameters. Moreover, the processor may be configured to calculate a PGR score. The PGR score may be calculated by combining the glycemic response score with one or more food factors from the user’s food log. Further, the processor may be configured to predict a real-time PGR graph and a real-time PGR score of the user. The real-time PGR graph and the real-time PGR score may be predicted using the trained machine learning model and the one or more user inputs captured by the system in real-time. In one embodiment, the real-time PGR graph may be predicted using the trained machine learning model and the real-time user inputs. In a related embodiment, the real-time PGR score may be predicted by combining one or more parameters derived from the predicted real-time PGR graph and one or more food factors from the user’s food log captured in real-time through the UI.
In another embodiment of the present disclosure, a method for predicting Personalized Glycemic Response (PGR) of a user, is disclosed. The PGR may comprise a graph visualizing the user’s glucose level based on the user’s food intake, the user’s physical activity and medication taken by the user. The method may comprise a step of receiving one or user input through a user interface (UI). The one or more user input may correspond to at least one of the user’s food log, user’s activity details, information on medication taken by the user, and a combination thereof. Further, the method may comprise a step of monitoring the user’s glucose data using a glucose monitor sensor. The method may further comprise a step of generating a PGR graph based on the user’s glucose data and the one or more user inputs. The user’s glucose data, by the glucose monitoring sensor and the one or more user inputs (food, exercise, and medication) may be captured for a predetermined time-interval. Further, the method may comprise a step of training a machine learning model based on the PGR graph, the user’s glucose data, and the one or more user inputs. The method may further comprise a step of deriving one or more parameters from the PGR graph. The one or more parameters may correspond to the user’s glucose pattern concerning the one or more user inputs. Further, the method comprises a step of determining a glycemic response score based on the one or more derived parameters. Further, the method comprises a step of calculating a PGR score. The PGR score may be calculated by combining the glycemic response score with one or more food factors from the user’s food log. Further, the method may comprise a step of predicting a real-time PGR graph and a real-time PGR score of the user. The real-time PGR graph and the real-time PGR score may be predicted using the trained machine learning model and the one or more user inputs captured by the method in real-time. In one embodiment, the real-time PGR graph may be predicted using the trained machine learning model and the real-time user inputs. In a related embodiment, the real-time PGR score may be predicted by combining one or more parameters derived from the predicted real-time PGR graph and one or more food factors from the user’s food log captured in real-time through the UI.
The foregoing summary is illustrative only and is not intended to be in any way limiting. In addition to the illustrative aspects, embodiments, and features described above, further aspects, embodiments, and features will become apparent by reference to the drawings and the following detailed description.
BRIEF DESCRIPTION OF DRAWINGS
The detailed description is described with reference to the accompanying figures. In the figures, the left-most digit(s) of a reference number identifies the figure in which the reference number first appears. The same numbers are used throughout the drawings to refer like features and components.
Figure 1A illustrates a block diagram describing a system (100) for predicting personalized Glycemic Response (PGR), in accordance with an embodiment of the present subject matter.
Figure 1B illustrates a block diagram describing a server (101) for predicting the PGR of a user, in accordance with an embodiment of the present subject matter.
Figure 2A-2B illustrates a flow diagram describing a method (200) for providing personalized Glycemic Response (PGR) prediction, in accordance with an embodiment of the present subject matter, and
Figure 3 illustrates an output (300) generated by the system (100), in accordance with an embodiment of the present subject matter.
DETAILED DESCRIPTION OF THE INVENTION
Reference throughout the specification to “various embodiments,” “some embodiments,” “one embodiment,” or “an embodiment” means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment. Thus, appearances of the phrases “in various embodiments,” “in some embodiments,” “in one embodiment,” or “in an embodiment” in places throughout the specification are not necessarily all referring to the same embodiment. Furthermore, the features, structures or characteristics may be combined in any suitable manner in one or more embodiments.
The words "comprising," "having," "containing," and "including," and other forms thereof, are intended to be equivalent in meaning and be open ended in that an item or items following any one of these words is not meant to be an exhaustive listing of such item or items or meant to be limited to only the listed item or items. It must also be noted that, the singular forms "a," "an," and "the" include plural references unless the context clearly dictates otherwise. Although any methods similar or equivalent to those described herein can be used in the practice or testing of embodiments of the present disclosure, the exemplary methods are described. The disclosed embodiments are merely exemplary of the disclosure, which may be embodied in various forms.
The integration of artificial intelligence (AI) machine learning model (ML) in health management has revolutionized the industry by leveraging advanced algorithms to analyse vast datasets, enhance diagnostic accuracy, personalize treatment plans, and streamline administrative processes.
The present invention is directed toward addressing the limitations associated with conventional glycemic response management. Existing approaches often lack the capability to provide personalized and real-time insights into an individual's glycemic responses. These systems typically rely on manual data entry, sporadic glucose monitoring, and do not account for factors like food intake, physical activity, and medication, leading to suboptimal diabetes management.
Further, by integrating a user interface for comprehensive data collection and a continuous glucose monitoring sensor, the present invention can generate personalized glycemic responses and train a machine learning model to make sense of this data. The resulting machine learning model becomes a powerful ally in predicting and visualizing glycemic patterns in real-time, considering not only the individual's glucose data but also their lifestyle choices and medication regimens. Additionally in an instance, the disclosed invention predicts the glycemic pattern of the user without even taking the input from the glucose monitoring sensor data.
In one embodiment of the present disclosure, a system for predicting personalized glycemic response of a user is disclosed. The system, based on receiving one or more user inputs (food intake, physical activities, and medication information) and based on monitoring user’s glucose data, generates a PGR graph and further trains a machine learning model based on the generated PGR graph. Further, the system may calculate a PGR score based on one or more parameters derived from the PGR graph. Furthermore, the system may predict a PGR graph (also known as a real-time PGR graph) based on the trained machine learning model and the one or more real-time user inputs (food intake, physical activities, and medication information). It is important to note that the real-time PRG graph is predicted by the system without taking input (user’s glucose data) from the glucose monitoring sensor. In an embodiment, the system may predict a PGR score (also known as a real-time PGR score) based on the real-time PGR graph, trained machine learning model and the one or more real-time user inputs, without taking input (user’s glucose data) from the glucose monitoring sensor.
Referring to Figure 1A, a block diagram describing a system (100) for predicting personalized glycemic response (PGR) of a user is illustrated in accordance with an embodiment of a present subject matter. The system (100) may comprise a server (101), a network (102), one or more user devices (103) and one or more glucose monitoring sensors (104). The server (101) may be configured to predict the personalized glycemic response for one or more users. In an embodiment, the personalized glycemic response comprises a PGR graph, a PGR score, and a combination thereof. The PGR graph corresponds to a visual representation of the user’s glucose pattern in view of the user’s food intake, user’s activity/exercise information and medication taken by the user. Further, the PGR score corresponds to a rating scale representing the user’s glucose pattern in view of the user’s food intake, user’s activity/exercise information and medication taken by the user.
In an embodiment, the one or more users may be connected to the server (101), via the network (102) for predicting their personalized glycemic response (Graph and Score). The one or more users comprises one or more user devices (103-1, 103-2, …103-n) referred to as user device (103) hereinafter. In a particular embodiment of the system the user device (103) is selected from a group including but not limited to a cell phone, personal digital assistant (PDA), lap top computer, stationary personal computer, IPTV remote control, web tablet, laptop computer, pocket PC, a television set capable of receiving IP based video services and mobile IP device. The user device (103) may be used to provide one or more user inputs to the server (101). The one or more user inputs comprise at least one of the user’s food log, user’s activity details, information on medication taken by the user, and a combination thereof. In one embodiment, the user’s food log corresponds to information of one or more meals consumed by the user. The information of one or more meals comprise detailed breakdown of micronutrients and macronutrients present in the food consumed by the user. In one exemplary embodiment, the micronutrients or macronutrients correspond to energy, Carbohydrates (carbs), protein, fat, fiber present in the food. In another embodiment, the user’s activity details comprise information about the user’s physical activity/exercise. In an exemplary embodiment, the user’s activity details correspond to details of exercise manually entered by the user on the user device (103). In another exemplary embodiment, the user’s activity details correspond to details of activities automatically tracked by one or more activity trackers. The one or more activity trackers, in one embodiment, may be embedded in the user device (103), and in another embodiment, may be communicatively coupled with the user device (103). The one or more activity trackers may comprise to one of, but not limited to, Accelerometer, Gyroscope, Altimeter, Bioimpedance sensor, Proximity sensor, GPS, Electrodermal activity or EDA sensor, Temperature sensor, Optical sensor (ex. Camera) and a combination thereof. In another embodiment, information on medication taken by the user corresponds to medication which affects the user’s glucose data. The one or more medication affecting user’s glucose data may comprise one of, but not limited to, Beta-blockers, Cibenzoline, quinidine, Glinides, Indomethacin, Insulin, Metformin, SGLT2 inhibitors, Sulfonylureas, Thiazolidinediones and a combination thereof. In an embodiment, the one or more user inputs may be provided to the user device (103) by using one or more sensors from one of, but not limited to, Accelerometer, Gyroscope, Altimeter, Bioimpedance sensor, Proximity sensor, GPS, Electrodermal activity or EDA sensor, Temperature sensor, Optical sensor (ex. Camera), heart rate sensor, SpO2 monitor, ECG sensor, UV sensor and a combination thereof.
In another embodiment, the one or more users may be connected to the server (101), via the network (102), using one or more glucose monitoring sensors (104-1, 104-2, …104-n) referred to as glucose monitoring sensor (104) hereinafter. In a particular embodiment the glucose monitoring sensor (104) corresponds to a continuous glucose monitoring (CGM) sensor coupled with the user’s body part. The CGM estimates glucose levels and wirelessly sends the information to a software program on a smartphone or insulin pump. The program calculates how much insulin your body needs, and the insulin pump delivers the insulin when glucose levels rise higher than target range. In an exemplary embodiment, for a user to enable the system (100) to predict personalized glycemic response (PGR), the user may need to communicate one or more user inputs (user’s food intake, user’s activity/exercise information and medication taken by the user) from the user device (103) and user’s glucose data from the glucose monitoring sensor (104) to the server (101) for a predetermined time-interval. In one implementation, the pre-determined time-interval corresponds to 14 days of time. For the period of 14 days, the user needs to provide the one or more user inputs to the server (101), along with the glucose monitoring sensor (104) periodically measuring and reporting the user’s glucose data to the server (101).
In another embodiment, the server (101), the user device (103) and the glucose monitoring sensor (104) are connected via the network (102). In one implementation, the network (102) may be a wireless network, a wired network, or a combination thereof. The network (102) can be implemented as one of the different types of networks, such as intranet, local area network (LAN), wide area network (WAN), the internet, and the like. The network (102) may either be a dedicated network or a shared network. The shared network represents an association of the different types of networks that use a variety of protocols, for example, Hypertext Transfer Protocol (HTTP), Transmission Control Protocol/Internet Protocol (TCP/IP), Wireless Application Protocol (WAP), and the like, to communicate with one another. In another embodiment, the network (102) may include any one of the following: the wireless network (e.g., Bluetooth, Near Field Communication (NFC)), a mobile telephone network (e.g., 3G, 4G, 5G, 6G).
In another embodiment, the server (101) may comprise various components to predict the personalized glycemic response (PGR graph and PGR score) based on the one or more user inputs and the user’s glucose data. The detailed explanation of the server (101) will be described below in explanation of Figure 1B.
Although the present disclosure is explained considering that the system (100) is implemented on a server (101), it may be understood that the system (100) may be implemented in a variety of computing systems, such as a laptop computer, a desktop computer, a notebook, a workstation, a virtual environment, a mainframe computer, a server, a network server, a cloud-based computing environment. It will be understood that the system (100) may be accessed by multiple users through one or more user devices (103). In one implementation, the system (100) may comprise the cloud-based computing environment in which the user may operate individual computing systems configured to execute remotely located applications. Examples of the user devices (103) may include, but are not limited to, a portable computer, a personal digital assistant, a handheld device, and a workstation.
The system (100) can be implemented using hardware, software, or a combination of both, which includes using where suitable, one or more computer programs, mobile applications, or “apps” by deploying either on-premises over the corresponding computing terminals or virtually over cloud infrastructure. The system (100) may include various micro-services or groups of independent computer programs which can act independently in collaboration with other microservices. The system (100) may also interact with a third-party or external computer system. Internally, the system (100) may be the central processor of all requests for transactions by the various actors or users of the system. a critical attribute of the system (100) is that it can concurrently and instantly complete an online transaction by a system user in collaboration with other systems.
Now referring to Figure 1B, a block diagram describing the server (101) for predicting the PGR of a user, is illustrated in accordance with an embodiment of the present subject matter. The server (101) comprises a processor (105), a user interface (UI) (106) and a memory (107). Further, the memory (107) comprises a machine learning model (108), a graph generation module (109), a score generation module (110). The memory (107) may further comprise data (111) which may further comprise training data (112) and user data (113). The processor (105) is coupled with the memory (107). The processor (105) is configured to execute programmed instructions stored in the memory (107). The processor (105), in one embodiment, may comprise a standard microprocessor, microcontroller, central processing unit (CPU), distributed or cloud processing unit, state machines, logic circuitries, and/or any devices that manipulate signals based on operational instructions and/or other processing logic that accommodates the requirements of the present invention.
Further, the UI (106) is an interface to other components of the server (101) and the system (100). The UI (106) may include a variety of software and hardware interfaces, for example, a web interface, a graphical user interface, and the like. The UI (106) may allow the system (100) to interact with the user directly or through the user devices (103). Further, the UI (106) may enable the system (100) to communicate with other computing devices, such as web servers and external data servers (not shown). The UI (106) can facilitate multiple communications within a wide variety of networks and protocol types, including wired networks, for example, LAN, cable, etc., and wireless networks, such as WLAN, cellular, or satellite. The UI (106) may include one or more ports for connecting a number of devices to one another or to another server. In one embodiment, the UI (106) allows the server (101) to be logically coupled to another user device (103), some of which may be built in. Illustrative components include tablets, mobile phones, scanner, printer, wireless device, etc. In an exemplary embodiment, the UI (106) may be configured to receive the one or more user inputs from the user device (103). In another exemplary embodiment, the UI (106) may be configured to receive the user’s glucose data from the glucose monitoring sensor (104). Further, the processors (105) may be configured to store the one or more user inputs and the user’s glucose data, received via the UI (106) to the user data (113) of the memory (107).
The memory (107) may include any computer-readable medium or computer program product known in the art including, for example, volatile memory, such as static random-access memory (SRAM) and dynamic random-access memory (DRAM), and/or non-volatile memory, such as read only memory (ROM), erasable programmable ROM, flash memories, hard disks, Solid State Disks (SSD), optical disks, magnetic tapes, memory cards, virtual memory and distributed cloud storage. The memory (107) may be removable, non-removable, or a combination thereof. The memory (107) may include routines, programs, objects, components, data structures, etc., which perform particular tasks or implement particular abstract data types. The memory (107) may include programs or coded instructions that supplement applications and functions of the system (100). In one embodiment, the memory (107), amongst other things, serves as a repository for storing data processed, received, and generated by one or more of the programs or the coded instructions. In yet another embodiment, the memory (107) may be managed under a federated structure that enables adaptability and responsiveness of the server (101). Further, the memory (107) comprises the machine learning model (108), the graph generation module (109), the score generation module (110).
In the present system, machine learning model (108), particularly neural networks, are utilized for various predictive tasks, such as estimating the glucose levels of an individual. Neural networks, which may comprise multiple layers, including input and output layers with hidden layers in between, are employed. Each layer in a neural network performs specific transformation operations, often referred to as neurons. These neurons calculate their output as a weighted sum of their inputs, adjusted by a bias and passed through an activation function, such as a rectified linear unit (ReLU) or sigmoid function.
In general, the training process of a neural network involves providing input data to an untrained network, generating predicted outputs, and then comparing these predictions to the expected outputs. The network's weights and biases are iteratively updated to minimize the difference between predicted and expected outputs, typically by computing the derivative of a cost function. Training converges when the cost function reaches a small magnitude, satisfying a convergence condition. Various types of neural networks are applicable in this system, including Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), such as Long Short-Term Memory (LSTM) networks, and others. CNNs are designed to receive input data in small, localized portions, allowing them to detect features throughout the dataset. They may incorporate convolutional and pooling layers, as well as fully connected layers for object detection in images or videos. RNNs, on the other hand, are well-suited for time-series data, such as continuous glucose monitoring. An RNN includes input layers for receiving sequential inputs and hidden recurrent layers to maintain a state across time steps, capturing dependencies in the input sequence. LSTMs, a specific type of RNN, consist of LSTM units, each comprising a cell, input gate, output gate, and forget gate. The cell preserves dependencies, while the gates regulate the flow of information into and out of the cell. The logistic function serves as the activation function for these gates. This architecture allows LSTMs to effectively encode time-series features, like glucose levels in a subject, for analysis and prediction. In one embodiment of the present disclosure, the machine learning model (108) may correspond to time series forecasting models, selected from one of Arima, LSTM, CNN-LSTM, ensemble models and a combination thereof.
In an embodiment, the machine learning model (108) may be trained based on population level user input and corresponding population level glucose data. In one embodiment, the population level user input corresponds to food log, activity details and medication information of a large number of the population, stored in the training data (112) of the memory (107). In another embodiment, the population level glucose data corresponds to glucose data of the large number of population corresponding to their population level user input. In yet another embodiment, the machine learning model (108) may be trained using Personalized Glycemic Response (PGR) (i.e., PGR Graph and PGR score) of the large number of population. In an implementation, the processor (105) may be coupled to the memory (107) and the machine learning model (108). The processor (105) may be configured to execute programmed instructions stored in the memory to train the machine learning model (108) based on the one or more user inputs and the user’s glucose data, stored in the memory (107). In an embodiment, training the machine learning model (108), by the processor (105), based on the one or more user inputs and the user’s glucose data, corresponds to re-training of the machine learning model (108) which was previously trained on the population level user input, population level glucose data and the population level Personalized Glycemic Response (PGR) (i.e., PGR Graph and PGR score). The trained machine learning model (108) is used to predict a real-time PGR graph (302, as illustrated in Figure 3) of a user based on the one or more user inputs (food log, activity details and medication information) captured through the UI (106) real-time.
In another embodiment, the graph generation module (109) may comprise one or more programmed instructions stored in the memory (107), which with the help of processor (105) generates a PGR graph (not illustrated) based on the one or more user inputs (food log, activity details and medication information) received from the user device (103) and the user’s glucose data received from the glucose monitoring sensor (104). The PGR graph generated by the graph generation module (109) may be used to identify patterns of one of user’s glucose response in the event of meal consumed by the user, user’s glucose response in the event of any exercise or physical activity performed by the user, user’s glucose response in the event of medication, which affects the glucose level data, consumed by the user, and a combination thereof.
In another embodiment, the score generation module (110) may comprise one or more programmed instructions stored in the memory (107), which with the help of processor (105) calculates a PGR score (not illustrated) based on the PGR graph (not illustrated), the one or more user inputs (food log, activity details and medication information) received from the user device (103) and the user’s glucose data received from the glucose monitoring sensor (104). In an implementation of the present disclosure, the score generation module (110) comprises two sub-modules as first sub-module and a second sub-module. The first sub-module is configured to calculate a glycemic response score. However, the second sub-module is configured to calculate a food score. The glycemic response score and the food score are combined, by the score generation module (110), to calculate the PGR score of the user. The first sub-module may comprise one or more programmed instructions stored in the memory (107), which with the help of processor (105) derives one or more parameters from the PGR graph generated by the graph generation module (109). The one or more parameters may correspond to user’s glucose pattern with respect to the one or more user’s inputs (food log, activity details and medication information). In an exemplary embodiment, the one or more parameters comprises one of Age, Gender, HbA1C, BMI, Heart Rate, step counts, caloric expenditure, Co-morbidities affecting metabolic responses, Positive Delta/Max rise (in mg/dl), Pre-meal glucose (in mg/dl) for identifying historic/Trend glucose reading within 15 mins of the meal event, Time to maximum delta/rise (in mins), Time in range (in %), Negative Delta/fall (in mg/dl), Pre-meal glucose (in mg/dl) for identifying historic/Trend glucose reading within 15 mins of the meal event, Time to maximum delta/fall (in mins), Time below 100mg/dl glucose (in %), and a combination thereof. Each parameter from the one or more parameters may be compared against a predefined threshold to identify a corresponding parameter score. The first sub-module with the help of processor (105) may determine a glycemic response score based on the combination of each parameter score of the one or more parameters. The second sub-module may comprise one or more programmed instructions stored in the memory (107), which with the help of processor (105) derives one or more food factors from the food log received by the UI (106). The one or more food factors may correspond to one or more meals consumed by the user along with corresponding detailed macronutrient and micronutrient breakdowns, energy, Carbohydrates (carbs), protein, fat, fiber and a combination thereof. In an embodiment the energy may be derived in percentage (%) of suggested Recommended Dietary Allowances (RDA). Similarly, Carbs, protein, and fat may be derived in % of the suggested RDA of Energy. Similarly, Fiber may be derived in g/1000 kcal. Each factor from the one or more food factors may be compared against a predefined threshold to identify a corresponding factor score. The second sub-module with the help of processor (105) may determine a food score based on the combination of each factor score of the one or more food factors. In an embodiment, the score generation module (110), with the help of processor (105), may calculate the PGR score based on the combination of glycemic response score calculated by the first sub-module and the food score calculated by the second sub-module.
In another embodiment, the processor (105) may be configured to execute programmed instructions stored in the memory (107) for predicting a real-time PGR graph (302, illustrated in Figure 3) and a real-time PGR score (301, illustrated in Figure 3). In an implementation, the processor (105) with the help of the machine learning model (108), the graph generation module (109), the score generation module (110) and one or more real-time user inputs (food log, activity details and medication information), may predict the real-time PGR graph (302) and the real-time PGR score (301). In a specific embodiment, the processor (105) with the help of the machine learning model (108) and the graph generation module (109) may predict the real-time PGR graph (302) based on one or more real-time user inputs (food log, activity details and medication information). The real time PGR graph (302) may be used to predict future patterns of one of user’s glucose response in the event of meal consumed by the user, user’s glucose response in the event of any exercise or physical activity performed by the user, user’s glucose response in the event of medication, which affects the glucose level data, consumed by the user, and a combination thereof. In another specific embodiment, the processor (105) with the help of the real-time PGR graph (302) and the score generation module (110) may predict the real-time PGR score (301) based on one or more real-time user inputs (food log, activity details and medication information). The processor (105) may predict the real-time PGR score (301) based on combining one or more parameters derived from the predicted real-time PGR graph (302) and one or more food factors from the user’s food log captured in real-time through the UI (106). It is important to note that, the real-time PGR graph (302) and the real-time PGR score (301) predicted based on the machine learning model (108), without taking the user’s glucose data from the glucose monitoring sensor (104).
Referring to Figure 2, a flow diagram describing a method (200) for predicting PGR of a user, is illustrated in accordance with an embodiment of the present disclosure. The method (200) comprises a step of receiving (201) one or more user inputs through a user interface (UI) (106). The said one or more user inputs corresponds to at least one of the user’s food log, user’s activity details, information on medication taken by the user and a combination thereof. Further, the method (200) comprises a step of monitoring (202) user’s glucose data using a glucose monitoring sensor (104). The method (200) comprises a step of generating (203) a PGR graph based on the user’s glucose data and the one or more user inputs captured for a predetermined time-interval. Furthermore, the method (200) comprises a step of training (204) a machine learning model (108) based on the PGR graph, the user’s glucose data, and the one or more user inputs. The method (200) comprises a step of deriving (205) one or more parameters from the generated PGR graph. The one or more derived parameters corresponds to the user’s glucose pattern concerning the one or more user inputs. The method (200) comprises a step of determining (206) a glycemic response score based on the one or more derived parameters. The method (200) comprises a step of calculating (207) a PGR score. The PGR score is calculated by combining the glycemic response score with one or more food factors from the user’s food log, and the method (200) comprises a step of predicting (208) a real-time PGR graph (302) and a real-time PGR score (301) of the user using the trained machine learning model (108) and the one or more user inputs captured in real-time through the user interface (UI) (106).
In another embodiment, the method (200) comprises displaying the real-time PGR graph (301) and the real-time PGR score (301) in real-time through the UI (106).
In yet another embodiment, the PGR graph corresponds to a graphical representation of the user’s glucose levels in view of the monitored user’s glucose data and the one or more user inputs, over the predetermined time-interval.
In yet another embodiment, the real-time PGR graph (302) corresponds to visualizing prediction of the user’s glucose level in view of the one or more user inputs, captured in real-time.
In yet another embodiment, the method (200) comprises predicting the real-time PGR score (301) based on combining one or more parameters derived from the predicted real-time PGR graph (302) and one or more food factors from the user’s food log captured in real-time through the UI (106).
In yet another embodiment, the user’s food log corresponds to one or more meals consumed by the user along with corresponding detailed macronutrient and micronutrient breakdowns, wherein the user’s activity details correspond to either exercise detail manually entered by the user or activity details tracked by one or more activity trackers, wherein information on medication corresponds to medication taken by the user which affects user’s glucose data.
In yet another embodiment, the method (200) comprises the machine learning model (108) corresponds to time series forecasting models, selected from one of Arima, LSTM, CNN-LSTM, ensemble models and a combination thereof.
In yet another embodiment, the machine learning model (108) corresponds to a general machine learning model trained on a population level users’ glucose data and population level user inputs on one of food log, activity details, medication information and a combination thereof.
In yet another embodiment, the one or more parameters, correspond to one of Age, Gender, HbA1C, BMI, Heart Rate, step counts, caloric expenditure, Co-morbidities that affect metabolic response of the one or more user’s input (120) and one of Positive Delta/Max rise (in mg/dl), Pre-meal glucose (in mg/dl), Time to maximum delta/rise (in mins), Time in range (in %), Negative Delta/fall (in mg/dl), Pre-meal glucose (in mg/dl), Time to maximum delta/fall (in mins), Time below 100mg/dl glucose (in %), and a combination thereof.
In yet another embodiment, the one or more food factors, from the user’s food log, correspond to one or more meals consumed by the user along with corresponding detailed macronutrient and micronutrient breakdowns, energy, Carbohydrates (carbs), protein, fat, fiber and a combination thereof.
Referring to Figure 3, an output (300) generated by the system (100), is illustrated in accordance with an embodiment of the present subject matter. The output (300) may comprise a real-time PGR score (301) and a real-time PGR graph (302). The real time PGR score (301) may correspond to future rating of one of user’s glucose response in the event of meal consumed by the user, user’s glucose response in the event of any exercise or physical activity performed by the user, user’s glucose response in the event of medication, which affects the glucose level data, consumed by the user, and a combination thereof. Further, the real time PGR graph (302) may be used to predict, through visual representation, future patterns of one of user’s glucose response in the event of meal consumed by the user, user’s glucose response in the event of any exercise or physical activity performed by the user, user’s glucose response in the event of medication, which affects the glucose level data, consumed by the user, and a combination thereof. It is important to note that, the real-time PGR graph (302) and the real-time PGR score (301) predicted based on the machine learning model (108), without taking the user’s glucose data from the glucose monitoring sensor (104). In one embodiment, the output (300) may be presented to the user on their corresponding user device (103).
The system (100) as disclosed in the disclosure may help in predicting personalized glycemic response with the following advantages:
• The PGR Scoring is a unique feature that quantifies glycemic response and nutritional value in a single, easily interpretable score.
• This system predicts blood sugar spikes and dips up to three hours in advance unlike existing systems which are reactive, providing data after a health event has occurred.
• While some existing solutions might only focus on glucose levels or dietary inputs, this system also accounts for exercise, sleep quality, gut health, and medication, providing a more comprehensive picture of overall health.
• The system also possesses two-fold PGR scoring when sensor data is available and PGR charting when the sensor is off.
• The system’s capability to distinguish between major and minor meals, as well as the type of glycemic response (positive or negative delta), allows for a much more nuanced understanding and management strategy.
• Unlike traditional systems that assess nutrients individually, this system examines the synergistic effects between different types of macronutrients and how they collectively impact glycemic response.
Various modifications to the embodiment will be readily apparent to those skilled in the art and the generic principles herein may be applied to other embodiments. However, one of ordinary skill in the art will readily recognize that the present disclosure is not intended to be limited to the embodiments illustrated but is to be accorded the widest scope consistent with the principles and features described herein.
The foregoing description shall be interpreted as illustrative and not in any limiting sense. A person of ordinary skill in the art would understand that certain modifications could come within the scope of this disclosure.
The embodiments, examples and alternatives of the preceding paragraphs or the description and drawings, including any of their various aspects or respective individual features, may be taken independently or in any combination. Features described in connection with one embodiment are applicable to all embodiments unless such features are incompatible.
, Claims:WE CLAIM:
1. A system (100) for predicting Personalized Glycemic Response (PGR) of a user, characterized in that, the system (100) comprises:
a user interface (UI) (106), wherein the user interface (106) is configured to receive one or more user inputs, wherein the one or more user inputs correspond to at least one of user’s food log, user’s activity details, information on medication taken by the user, and a combination thereof;
a glucose monitoring sensor (104), configured to monitor user’s glucose data;
a memory (107);
a processor (105) coupled with the memory (107), wherein the processor (105) is configured to execute programmed instructions stored in the memory (107), wherein the programmed instructions comprise:
generating a PGR graph based on the user’s glucose data and the one or more user inputs, wherein the user’s glucose data and the one or more user inputs are captured for a predetermined time-interval;
training a machine learning model (108), based on the PGR graph, the user’s glucose data and the one or more user inputs;
deriving one or more parameters from the generated PGR graph, wherein one or more parameters correspond to user’s glucose pattern with respect to the one or more user’s inputs;
determining a glycemic response score based on the one or more derived parameters;
calculating a PGR score, wherein the PGR score is calculated by combining the glycemic response score with one or more food factors from the user’s food log, and
predicting a real-time PGR graph (302) and a real-time PGR score (301) of the user, using the trained machine learning model (108) and the one or more user inputs, wherein the one or more user inputs are captured in real-time through the UI (106).

2. The system (100) as claimed in claim 1, wherein the PGR graph corresponds to a graphic representation of user’s glucose levels in view of the monitored user’s glucose data and the one or more user inputs, over the predetermined time-interval.

3. The system (100) as claimed in claim 1, wherein the real-time PGR graph (302) corresponds to visualize prediction of the user's glucose level in view of the one or more user inputs captured in real-time.

4. The system (100) as claimed in claim 1, wherein the processor (105) predicts the real-time PGR score (301) based on combining one or more parameters derived from the predicted real-time PGR graph (302) and one or more food factors from the user’s food log captured in real-time through the UI (106).

5. The system (100) as claimed in claim 1, wherein the user’s food log corresponds to one or more meals consumed by the user along with corresponding detailed macronutrient and micronutrient breakdowns, wherein the user’s activity details correspond to either exercise detail manually entered by the user or activity details tracked by one or more activity trackers, wherein the information on medication corresponds to medication taken by the user which affects user’s glucose data.

6. The system (100) as claimed in claim 1, wherein the glucose monitoring sensor (104) corresponds to a continuous glucose monitoring (CGM) sensor coupled with user’s body part.

7. The system (100) as claimed in claim 1, wherein the machine learning model (108) corresponds to time series forecasting models, selected from one of Arima, LSTM, CNN-LSTM, ensemble models and a combination thereof.

8. The system (100) as claimed in claim 1, wherein the machine learning model (108) corresponds to a general machine learning model trained on a population level users’ glucose data and population level user inputs on one of food log, activity details, medication information and a combination thereof.

9. The system (100) as claimed in claims 1 and 4, wherein the one or more parameters, correspond to one of Age, Gender, HbA1C, BMI, Heart Rate, step counts, caloric expenditure, Co-morbidities affecting metabolic responses, Positive Delta/Max rise (in mg/dl), Pre-meal glucose (in mg/dl), Time to maximum delta/rise (in mins), Time in range (in %), Negative Delta/fall (in mg/dl), Pre-meal glucose (in mg/dl), Time to maximum delta/fall (in mins), Time below 100mg/dl glucose (in %), and a combination thereof.

10. The system (100) as claimed in claim 1, wherein the one or more food factors, from the user’s food log, correspond to one or more meals consumed by the user along with corresponding detailed macronutrient and micronutrient breakdowns, energy, Carbohydrates (carbs), protein, fat, fiber and a combination thereof.

11. A method (200) for predicting Personalized Glycemic Response (PGR) of a user, characterized in that, the method (200) comprising steps of:
receiving (201) one or more user inputs through a user interface (UI) (106), said one or more user inputs corresponding to at least one of user's food log, user's activity details, information on medication taken by the user, and a combination thereof;

monitoring (202) user's glucose data using a glucose monitoring sensor (104);

generating (203) a PGR graph based on the user's glucose data and the one or more user inputs, wherein the user’s glucose data and the one or more user inputs are captured for a predetermined time-interval;

training (204) a machine learning model (108) based on the PGR graph, the user's glucose data, and the one or more user inputs;

deriving (205) one or more parameters from the generated PGR graph, wherein the one or more parameters correspond to the user's glucose pattern concerning the one or more user inputs;

determining (206) a glycemic response score based on the one or more derived parameters;

calculating (207) a PGR score, wherein the PGR score is calculated by combining the glycemic response score with one or more food factors from the user's food log, and

predicting (208) a real-time PGR graph (302) and a real-time PGR score (301) of the user, using the trained machine learning model (108) and the one or more user inputs, wherein the one or more user inputs are captured in real-time through the UI (106).

12. The method (200) as claimed in claim 11, wherein the method (200) comprises displaying the real-time PGR graph (302) and the real-time PGR score (301) in real-time through the UI (106).

13. The method (200) as claimed in claim 11, wherein PGR graph corresponds to a graphic representation of user’s glucose levels in view of the monitored user’s glucose data and the one or more user inputs, over the predetermined time-interval.

14. The method (200) as claimed in claim 11, wherein the real-time PGR graph (302) corresponds to visualize prediction of the user's glucose level in view of the one or more user inputs captured in real-time.

15. The method (200) as claimed in claim 11, wherein the method (200) comprises predicting the real-time PGR score (301) based on combining one or more parameters derived from the predicted real-time PGR graph (302) and one or more food factors from the user’s food log captured in real-time through the UI (106).

16. The method (200) as claimed in claim 11, wherein the user’s food log corresponds to one or more meals consumed by the user along with corresponding detailed macronutrient and micronutrient breakdowns, wherein the user’s activity details correspond to either exercise detail manually entered by the user or activity details tracked by one or more activity trackers, wherein the information on medication corresponds to medication taken by the user which affects user’s glucose data.

17. The method (200) as claimed in claim 11, wherein the machine learning model (108) corresponds to time series forecasting models, selected from one of Arima, LSTM, CNN-LSTM, ensemble models and a combination thereof.

18. The method (200) as claimed in claim 11, wherein the machine learning model (108) corresponds to a general machine learning model trained on a population level users’ glucose data and population level user inputs on one of food log, activity details, medication information and a combination thereof.

19. The method (200) as claimed in claims 11 and 15, wherein the one or more parameters, correspond to one of Age, Gender, HbA1C, BMI, Heart Rate, step counts, caloric expenditure, Co-morbidities affecting metabolic responses, Positive Delta/Max rise (in mg/dl), Pre-meal glucose (in mg/dl), Time to maximum delta/rise (in mins), Time in range (in %), Negative Delta/fall (in mg/dl), Pre-meal glucose (in mg/dl), Time to maximum delta/fall (in mins), Time below 100mg/dl glucose (in %), and a combination thereof.

20. The method (200) as claimed in claim 11, wherein the one or more food factors, from the user’s food log, correspond to one or more meals consumed by the user along with corresponding detailed macronutrient and micronutrient breakdowns, energy, Carbohydrates (carbs), protein, fat, fiber and a combination thereof.

Dated this 27th day of October 2023

Deepak Pawar
Agent for the Applicant
IN/PA-2052