Method and server computer for finding timing of communications regarding physiological information and associated timestamps

Abstract

A method, server computer and network system for finding timing of communications relating to physiological information and associated timestamps are provided. [Solution] The method includes receiving physiological measurement information of a first user from a detection device via a network, obtaining a first communication timing of first health-specific information according to the physiological measurement information and a corresponding timestamp, transmitting the first health-specific information to a client computer at the first timing, receiving a first response from the client computer regarding the first timing, and improving a first machine learning model using the first response.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This patent application is based on Provisional Patent Application No. 63 / 687,005, filed August 26, 2024. [Technical Field]

[0002] The present invention relates to a network system, and more particularly to a network system for notifying a patient of health-specific information with timing calculated according to physiological measurements and corresponding timestamps. [Background technology]

[0003] Patients struggle to remember all of the lifestyle instructions from their doctors and other authority sources and often fail to implement these instructions as part of their daily living tasks. There are several possible reasons for this failure. First, these instructions may be difficult to follow because the timing of implementing them conflicts with the patient's other daily activities. Second, the instructions may be dependent on some physiological measurement. A notification mechanism is needed to ensure that responsive instructions are delivered to the patient in a timely manner.

[0004] A patient was instructed to take his or her antihypertensive medication after breakfast at the same time every day to control his or her blood pressure throughout the day. However, the patient must get up early to go to work several days a week, such as Mondays and Fridays. When he or she gets up at dawn and rushes to work, he or she often forgets to take breakfast and his or her medication, resulting in poor blood pressure control. This illustrates the potential for the patient's other daily activities to conflict with following instructions.

[0005] For the same patient mentioned above, one day he forgot to eat breakfast and take his blood pressure medication. He worked until 11:00 AM and then took a lunch break. During his break, he measured his blood pressure and it was 160 / 100 mmHg. He wondered, "Should I take my blood pressure medication now?" Without timely instructions, the patient would be confused and likely to continue working that afternoon without taking any action regarding his blood pressure. This means that instructions must be provided in a timely manner and be responsive to physiological measurements.

[0006] In the same patient mentioned above, he had been suffering from gastroenteritis for several days and had eaten very little. This morning, he woke up at his usual time. He had no appetite for breakfast, but he remembered that his medication was waiting for him. Nevertheless, when his blood pressure was measured, it was 100 / 50 mmHg, significantly lower than normal. He wondered, "Should I take my blood pressure medication now?" This again illustrates the need for timely instructions and responsiveness to physiological measurements.

[0007] These three illustrative scenarios demonstrate that daily activities can affect patients' daily schedules and interfere with their decision-making, and patients want timely guidance when making decisions after physiological measurements.

[0008] Therefore, there is a need to provide a mechanism for timely notification of instructions to accommodate patients' viable options and respect their decisions. Summary of the Invention [Means for solving the problem]

[0009] According to an embodiment of the present application, there is provided a method for finding a communication timing for physiological information and an associated timestamp, the method including: receiving physiological measurement information of a first user from a sensing device via a network; obtaining a first communication timing of the first health-specific information according to the physiological measurement information and the corresponding timestamp; and transmitting the first health-specific information to a client computer at the first timing.

[0010] In one embodiment of the present application, a server computer for finding communication timings related to physiological information and associated timestamps is provided, the server computer including: a networking device connected to a detection device and a client computer via a network; and a processor configured to execute instructions stored in a non-volatile memory to cause the networking device to receive physiological measurement information of a first user from the detection device, obtain a first communication timing of the first health-specific information according to the physiological measurement information and the corresponding timestamp, and cause the networking device to transmit the health-specific information to the client computer at the first timing.

[0011] According to one embodiment of the present application, there is provided a network system for finding timing of communication relating to physiological information and associated timestamps, the network system comprising a server, a sensing device, and a client computer.

[0012] Based on the method and server provided herein, a prompt notification mechanism is presented to a patient or user that responds to viable options and provides timely notifications to respect their decisions. The mechanism can automatically analyze the periodic schedule and communication timing based on physiological measurement information and corresponding timestamps. Therefore, health-specific information, such as reminders, action suggestions, and / or summaries of physiological measurement information, can be notified at the communication timing found by the mechanism. This allows the patient or user to follow instructions at communication timings that do not conflict with the user's other activities. Also, response reports of physiological measurement information can be delivered to the patient or user at the appropriate time. Following this mechanism, the mechanism can provide accurate information to the user at the appropriate time, allowing the user to make better, proactive, and thoughtful choices.

[0013] The advantages and principles associated with the present invention will be better understood from the following detailed description and drawings. [Brief explanation of the drawings]

[0014] [Figure 1A] 1 is a block diagram of a network system 100 for providing notifications to a user at calculated times, according to one embodiment of the present application. [Figure 1B] 1B illustrates a modification of the network system 100 shown in FIG. 1A for providing notifications to a user at calculated times according to another embodiment of the present application. [Figure 1C] 1B illustrates a modification of the network system 100 shown in FIG. 1A for providing notifications to a user at calculated times according to another embodiment of the present application. [Figure 2] 2 is a block diagram of a computer 200 according to an embodiment of the present application. [Figure 3] FIG. 3 is a flowchart diagram of a method 300 for finding a timing to transmit health-specific information according to an embodiment of the present application. [Figure 4]FIG. 4 is a flow chart diagram of a method 400 for finding patterns between communication timing and corresponding measurements, according to an embodiment of the present application. [Figure 5] FIG. 5 is a flowchart diagram of a method 500 for finding a timing to transmit health-specific information according to an embodiment of the present application. [Figure 6A] FIG. 6 is a flowchart diagram of a method 600 for finding a timing to transmit health-specific information according to an embodiment of the present application. [Figure 6B] FIG. 6B illustrates a variation of the flowchart diagram of method 600 shown in FIG. 6A. [Figure 7] FIG. 7 is a flowchart diagram of a method 700 for finding a timing to transmit health-specific information according to an embodiment of the present application. [Figure 8] FIG. 8 illustrates a method 800 for finding a timing for transmitting health-specific information according to an embodiment of the present application. DETAILED DESCRIPTION OF THE INVENTION

[0015] Some embodiments of the present application are described in detail below. However, in addition to the following description, the present invention can be applied to other embodiments, and the scope of the present invention is not limited by these, but by the scope of the claims. Furthermore, for the sake of understanding and clarity of the description, some components in the drawings are not necessarily drawn to scale, may be exaggerated relative to other components, and may be irrelevant. If the relationship between two steps is not described, the order of their execution is not bound by the order shown in the flowchart.

[0016] One aspect of the present application is to provide a network system for notifying a user at calculated timings, which collects physiological measurement information (sometimes referred to herein as physiological measurement, measurement information, or physiological information) from a user via a sensing device, generates health-specific information for the user based on the physiological measurement information and corresponding timestamps to remind the user of available actions or review events, finds a communication timing for delivering the health-specific information to the user's computer (smartphone, smart band, and / or smart watch), collects the user's response regarding the timing, and repeats the process to refine the timing discovery strategy based on the response.

[0017] Please refer to Figure 1A, which shows a block diagram of a network system 100 for providing notifications to users at calculated times, according to one embodiment of the present application. The network system 100 may include a server 110, a network 120, a user sensing device 130, a user client computer 140, an optional artificial intelligence (AI) server 112, and an optional message server 150.

[0018] Network 120 is configured to support data exchange between server 110 and sensing device 130, and between server 110 and client computer 140. For example, network 120 may comprise an access network such as a WiFi / IEEE 802.11 network, a 3G / 4G / 5G / 6G compliant network, an IEEE 802.3 network, a PSTN network, an optical fiber network, and / or an xDSL network. Server 110, sensing device 130, client computer 140, optional AI server 112, and / or message server 150 may be configured to connect to the access network. Furthermore, network 120 may comprise a backbone network, e.g., a telecommunications network, for connecting the access networks.

[0019] The sensing device 130 is configured to measure physiological characteristics of the user, such as glucose, blood pressure, heart rate, respiratory rate, blood oxygen level, and / or electrocardiogram measurements, etc. Accordingly, the sensing device 130 comprises one or more specific sensors for measuring the physiological characteristics of the user and a networking device for connecting to the network 120.

[0020] In one example, sensing device 130 adds a local timestamp to the measurement information before transmitting the measurement information to server 110. The clock maintained by sensing device 130 may be synchronized with server 110 or a common source such as a timing service provider. Thus, the timestamp included in the measurement information received by server 110 is synchronized with the server's internal clock.

[0021] In another example, the sensing device 130 does not add a local timestamp to the measurement information. Instead, the server 110 generates a timestamp based on its internal clock for the received measurement information. Of course, the server 110 can always generate a timestamp based on its internal clock for the received measurement information, whether or not it is timestamped.

[0022] The measurement information may be transmitted via a message server 150. The message server 150 may implement one or more public messaging services for the server 110 and the sensing device 130. For example, the message server 150 may be an email server, an instant messaging server (such as Line, WhatsApp®, Skype®, or Facetime®), or a message queue server. One of the features provided by the message server 150 is asynchronous transmission. This means that messages, i.e., measurement information, are temporarily stored in the message server 150 and may be retransmitted by the message server 150 until they are safely received by the server 110. Another feature provided by the message server 150 is encrypted transmission to ensure the security of the transmission. However, the message server 150 is optional in the present embodiment. The sensing device 130 may transmit the user's measurement information to the server 110 without the intervention of the message server 150.

[0023] After receiving multiple physiological measurement information from the sensing device 130, the server 110 may analyze the physiological measurement information and their corresponding timestamps to obtain the user's behavior or the user's periodic schedule. The accumulated timestamps and corresponding physiological measurement values ​​can be used to infer the user's daily tasks. Daily or weekly tasks may include meals (breakfast, lunch, afternoon tea, dinner), exercise, work, and sleep. A collection of daily tasks constitutes the user's behavior or daily schedule. Similarly, a collection of weekly or monthly tasks constitutes the user's behavior or periodic schedule. In one embodiment, the analysis of the physiological information and the corresponding timestamps may be performed by applying one or more machine learning algorithms / models to obtain the user's periodic schedule. The periodic schedule includes at least one task. One or more types of physiological measurement information may be considered in the analysis.

[0024] In one embodiment, a first machine learning algorithm / model is used to infer communication timing based on the user's physiological measurements and corresponding timestamps. In an alternative embodiment, a second machine learning algorithm / model is used to infer communication timing based on the user's periodic schedule, where the user's periodic schedule is the result of an analysis of the user's physiological measurements and corresponding timestamps. The analysis may further utilize a third machine learning algorithm / model to infer the user's periodic schedule based on the user's physiological measurements and corresponding timestamps.

[0025] When these tasks (contained in the behavioral and periodic schedules) and their associated timing are available, task-specific health-specific information may be sent to the user's client computer to remind the user what to do with an actionable plan or to review events with the user as they occur. In this way, the network system 100 can provide accurate information to the user at the appropriate time, allowing the user to proactively and thoughtfully make better choices.

[0026] The server 110 may itself perform analysis corresponding to the user's measurement information and timestamp. In an alternative embodiment, the server 110 may further utilize the AI ​​server 112 to generate task-specific health-specific information for the user. In one example, the AI ​​server 112 may be configured to train or infer various machine learning algorithms / models used in embodiments of the present application. The machine learning model may be a combination of various neural networks, such as a convolutional neural network, a deep neural network, or a transformer network. Those skilled in the art will appreciate that the AI ​​server 112 may be provided as a cloud service by a vendor. For example, Google® cloud services, Amazon® AWS services, Microsoft® Azure® services, etc. offer various neural network services that may be considered as the AI ​​server 112 in this application. However, even without the AI ​​server 112, the server 110 may have sufficient computing power to implement the machine learning algorithms / models presented in this application.

[0027] The task-specific health-specific information may include an actionable plan and / or a review of measurement information. For example, the actionable plan may include exercise, rest, sleep advice, and / or dietary suggestions. The review of measurement information may include charts and descriptions of physiological information. Additionally, the review may include a survey related to the task, for example, a survey regarding sleep quality after sleep.

[0028] In one embodiment, some of the content of the task-specific health-specific information may be generated by a large-scale language model (LLM) implemented by the AI ​​server 112. Based on the measurement information and corresponding timestamps, the server 110 may create or provide appropriate prompts to the LLM implemented by the AI ​​server 112 and receive the health-specific information from the AI ​​server 112. For example, the LLM may be a commercially available product such as ChatGPT®, DeepSeek®, Perplexity®, Co-pilot, and / or Gemini®.

[0029] Similarly, one or more machine learning models may be utilized to learn and predict when to send health-specific information to a user. The machine learning model may be a combination of multiple neural networks. In an initial stage, a general machine learning model may be used to output at least one timing for sending health-specific information to a specific user. After collecting user responses, a user-specific, tailored machine learning model may be derived, fine-tuned, and / or retrained from the general machine learning model based on the user's responses to the health-specific information. The general machine learning model may also be derived, fine-tuned, improved, refined, and / or retrained based on multiple users' responses to the health-specific information.

[0030] The server 110 is configured to transmit the health-specific information to the user when the health-specific information and timing become available. The transmission of the health-specific information may occur end-to-end from the server 110 over the network 120 to the user's client computer 140. The transmission is synchronized so that the server 110 knows whether the health-specific information has been delivered to the client computer 140 in a timely manner. The client computer 140 may be a smartphone, a wearable electronic device such as a smart watch or smart goggles, a tablet computer, a laptop, and / or any computer that connects to the network 120.

[0031] Alternatively, transmission of the health-specific information may be completed through message server 150. Because the transmission is asynchronous, message server 150 may notify server 110 when client computer 140 receives the health-specific information.

[0032] The user may respond to the health-specific information with their opinion regarding the timing. For example, the user may state one of two options regarding the timing: good or bad. In another example, the user may give the timing a score from 1 to 10. The user's response may be transmitted back from the client computer 140 to the server 110. If the timing matches the time of receipt by the client computer 140, the server 110 can pair the timing of the health-specific information with the associated user response. In other words, the timing of the health-specific information is labeled by the user's response as a sample for the machine learning model.

[0033] Those skilled in the art of AI will appreciate that the samples can be further used to derive, fine-tune, improve, refine, and / or retrain general machine learning models and / or user-specific tailored machine learning models. In the next iteration, the tailored machine learning model can be used to infer the timing of the next health-specific information based on the specified user's measurement information and associated timestamps.

[0034] Please refer to Figure 1B, which shows a modification of the network system 100 shown in Figure 1A for notifying a user at a calculated timing according to another embodiment of the present application. Compared to the network system 100 shown in Figure 1A, the sensing device 130 shown in Figure 1B is not directly connected to the network 120. Instead, the sensing device 130 is connected to the server 110 via the user's client computer 140.

[0035] For example, sensing device 130 may be a wearable device connected to client computer 140 via short-range wireless communication, such as Bluetooth® or Near Field Communication (NFC), or via wired communication, such as USB®, Firewire®, or UART (RS-232 / 422). An application or driver program corresponding to sensing device 130 may be configured to receive measurement information from sensing device 130 and transmit the measurement information, with or without a timestamp, to server 110 over network 120. As with the above-described embodiment, transmission of the measurement information may be via message server 150 or may be end-to-end from client computer 140 to server 110.

[0036] Please refer to Fig. 1C, which shows a modified example of the network system 100 shown in Fig. 1A for notifying a user at a calculated timing according to another embodiment of the present application. Compared to the network system 100 shown in Fig. 1A, the server 110 communicates with the detection device 130 and the client computer 140 via different networks.

[0037] 1C may be applied to remote monitoring. That is, the sensing device 130 is configured to measure physiological information of a first user, and the client computer 140 is configured to report to a second user. For example, the first user may be a patient, and the second user may be a caretaker of the patient.

[0038] To transmit the health-specific information to the client computer 140, if the client computer 140 is a smartphone or feature phone, the server 110 may utilize a short message service / multimedia message service (SMS / MMS) provided by a carrier. Thus, the server 110 may transmit the health-specific information to the SMS / MMS server 152, which may then forward the health-specific information to the client computer 140. Conversely, the user of the client computer 140 may also respond to the server 110 regarding the timing of the health-specific information using the SMS / MMS server 152.

[0039] The SMS / MMS server 152 may provide a scheduling function for the server 110, i.e., the server 110 can upload the health-specific information and pre-set the timing for sending the health-specific information. Thus, at a predetermined time, the SMS / MMS server 152 can deliver the health-specific information packaged in a short message or a multimedia message to the client computer 140 via the telecommunications network 122.

[0040] Please refer to FIG. 2, which shows a block diagram of a computer 200 according to one embodiment of the present application. The computer 200 may be embodied as the server 110, the AI ​​server 112, the sensing device 130, the client computer 140, the message server 150, and / or the SMS / MMS message server 152. Some components of the computer 200 may be omitted or modified depending on their role. For example, the AI ​​server 112 may not have a display 240 and input devices 280 such as a keyboard and mouse. The graphics processing unit (GPU) 230 of the AI ​​server 112 may be modified into several tens of neural network processing units (NPUs).

[0041] The computer 200 includes at least one central processing unit (CPU) 210, a memory module 220 for system operation, an optional GPU 230 for generating content to be displayed on the display 240, peripheral connection devices 250 implementing industry-standard interfaces such as PCI, PCI-Express, SCSI, SATA, USB, etc., a networking device 260 for connecting to the network 120 and / or the telecommunications network 122, a storage device 270 for storing an operating system, application programs, and data for implementing the processes provided herein, one or more input devices 280, and one or more output devices 290. The operating system executed by the CPU 230 is configured to control the computer 200. Various driver programs and application programs may be used under the operating system to implement the processes of the methods embodied herein.

[0042] In one embodiment, the sensing device 130 may have, as one of the input devices 280, a specific sensor for measuring a user's physiological characteristics, including but not limited to, glucose, blood pressure, heart rate, respiratory rate, blood oxygen level, and / or electrocardiogram measurements. The client computer 140 may have a display 240 or a touchscreen for displaying the health-specific information and / or may have a speaker for reading out the health-specific information. Furthermore, the client computer 140 may have an input device 280, such as a touchscreen, touch panel, keyboard, or buttons, for receiving a user's response.

[0043] To enhance computing power, server 110, AI server 112, message server 150, and / or SMS / MMS server 152 may be configured as a cluster of multiple physical or virtual computers 200. Those skilled in the art are familiar with computer architecture and computer configurations and will understand that variations of computer 200 shown in Figure 2 may be applied to the various computers shown in Figures 1A-1C.

[0044] Please refer to Figure 3, which shows a flowchart diagram of a method 300 for finding a timing to transmit health-specific information according to one embodiment of the present application. The method 300 may be realized by the network system 100 shown in Figure 1. In particular, the method 300 may be embodied as computer instructions and data executed by the server 110 shown in Figure 1. If there is no causal relationship between the two steps shown in Figure 3, the present application does not limit the order in which the two steps are performed. The method 300 may start from step 305.

[0045] Step 305: Receive the user's physiological information and an associated timestamp from the sensing device. The associated timestamp may be included in the physiological information transmitted from the sensing device or may be generated locally.

[0046] Step 310: Analyze the physiological information and associated timestamps to obtain the user's behavior and / or daily schedule and generate health-specific information for the user.

[0047] Step 315: Utilizing the first machine learning model together with the user's behavior and / or daily schedule, find a first communication timing for health-specific information related to physiological information.

[0048] The first machine learning model includes a set of methods for extracting necessary information from input information. Depending on the multiplicity of input variables and the predicted output, a method called a "model" is selected to learn patterns from the input data. This model is then applied to newly input variables to obtain a predicted output.

[0049] In one example, the behavioral and / or daily schedule includes one or more upload times of the user as input data to a first machine learning model, which may be trained to apply a "k-means" method to the upload times to output a first timing (e.g., the most common time).

[0050] Those skilled in the art will appreciate that the first machine learning model may be multimodal and accept different types of input data in other scenarios. In the above example, the user's upload time is used as the only input data. However, the first machine learning model may also accept different types of input data other than the upload time.

[0051] Step 320: Send the health-specific information to the client computer at a first timing.

[0052] Step 325: Receive a first response from the client computer for a first timing.

[0053] In one embodiment, the collected first response may include uploaded or received metadata (e.g., a timestamp) corresponding to the first response. The timestamp and content of the first response may be further input into a first machine learning model or algorithm to calculate a first timing mean time for communication with the user.

[0054] Step 330: Improve the first machine learning model with the first response.

[0055] For example, if a user interacts with the server 110 over several months, e.g., from June 2025 to December 2025, the records may include multiple sets of physiological information, daily schedules, behaviors, health-specific information, and first responses and their corresponding timings and / or timestamps. Some or all of these records may be treated as input data variables to a first machine learning model. The first machine learning model may include a "k-means" method, and the best or most common time may be signaled by the first machine learning model.

[0056] On the day following a specified user's interaction, say the first day of 2026, records collected from June to December 2025 are applied to a first machine learning model to find the most recent first timing of communication with the user.

[0057] While in the above example, records collected over several months containing a large number of sets of first responses were used as input data, in an alternative example, the first machine learning model may be refined using only one set of first responses. In other words, the refinement process may be performed each time method 300 is performed. Alternatively, the refinement process may be performed once after method 300 has been performed multiple times.

[0058] Please refer to Figure 4, which shows a flowchart diagram of a method 400 for finding patterns between communication timings and corresponding measurements according to one embodiment of the present application. The method 400 may be implemented by the network system 100 shown in Figure 1. In particular, the method 400 may be embodied as computer instructions and data executed by the server 110 shown in Figure 1. If there is no causal relationship between the two steps shown in Figure 4, the present application does not limit the order in which the two steps are performed. The method 400 may start from step 405.

[0059] Step 405: Obtain training data and corresponding timestamps related to user health measurements from sensing devices. In one embodiment, the training data may be manually or automatically labeled with default timing. In an alternative embodiment, the training data may include pre-labeled samples. Each sample may include a measurement and corresponding timestamp, as well as a label for preferred communication timing.

[0060] Step 410: Run or train a second machine learning algorithm / model based on the training data to find patterns between communication timing and associated measurements.

[0061] Step 415: A query is sent to the client computer for the user to confirm or reject the inferred communication timing regarding the relevant measurements and corresponding timestamps. If the user confirms the inferred timing, the flow proceeds to step 420.

[0062] Step 420: Apply the newly confirmed second timing to subsequent communications with the client computer.

[0063] Step 425: Collect a second response from the client computer regarding the second communication timing. The second response may include a preferred communication timing.

[0064] Step 430: Improve a second machine learning algorithm / model with the second communication timing and the second response.

[0065] Please refer to Figure 5, which shows a flowchart diagram of a method 500 for finding a timing to transmit health-specific information according to one embodiment of the present application. Method 500 is a variation of method 300 shown in Figure 3. This method 500 is configured to illustrate an embodiment where the physiological information is a user's glucose information. The sensing device may be a continuous glucose monitoring (CGM) device that tracks blood glucose levels throughout the day without fingersticks.

[0066] Similar to method 300, method 500 may be implemented by the network system 100 shown in Fig. 1. In particular, method 500 may be embodied as computer instructions and data executed by the server 110 shown in Fig. 1. If there is no causal relationship between the two steps shown in Fig. 5, the present application does not limit the order in which the two steps are performed. Method 500 may start from step 505.

[0067] Step 505: Receive the user's glucose information and associated timestamp from the sensing device. The glucose information may include one or more blood glucose measurements. After a meal, blood glucose levels rise. When the user is hungry, blood glucose levels are low. Based on the blood glucose levels and associated timestamp, the server can determine when the user ate.

[0068] Step 510: The user's glucose information and associated timestamps are analyzed to obtain the user's behavior and daily eating schedule. In one embodiment, the analysis may be based on an algorithm other than a machine learning model. The algorithm may determine the daily eating schedule. In an alternative embodiment, the analysis may include a machine learning model that associates blood glucose levels and associated timestamps with the daily eating schedule.

[0069] Step 515: Infer a glucose machine learning model according to the behavior and daily meal schedule to find at least one communication timing of health-specific information related to the glucose information.

[0070] In one embodiment, the glucose machine learning model may provide pre-meal timing, which may correspond to meal-related information such as meal preparation information, including reducing salt and sugar intake.

[0071] In an alternative embodiment, the glucose machine learning model may provide alternative timings after meals that may correspond to exercise suggestions, such as taking a walk, to help the user achieve their daily exercise plan.

[0072] In another embodiment, the glucose machine learning model may provide timing for post-meal reports that are generated according to physiological measurements of blood glucose levels after a meal, allowing the user to review glucose variability in the post-meal report, and allowing the user's nutritionist to adjust or fine-tune their meal plan.

[0073] Step 520: Transmit health-specific information to the user's client computer at least one communication timing. In one embodiment, the server may transmit meal-related information at a timing before a meal and exercise suggestions at another timing after a meal.

[0074] Step 525: Collect a response from the client computer regarding at least one communication timing.

[0075] Step 530: Improve the glucose machine learning model with the response from the client computer.

[0076] Please refer to Figure 6A, which shows a flowchart diagram of a method 600 for finding a timing to transmit health-specific information according to one embodiment of the present application. Method 600 is a variation of method 300 shown in Figure 3. This method 600 is configured to illustrate an embodiment where physiological information is related to a user's sleep information.

[0077] In one embodiment, the sensing device may be a smart wearable device such as a smart ring, a smart watch, or a smart band. This type of smart wearable device may include an oximeter, a blood pressure monitor, a heart rate sensor, and / or a motion sensor. Therefore, the sensed physiological information can determine whether the user has fallen asleep. The determination of the user's sleep state may be performed by the sensing device 130 or the server 110. That is, the sensing device 130 may send a status indicating whether the user is sleeping to the server 110. Alternatively, the sensing device 130 may send the measured physiological information and motion information to the server 110, and the server 110 may determine whether the user is sleeping based on the information sent by the sensing device 130.

[0078] Similar to method 600, method 600 may be implemented by the network system 100 shown in Fig. 1. In particular, method 600 may be embodied as computer instructions and data executed by the server 110 shown in Fig. 1. If there is no causal relationship between the two steps shown in Fig. 6A, the present application does not limit the order in which the two steps are performed. Method 600 may start from step 605.

[0079] Step 605: Receive physiological information related to the user's activity level and associated timestamps from the sensing device. In one embodiment, the physiological information may include motion information of the user's body parts, such as a hand, wrist, or fingers.

[0080] Step 610: Analyze the physiological information related to activity level and the associated timestamps to obtain the user's behavior and daily sleep schedule.

[0081] Step 615: Infer a sleep machine learning model with the behavior and daily sleep schedule to find timings for communication of at least one of health-specific information related to physiological information associated with activity level.

[0082] In one embodiment, the sleep machine learning model may provide timing for bedtime reminders before bedtime. The bedtime reminders may include sleep-related information such as sleep rituals and tooth brushing. In an alternative embodiment, the sleep learning model may provide timing for a sleep quality survey after waking up to gather input on the user's sleep quality. In addition to a sleep schedule, the sleep machine learning model may also provide timing for rest reminders or nap reminders to the user.

[0083] Step 620: Send health-specific information to the user's client computer at at least one communication timing. The health-specific information may be any of the above-mentioned bedtime reminder, rest reminder, nap reminder, and wake-up questionnaire.

[0084] Step 625: Collect a response from the client computer regarding at least one communication timing.

[0085] Step 630: Improve the sleep machine learning model with the responses from the client computers.

[0086] Please refer to Figure 6B, which shows a variation of the flowchart diagram of the method 600 shown in Figure 6A. As shown in Figure 6B, steps 605 and 610 are replaced with step 607. The other processes shown in Figure 6A are also adopted in the flowchart shown in Figure 6B. The method 600 shown in Figure 6B may start with step 607.

[0087] In step 607, the server 110 is configured to receive the user's behavior and daily sleep schedule from a sensing device that measures physiological information related to the user's activity level. In other words, the analysis is performed by the sensing device 130, not the server 110.

[0088] In one example, a daily sleep schedule was output as the analysis result. The user's daily sleep schedule is from 10:00 PM to 7:00 AM. Therefore, the timing of the bedtime reminder provided by the sleep machine learning model is 8:00 PM, i.e., two hours before going to bed. In addition, the timing of the first wake-up questionnaire provided by the sleep machine learning model is 7:15 AM, i.e., 15 minutes after waking up.

[0089] However, based on the user's response, the 7:15 AM wake-up survey timing seems inconvenient to the user, so the sleep machine learning model adjusts the timing from 7:15 AM to 1:00 PM based on the user's response.

[0090] Please refer to FIG. 7, which shows a flowchart diagram of a method 700 for finding a timing to transmit health-specific information according to one embodiment of the present application. The method 700 may be realized by the network system 100 shown in FIG. 1. In particular, the method 700 may be embodied as computer instructions and data executed by the server 110 shown in FIG. 1. If there is no causal relationship between the two steps shown in FIG. 7, the present application does not limit the order in which the two steps are performed. The method 700 may start from step 705.

[0091] Step 705: Receive physiological measurement information of the first user from the sensing device via the network.

[0092] Step 710: Obtain a first communication timing of the first health-specific information according to the physiological measurement information and the corresponding timestamp. The timestamp may be received together with the physiological measurement information or may be locally generated. The first communication timing may be an inference result of a first machine learning model.

[0093] Step 715: Generate first health-specific information by inferring a large-scale language model according to the physiological measurement information and the corresponding timestamp. Alternatively, the first health-specific information may be generated by a deterministic algorithm according to the physiological measurement information and the corresponding timestamp.

[0094] Step 720: Send the first health-specific information to a client computer at a first timing. The client computer may be used by the user of the sensing device. Alternatively, the client computer may be used by another user who is a caretaker of the user of the sensing device. In one example, the sensing device and the client computer may be in different locations. The sensing device may be in a room where the user is staying, and the client computer may be in an office where another user is staying. The sensing device and the client computer are connected to different access networks.

[0095] Step 725: Collect a first response from the client computer for a first timing.

[0096] Step 730: Improve the first machine learning model with the first response.

[0097] Please refer to FIG. 8, which illustrates a method 800 for finding a timing to transmit health-specific information according to one embodiment of the present application. The method 800 is a variation of the method 700. The method 800 may be implemented by the network system 100 illustrated in FIG. 1. In particular, the method 800 may be embodied as computer instructions and data executed by the server 110 illustrated in FIG. 1. If there is no causal relationship between the two steps illustrated in FIG. 8, the present application does not limit the order in which the two steps are performed.

[0098] Method 800 reuses some steps from method 700. Method 800 may start at step 705. After step 705, the flow may proceed to step 805.

[0099] Step 805: Analyze the physiological measurement information and corresponding timestamps to obtain the first user's periodic schedule. The periodic schedule may be a daily sleep schedule, a daily meal schedule, or a weekly exercise schedule. Each periodic schedule includes at least one task and its time period, such as sleep time, breakfast, lunch, dinner, morning jogging, etc. Next, the flow may proceed to step 810 or step 812.

[0100] Step 810: Infer a second machine learning model according to the periodic schedule of the first user to obtain a first communication timing. The second machine learning model shown in Figure 8 may be different from the first machine learning model shown in Figure 7 because its input is the analysis result, i.e., the user's periodic schedule, rather than physiological measurements and corresponding timestamps.

[0101] In one embodiment, the first communication timing may correspond to a reminder of a periodically scheduled task, such as a sleep ritual before bedtime or a meal plan presented before a meal. Flow may then proceed to step 812 or step 815.

[0102] Step 812: Infer a second machine learning model according to the first user's periodic schedule to obtain a second communication timing. In one embodiment, the second communication timing may correspond to a report of physiological measurement information received after an action or at least one task of the periodic schedule, such as a glucose fluctuation after a meal or a survey on sleep quality after waking up. The flow may then proceed to step 815.

[0103] Optionally, the flow may proceed to step 720, step 725, and step 830. Step 830: Refine a second machine learning model with the first response. If a second communication timing is obtained in step 812, the flow may further proceed to step 835, step 840, step 845, and step 850.

[0104] Step 835: Generate second health-specific information based on physiological measurement information received during or after the performance of the periodically scheduled task. For example, a glucose variability report including blood glucose levels measured during or after a meal should be generated after the task / meal.

[0105] Step 840: The second health information specifying information is sent to the client computer at a second timing.

[0106] STEP 845: Collect a second response from the client computer for a second timing.

[0107] Step 850: Improve a second machine learning model with the second response.

[0108] According to an embodiment of the present application, there is provided a method for finding communication timings related to physiological information and associated timestamps. The method includes receiving physiological measurement information of a first user from a detection device via a network, obtaining a first communication timing of first health-specific information according to the physiological measurement information and the corresponding timestamp, and transmitting the first health-specific information to a client computer at the first timing. The corresponding timestamp may be either a timestamp of uploading the physiological measurement information at the detection device or a timestamp of receiving the physiological measurement information at a server.

[0109] Preferably, for preparing the first health-specific information, the method further includes generating the first health-specific information by inferring a large-scale language model according to the physiological measurement information and the corresponding timestamp.

[0110] Preferably, to achieve remote sensing, the client computer is operated by a second user, and the sensing device and the client computer are connected to different access networks that are each part of the network.

[0111] Preferably, in order to utilize artificial intelligence, the obtaining step further includes inferring a first machine learning model according to the physiological measurement information and the corresponding timestamp to obtain the first communication timing.

[0112] Preferably, to accommodate user needs, the method further includes collecting a first response from the client computer over the network at a first timing, and improving the first machine learning model with the first response.

[0113] Preferably, in order to associate the timing with the task, the obtaining step further includes analyzing the physiological measurement information and corresponding timestamps to obtain a periodic schedule of the first user, and inferring a second machine learning model according to the periodic schedule of the first user to obtain a first communication timing.

[0114] Preferably, to accommodate user needs, the method further includes collecting a first response from the client computer over the network at a first timing, and improving the second machine learning model with the first response.

[0115] Preferably, to utilize artificial intelligence, the analyzing step further includes inferring a third machine learning model according to the physiological measurement information and corresponding timestamps to obtain the periodic schedule of the first user.

[0116] Preferably, to provide a reminder of an upcoming task, the periodic schedule includes at least one task, the first communication timing precedes the at least one task, and the first health-specific information is configured to remind the first user of an action prior to the at least one task.

[0117] Preferably, to provide a report of the completed tasks, the method further includes: obtaining a second communication timing of second health-specific information according to a periodic schedule; generating the second health-specific information based on physiological measurement information received during or after the performance of the at least one task; and transmitting the health-specific information to the client computer at the second communication timing, wherein the second communication timing is after the at least one task.

[0118] In one embodiment of the present application, a server computer for finding communication timings related to physiological information and associated timestamps is provided, the server computer including: a networking device connected to a detection device and a client computer via a network; and a processor configured to execute instructions stored in a non-volatile memory to cause the networking device to receive physiological measurement information of a first user from the detection device, obtain a first communication timing of the first health-specific information according to the physiological measurement information and the corresponding timestamp, and cause the networking device to transmit the health-specific information to the client computer at the first timing.

[0119] Preferably, for preparing the first health-specific information, the processor is further configured to generate the first health-specific information by inferring a large-scale language model according to the physiological measurement information and the corresponding timestamp.

[0120] Preferably, to achieve remote sensing, the client computer is operated by a second user, and the sensing device and the client computer are connected to different access networks that are each part of the network.

[0121] Preferably, to utilize artificial intelligence, the obtaining further includes inferring a first machine learning model according to the physiological measurement information and the corresponding timestamp to obtain the first communication timing.

[0122] Preferably, to adapt to the needs of the user, the processor is further configured to collect a first response from the client computer over the network for a first timing, and refine the first machine learning model with the first response.

[0123] Preferably, in order to associate the timing with the task, the obtaining further includes analyzing the physiological measurement information and corresponding timestamps to obtain a periodic schedule of the first user, and inferring a second machine learning model according to the periodic schedule of the first user to obtain a first communication timing.

[0124] Preferably, to adapt to the needs of the user, the processor is further configured to collect a first response from the client computer over the network with respect to a first timing, and refine the second machine learning model with the first response.

[0125] Preferably, to utilize artificial intelligence, the analyzing further includes inferring a third machine learning model according to the physiological measurement information and corresponding timestamps to obtain the periodic schedule of the first user.

[0126] Preferably, to provide a reminder of an upcoming task, the periodic schedule includes at least one task, the first communication timing precedes the at least one task, and the first health-specific information is configured to remind the first user of an action prior to the at least one task.

[0127] Preferably, to provide a report of the completed tasks, the processor is further configured to obtain a second communication timing of the second health-specific information according to a periodic schedule, generate the second health-specific information based on the physiological measurement information received during or after the performance of the at least one task, and transmit the health-specific information to the client computer at the second communication timing, wherein the second communication timing is after the at least one task.

[0128] According to one embodiment of the present application, there is provided a network system for finding timing of communication relating to physiological information and associated timestamps, the network system comprising a server, a sensing device, and a client computer.

[0129] Based on the method and server provided herein, a prompt notification mechanism is presented to a patient or user that responds to viable options and provides timely notifications to respect their decisions. The mechanism can automatically analyze the periodic schedule and communication timing based on physiological measurement information and corresponding timestamps. Therefore, health-specific information, such as reminders, action suggestions, and / or summaries of physiological measurement information, can be notified at the communication timing found by the mechanism. This allows the patient or user to follow instructions at communication timings that do not conflict with the user's other activities. Also, response reports of physiological measurement information can be delivered to the patient or user at the appropriate time. Following this mechanism, the mechanism can provide accurate information to the user at the appropriate time, allowing the user to make better, proactive, and thoughtful choices.

[0130] While the present invention has been described in what are presently considered to be the most practical and preferred embodiments, it is to be understood that the invention is not limited to the described embodiments, but rather is intended to cover various modifications and similar arrangements included within the spirit and scope of the appended claims, which are to be accorded the broadest interpretation so as to encompass all such modifications and similar structures.

Claims

1. 1. A method for finding timing of communication relating to physiological information and associated timestamps, comprising: receiving physiological measurement information of the first user from the sensing device via a network; obtaining a first communication timing of a first health specification information according to the physiological measurement information and a corresponding time stamp; transmitting the first health specification information to a client computer at the first communication timing; A method comprising:

2. The method of claim 1 , further comprising generating the first health-specific information by inferring a large-scale language model according to the physiological measurement information and corresponding timestamps.

3. The method of claim 1 , wherein the client computer is operated by a second user, and the sensing device and the client computer are each connected to different access networks that are part of the network.

4. The method of claim 1 , wherein the obtaining step further comprises: inferring a first machine learning model according to the physiological measurement information and the corresponding timestamp to obtain the first communication timing.

5. collecting a first response from the client computer via the network regarding the first communication timing; improving the first machine learning model with the first response; The method of claim 4 further comprising:

6. The obtaining step includes: analyzing the physiological measurement information and the corresponding timestamps to obtain a periodic schedule of the first user; Inferring a second machine learning model according to the periodic schedule of the first user to obtain the first communication timing; The method of claim 1 further comprising:

7. collecting a first response from the client computer via the network regarding the first communication timing; improving the second machine learning model with the first response; The method of claim 6 further comprising:

8. The method of claim 6 , further comprising: inferring a third machine learning model according to the physiological measurement information and the corresponding timestamps to obtain a periodic schedule of the first user.

9. 7. The method of claim 6, wherein the periodic schedule includes at least one task, the first communication timing occurs prior to the at least one task, and the first health-specific information is configured to remind the first user of an activity prior to the at least one task.

10. obtaining a second communication timing of the second health-specific information according to the periodic schedule; generating the second health-specific information based on the physiological measurement information received during or after performance of the at least one task; transmitting health information specifying information to the client computer at the second communication timing, the second communication timing being after the at least one task; The method of claim 9 further comprising:

11. a server computer for finding timing of communications relating to physiological information and associated timestamps, a networking device connected to the sensing device and the client computer via a network; 1. A processor configured to execute instructions stored in a non-volatile memory, comprising: causing the networking device to receive physiological measurement information of a first user from the sensing device; Obtaining a first communication timing of first health specification information according to the physiological measurement information and the corresponding time stamp; causing the networking device to transmit health information specifying information to a client computer at the first communication timing; A processor to achieve this A server computer comprising:

12. The server computer of claim 11 , wherein the processor is further configured to generate the first health-specific information by inferring a large-scale language model according to the physiological measurement information and corresponding timestamps.

13. 12. The server computer of claim 11, wherein the client computer is operated by a second user, and the sensing device and the client computer are each connected to different access networks that are part of the network.

14. The server computer of claim 11 , wherein the obtaining further comprises: inferring a first machine learning model according to the physiological measurement information and the corresponding time stamp to obtain the first communication timing.

15. The processor: collecting a first response from the client computer over the network with respect to the first timing; Improving the first machine learning model with the first response.

15. The server computer of claim 14, further configured to:

16. The obtaining includes: analyzing the physiological measurement information and the corresponding timestamps to obtain a periodic schedule of the first user; Inferring a second machine learning model according to the periodic schedule of the first user to obtain the first communication timing; The server computer of claim 11 further comprising:

17. The processor: collecting a first response from the client computer over the network with respect to the first timing; Improving the second machine learning model with the first response.

17. The server computer of claim 16, further configured to:

18. The analyzing step may include: Inferring a third machine learning model according to the physiological measurement information and the corresponding time stamp to obtain the periodic schedule of the first user. The server computer of claim 16 further comprising:

19. 17. The server computer of claim 16, wherein the periodic schedule includes at least one task, the first communication timing precedes the at least one task, and the first health-specific information is configured to remind the first user of an action before the at least one task.

20. The processor: obtaining a second communication timing of the second health-specific information according to the periodic schedule; generating the second health-specific information based on the physiological measurement information received during or after performance of the at least one task; The health information specifying information is transmitted to the client computer at the second communication timing. It is further structured as follows:

20. The server computer of claim 19, wherein the second communication timing is after the at least one task.

21. 12. A network system for finding timing of communications relating to physiological information and associated timestamps, comprising: a server computer, a sensing device, and a client computer according to claim 11.

Images