Personalized closed-loop infusion system and method

Abstract

A personalized closed-loop infusion system and method. Relevant information of patients is uploaded to a cloud server (104), and is grouped on the basis of identifiable features of the patients; the relevant information of different patients, such as blood glucose data records and insulin infusion data records, is compared in a same group; and when the relevant information of the patients is the same or similar, insulin algorithm parameters of patients with well-controlled blood glucose can be used for modifying an insulin algorithm of patients with poorly-controlled blood glucose, and a closed-loop artificial pancreas system can quickly obtain personalized insulin algorithms suitable for the patients without the need of long-period learning of physiological characteristics and lifestyle habits of the patients, thereby benefiting diabetes treatment of the patients.

Description

Personalized closed-loop infusion systems and methods

[0001] Cross-reference of related applications

[0002] This application claims the benefit of and priority of the following patent application: PCT patent application filed on May 27, 2024, application number PCT / CN2024 / 095396. Technical Field

[0003] This invention relates primarily to the field of medical devices, and in particular to a personalized closed-loop infusion system and method. Background Technology

[0004] In healthy individuals, the pancreas automatically secretes the necessary insulin / glucagon based on blood glucose levels, thus maintaining a reasonable range of blood sugar fluctuations. However, in diabetic patients, pancreatic function is abnormal, and the pancreas is unable to secrete the required insulin normally. Diabetes is a metabolic disease and is a lifelong condition. Current medical technology cannot cure diabetes; it can only control the occurrence and development of diabetes and its complications by stabilizing blood sugar levels.

[0005] Diabetic patients need to monitor their blood glucose levels before injecting insulin. Currently, most monitoring methods involve continuous blood glucose monitoring using in-vivo glucose monitoring devices. These devices use disposable percutaneous sensors inserted into the skin to measure the glucose concentration in the interstitial fluid and transmit the data in real time to an external device for patient viewing. This monitoring method is called Continuous Glucose Monitoring (CGM). Based on the blood glucose levels detected by CGM, the insulin pump automatically adjusts the required insulin infusion volume and delivers it subcutaneously, thus forming a closed-loop artificial pancreas system.

[0006] Currently, the insulin algorithm built into closed-loop artificial pancreas systems learns from individual patient blood glucose data and insulin infusion patterns to adjust the insulin algorithm parameters, thereby achieving precise control of the patient's blood glucose levels. However, this data-driven adjustment has limitations. Closed-loop artificial pancreas systems require a long period of learning about the patient's physiological characteristics and lifestyle habits to obtain insulin algorithm parameters suitable for that patient. This may lead to delayed or inaccurate adjustments to the insulin algorithm parameters, affecting the patient's treatment.

[0007] Therefore, there is an urgent need for a personalized closed-loop infusion system and method that can quickly adjust insulin algorithm parameters and is applicable to individual patients.

[0008] Summary of the Invention

[0009] This invention discloses a personalized closed-loop infusion system and method. The system uploads patient information to a cloud server and groups this information according to the patient's identifiable characteristics. Within the same group, it compares the information of different patients, such as blood glucose data records and insulin infusion data records. Given similar patient information, the insulin algorithm parameters of patients with better blood glucose control can be used to correct the insulin algorithm of patients with poorer blood glucose control. This closed-loop artificial pancreas system no longer needs to spend a long time learning the patient's physiological characteristics and lifestyle habits to quickly obtain a personalized insulin algorithm suitable for the patient, which is beneficial for the treatment of diabetes.

[0010] This invention discloses a closed-loop artificial pancreas system, comprising a detection module for acquiring the patient's actual blood glucose data; an insulin algorithm for determining the patient's insulin infusion volume based at least on the actual blood glucose data; an infusion module for completing insulin infusion based on the insulin infusion volume; and a cloud server for receiving patient-related information and classifying the information into groups according to identifiable features; wherein the cloud server is also used to analyze the patient's relevant information, determine correction parameters for the insulin algorithm based on the analysis results, and correct the insulin algorithm based on the correction parameters.

[0011] According to one aspect of the present invention, the patient's relevant information includes at least blood glucose data records, insulin infusion records, meal records, exercise records, and sleep records.

[0012] According to one aspect of the invention, identifiable features are associated with a patient’s physiological state or lifestyle.

[0013] According to one aspect of the invention, the groups are independent of each other.

[0014] According to one aspect of the invention, groups having at least one identical identifiable feature are subsets of each other.

[0015] According to one aspect of the invention, the insulin algorithm parameters corresponding to different patients within the same group can be mutually modified.

[0016] According to one aspect of the invention, the correction of insulin algorithm parameters is performed actively or passively by a cloud server.

[0017] According to one aspect of the invention, patient confirmation is required before modifying insulin algorithm parameters.

[0018] According to one aspect of the invention, it further includes a program module for controlling the infusion module to complete insulin infusion.

[0019] According to one aspect of the invention, the insulin algorithm is located in a detection module, a program module, an infusion module, or a cloud server.

[0020] This invention also discloses a closed-loop insulin infusion method, comprising: a detection module, an infusion module, a cloud server, and an insulin algorithm, including the following steps:

[0021] I. The detection module acquires actual blood glucose data, and the cloud server receives relevant information uploaded by the patient;

[0022] II. The cloud server groups patient-related information according to the patient's identifiable characteristics;

[0023] III. Cloud server analyzes relevant information of patients in the same group;

[0024] IV. Based on the analysis results, determine the correction parameters for the insulin algorithm, and then correct the insulin algorithm based on the correction parameters;

[0025] V. Based on the corrected insulin algorithm and actual blood glucose data, the insulin infusion volume is calculated, and the infusion module completes the insulin infusion.

[0026] According to one aspect of the present invention, step I further includes uploading the patient's insulin algorithm parameters to a cloud server.

[0027] According to one aspect of the invention, in step III, the relevant patient information analyzed includes at least blood glucose data records.

[0028] According to one aspect of the present invention, in step IV, based on the analysis results of the recorded blood glucose data, the step further includes determining the insulin algorithm parameters of patients with better blood glucose control as correction parameters, and pushing the correction parameters to other patients.

[0029] According to one aspect of the invention, the method further includes the patient sending an instruction to a cloud server, and in response to the instruction, the cloud server pushing corrective parameters to the patient.

[0030] According to one aspect of the present invention, in step IV, based on the analysis results of the recorded blood glucose data, the cloud server further includes a step of calculating correction parameters based on the insulin algorithm parameters of patients with better blood glucose control, and pushing the correction parameters to other patients.

[0031] According to one aspect of the invention, step IV further includes the step of the patient downloading correction parameters to a cloud server.

[0032] According to one aspect of the invention, step IV further includes a step of patient confirmation of using modified parameters to modify the insulin algorithm.

[0033] Compared with the prior art, the technical solution of the present invention has the following advantages:

[0034] The personalized closed-loop infusion system and method disclosed in this invention uploads relevant patient information to a cloud server and groups the patient information according to the patient's identifiable characteristics. Within the same group, the relevant information of different patients, such as blood glucose data records and insulin infusion data records, is compared. Under the premise of the same or similar patient information, the insulin algorithm parameters of patients with better blood glucose control can be used to correct the insulin algorithm of patients with poor blood glucose control. The closed-loop artificial pancreas system no longer needs to learn the patient's physiological characteristics and lifestyle habits for a long time, and can quickly obtain a personalized insulin algorithm suitable for the patient, which is beneficial to the patient's diabetes treatment.

[0035] Furthermore, the correction of insulin algorithm parameters can be done actively or passively by the cloud server, or by patients setting their preferences to meet their different needs.

[0036] Furthermore, the cloud server can obtain insulin algorithm correction parameters by calling the patient's insulin algorithm parameters or by calculating them based on the patient's insulin algorithm parameters. The methods for obtaining correction parameters are diverse, meeting the different needs of patients. Attached Figure Description

[0037] Figure 1 is a schematic diagram of the module relationships of a general closed-loop artificial pancreas insulin infusion control system;

[0038] Figure 2 is a schematic diagram of the integrated CGM according to an embodiment of the present invention;

[0039] Figure 3 is a schematic diagram of the structure of a split CGM according to an embodiment of the present invention;

[0040] Figure 4a is a schematic diagram of the structure of an integrated insulin pump according to an embodiment of the present invention;

[0041] Figure 4b is a schematic diagram of a split-type insulin pump according to an embodiment of the present invention;

[0042] Figure 5a is a schematic diagram of the main interface of the control system in the first working mode according to an embodiment of the present invention;

[0043] Figure 5b is a schematic diagram of the main interface of the control system in the second working mode according to an embodiment of the present invention;

[0044] Figures 6a-6c are schematic diagrams illustrating different operations of the control system for activating the insulin pump function according to an embodiment of the present invention.

[0045] Figures 7a-7c are schematic diagrams illustrating different operations of the automatic mode function of the control system according to an embodiment of the present invention.

[0046] Figures 8a-8b are schematic diagrams of the APP interface before and after the automatic mode function of the control system is turned on according to an embodiment of the present invention;

[0047] Figure 9a is a schematic diagram of the interface when the system activates the "Grand Meal" mode according to an embodiment of the present invention;

[0048] Figures 9b and 9c are schematic diagrams of different interfaces for selecting "regular" and "grand meal" in the system of the embodiment of the present invention.

[0049] Figure 9d is a schematic diagram of the interface when the system starts the regular meal mode according to an embodiment of the present invention;

[0050] Figure 10 is a schematic diagram of the infusion strategy for pre-infusion and supplementary infusion according to an embodiment of the present invention;

[0051] Figure 11a is a schematic diagram of a closed-loop infusion system based on food image recognition according to an embodiment of the present invention;

[0052] Figure 11b is a schematic diagram of a closed-loop infusion method based on food image recognition according to an embodiment of the present invention;

[0053] Figure 11c is a schematic diagram of a closed-loop infusion method based on food image recognition according to another embodiment of the present invention;

[0054] Figure 12 is a schematic diagram of another closed-loop infusion method based on food image recognition according to an embodiment of the present invention;

[0055] Figure 13a is a schematic diagram of a personalized closed-loop infusion system based on a cloud server according to an embodiment of the present invention;

[0056] Figure 13b is a schematic diagram of a personalized closed-loop infusion method according to an embodiment of the present invention. Detailed Implementation

[0057] As mentioned earlier, existing closed-loop artificial pancreas systems have limitations in their correction based on individual data. These systems require a long period of learning about the patient's physiological characteristics and lifestyle habits to obtain insulin algorithm parameters suitable for the patient. This may lead to delayed or inaccurate correction of the insulin algorithm parameters, affecting the patient's treatment.

[0058] To address this issue, this invention provides a personalized closed-loop infusion system and method. The system uploads patient information to a cloud server and groups this information according to the patient's identifiable characteristics. Within the same group, it compares the relevant information of different patients, such as blood glucose data records and insulin infusion data records. Given similar or identical patient information, the insulin algorithm parameters of patients with better blood glucose control can be used to correct the insulin algorithm of patients with poorer blood glucose control. This closed-loop artificial pancreas system no longer requires extensive learning of the patient's physiological characteristics and lifestyle habits, and can quickly obtain a personalized insulin algorithm suitable for the patient, thus benefiting the treatment of diabetes.

[0059] Various exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be understood that, unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps set forth in these embodiments should not be construed as limiting the scope of the invention.

[0060] Furthermore, it should be understood that, for ease of description, the dimensions of the various components shown in the accompanying drawings are not necessarily drawn to actual scale; for example, the thickness, width, length, or distance of some units may be enlarged relative to other structures.

[0061] The following description of exemplary embodiments is merely illustrative and is not intended to limit the invention or its application or use in any way. Techniques, methods, and apparatus known to those skilled in the art may not be discussed in detail herein, but where applicable, such techniques, methods, and apparatus should be considered part of this specification.

[0062] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined or described in a figure, it will not need to be discussed further in the subsequent description of the figures.

[0063] Figure 1 is a schematic diagram of the module relationships of a general closed-loop artificial pancreas insulin infusion control system.

[0064] The closed-loop artificial pancreas insulin infusion control system disclosed in this embodiment of the invention mainly includes a detection module 100, a program module 101, and an infusion module 102.

[0065] The detection module 100 is used to continuously detect the patient's current blood glucose level. Generally, the detection module 100 is a continuous glucose monitoring (CGM) device, which can detect the patient's current blood glucose level in real time, monitor blood glucose changes, and send the current blood glucose information to the program module 101. The CGM includes an implantable sensor connected to a transmitter. The transmitter also includes a memory, processor, communication interface, etc. The transmitter is used to transmit at least the blood glucose data monitored by the CGM, as well as the CGM's identifier information.

[0066] The infusion module 102 includes the necessary mechanical structures and electronic control units for insulin infusion, such as a reservoir, drive structure, infusion tubing and needle, power supply, and circuit board. It is controlled by the program module 101. Generally, the infusion module 102 is an insulin pump, and the electronic control unit includes a memory, processor, and communication interface. Based on the current insulin infusion volume data issued by the program module 101, the infusion module 102 infuses the required insulin into the patient's body. Simultaneously, the infusion status of the infusion module 102 can be fed back to the program module 101 in real time.

[0067] The program module 101 controls the operation of the detection module 100 and the infusion module 102. Based on the blood glucose value detected by the detection module 100, the program module 101 generates an insulin infusion command and controls the infusion module 102 to perform the infusion. It includes a memory, a processor, a communication interface, a display, and a patient interface. The memory stores programming instructions, which the processor can execute. The program module 101 is connected to both the detection module 100 and the infusion module 102. This connection can be a conventional electrical connection or a wireless connection.

[0068] The embodiments of the present invention do not limit the specific locations and connection relationships of the detection module 100, the program module 101 and the infusion module 102, as long as the aforementioned functional conditions are met.

[0069] In one embodiment of the invention, the three components are electrically connected to form a single integrated structure. Therefore, all three are adhered to the same location on the patient's skin. By connecting the three modules into a single unit and adhering them to the same location, the number of devices that need to be applied to the patient's skin is reduced, thereby lessening the interference with the patient's activities caused by having too many devices. Simultaneously, it effectively solves the problem of reliable wireless communication between separate devices, further enhancing the patient experience.

[0070] In another embodiment of the invention, program module 101 and infusion module 102 are interconnected to form an integral structure, while detection module 100 is separately disposed in another structure. In this case, detection module 100 and program module 101 transmit wireless signals to each other to achieve interconnection. Therefore, program module 101 and infusion module 102 are attached to one location on the patient's skin, while detection module 100 is attached to other locations on the patient's skin.

[0071] In another embodiment of the invention, program module 101 and detection module 100 are interconnected to form a single device, while infusion module 102 is disposed separately in another structure. Infusion module 102 and program module 101 transmit wireless signals to each other to achieve interconnection. Therefore, program module 101 and detection module 100 can be attached to one location on the patient's skin, while infusion module 102 can be attached to other locations on the patient's skin.

[0072] In another embodiment of the invention, the three components are disposed in different structures. Therefore, they are respectively attached to different locations on the patient's skin. At this time, the program module 101 transmits wireless signals to the detection module 100 and the infusion module 102 to achieve interconnection.

[0073] In another embodiment of the invention, the three modules are disposed in different structures. Therefore, the detection module 100 and the infusion module 102 are respectively attached to different locations on the patient's skin, while the program module 101 does not need to be attached to the skin. Control of the detection module 100 and the infusion module 102 is achieved through a handheld or portable device, such as a PDM or a smartphone. In this case, the program module 101 transmits wireless signals to both the detection module 100 and the infusion module 102 to achieve mutual connection.

[0074] The wireless communication described in the foregoing embodiments can be, for example, but not limited to, radio frequency (RF) communication (e.g., RFID, Zigbee communication protocol, WiFi, infrared, wireless universal serial bus (USB), ultra-wideband (UWB)). Communication protocols and cellular communications, such as Code Division Multiple Access (CDMA) or Global System for Mobile Communications (GSM).

[0075] Figure 2 is a schematic diagram of an integrated CGM according to an embodiment of the present invention. Figure 3 is a schematic diagram of a split CGM according to an embodiment of the present invention.

[0076] A CGM consists of a sensor and a transmitter. It is implanted under the skin of the patient using an auxiliary installer. The sensor collects and transmits the blood glucose level data. The transmitter connects to the sensor and receives the blood glucose data from the implanted sensor, converting it into a wireless signal for output. Each CGM has a unique identifier, such as a device identifier, hardware identifier, universally unique identifier, serial number, protocol-based identifier (e.g., BLE ID), or manufacturer identifier. This identifier consists of a combination of random numbers and letters and can be placed on the CGM's casing or packaging, and may vary depending on the type of CGM.

[0077] Figure 2 is a schematic diagram of an integrated CGM. The sensor and transmitter of the CGM are integrated before use, and it is a single-use product that is discarded after use. As shown in Figure 2, the integrated CGM includes a sensor 201, a housing 202, and a transmitter (not shown) disposed within the housing 202. The sensor 301 monitors the patient's blood glucose data, transmitting this data to the transmitter via internal circuitry, and then the transmitter sends it to the receiver. Identifiers can be placed on the outer casing or packaging of the CGM, or inside the CGM itself.

[0078] Figure 3 is a schematic diagram of the split-type CGM. Before use, the sensor and transmitter of the CGM are two separate components, packaged separately, and integrated together only during use. The split-type CGM includes a base shell 301 and a transmitter 302. The sensor 3011 is mounted on the base shell, and the transmitter 302 has a separate housing. The base shell 301 and the transmitter 302 housings are respectively equipped with snap-fit ​​structures 3012 and 3022. During use, the base shell 301 and the transmitter 302 are snapped together into a single unit using these snap-fit ​​structures. The sensor 3011 is electrically connected to the transmitter 302 via an electrical connector 3013. The sensor 301 monitors the patient's blood glucose data and transmits this data to the transmitter 302 via the electrical connector 3013, which then sends it to the receiver.

[0079] In one embodiment of the invention, both the sensor and transmitter of the split-type CGM are single-use products, discarded after use. Therefore, the identifier can be placed on the housing or outer packaging of the sensor or transmitter. In another embodiment of the invention, only the sensor of the split-type CGM is a single-use product, while the transmitter is a reusable product. Therefore, preferably, in this embodiment, the identifier is placed on the housing or outer packaging of the transmitter, which can reduce the frequency of binding patient information and identifiers and improve the patient experience. This will be described in detail below.

[0080] When the identifier is set on the housing or outer packaging of the CGM or transmitter, it can be set in the form of, but is not limited to, a QR code, barcode, or NFC tag.

[0081] Figure 4a is a schematic diagram of an integrated insulin pump according to an embodiment of the present invention; Figure 4b is a schematic diagram of a split insulin pump according to an embodiment of the present invention.

[0082] In this embodiment of the invention, the insulin pump is a patch-type insulin pump, that is, an insulin pump that does not include a long catheter. It includes an infusion structure and a control structure, and is entirely attached to the patient's skin surface by the same adhesive patch. The drug is directly infused from the reservoir along the infusion needle to the subcutaneous tissue.

[0083] Each insulin pump has a unique identifier, such as a device identifier, hardware identifier, universally unique identifier, serial number, communication protocol-based identifier, manufacturer identifier, etc. The identifier is formed by a combination of multiple random numbers and letters, which can be set on the insulin pump housing or packaging, and can also have different settings depending on the type of insulin pump.

[0084] Figure 4a is a schematic diagram of an integrated insulin pump, in which the infusion structure 410 and the control structure 400 of the insulin pump are located inside the same housing 10. The two are connected by wires and are attached to a certain position on the patient's skin by an adhesive patch 420. The whole unit is discarded after single use. The identifier can be set on the outer housing of the insulin pump, the outer packaging, or inside the insulin pump.

[0085] Figure 4b is a schematic diagram of a split-type insulin pump, in which the infusion structure 410 and the control structure 400 are respectively housed in two different housings, which are connected by a waterproof plug or directly snapped together and electrically connected to form a whole. Identifiers can be placed on the outer shell or packaging of the infusion structure and / or control structure, or inside the insulin pump.

[0086] In one embodiment of the invention, both the infusion structure and control structure of the split insulin pump are single-use products, discarded after use. Therefore, the identifier can be placed on the housing or outer packaging of the infusion structure and / or control structure. In another embodiment of the invention, only the infusion structure of the split insulin pump is a single-use product, while the control structure is a reusable product. Therefore, preferably, in this embodiment, the identifier is placed on the housing or outer packaging of the control structure, which can reduce the frequency of binding patient information and identifiers and improve the patient experience. This will be described in detail below.

[0087] When the identifier is set on the casing or outer packaging of the insulin pump or control structure, it can be set in the form of, but not limited to, a QR code, barcode, or NFC tag.

[0088] Depending on the severity of the patient's condition and their individual health status, some patients may only require continuous glucose monitoring (CGM), while others require both CGM and an insulin pump for medication infusion. When a doctor determines that a patient only needs CGM for continuous glucose monitoring, since CGM only involves monitoring the patient's blood glucose, self-use of CGM will not pose a risk to the patient's life. Therefore, the patient can purchase CGM independently. Before the CGM is installed on the patient's skin, the patient can search for and download a dedicated app for controlling the CGM from their smartphone's app store, create a new account on the app, and pair their personal information with the CGM information to achieve pairing and control between the smartphone and the CGM. In this embodiment of the invention, the CGM and insulin pump are developed and manufactured by the same manufacturer, and therefore can be controlled by the same dedicated app on the smartphone. Since not all patients need to use an insulin pump, the default home screen of the dedicated app only includes CGM-related content, as shown in Figure 5a. This simplifies the app interface, improves the patient's visual experience, and prevents accidental operation of the insulin pump function from affecting the normal use of the CGM.

[0089] In one embodiment of the present invention, when a doctor determines that a patient needs to use an insulin pump for drug infusion, as shown in Figure 6a, the doctor sends a request to the backend administrator to add the patient's account to the whitelist, allowing the patient to use the insulin pump function. The backend administrator receives the doctor's request to add the patient's account to the whitelist and simultaneously sends feedback to the doctor confirming the whitelist addition is complete. Furthermore, the backend administrator directly opens the insulin pump function in the patient's app interface. At this time, the app interface changes from Figure 5a to Figure 5b, with Figure 5b adding two insulin infusion-related function keys, "Insulin Delivery" and "Easyloop," compared to Figure 5a. In another embodiment of the present invention, the backend administrator does not directly open the insulin pump function in the patient's app interface, but instead sends a security code to the patient's account. The patient can then use the security code to open the insulin pump function in the app interface when needed or convenient.

[0090] In another embodiment of the invention, when a doctor determines that a patient needs to use an insulin pump for drug infusion, as shown in Figure 6b, the patient can directly send an application to the backend administrator to enable the insulin pump function. Upon receiving the application, the backend administrator verifies whether the patient's account is on the whitelist. If the patient's account is on the whitelist, the backend administrator opens the insulin pump function on the patient's app interface, and the app interface changes from Figure 5a to Figure 5b. If the patient's account is not on the whitelist, a feedback message is sent to the patient's account, reminding the patient to ask the doctor to send an application to the backend administrator to add the patient to the whitelist. After the doctor sends the application to the backend administrator, the backend administrator can directly open the insulin pump function on the patient's app interface. If no message is received from the backend administrator within a certain time, such as 1 minute, 2 minutes, or 5 minutes, the patient can send another application to the backend administrator to enable the insulin pump function, or ask the doctor to send an application to add the patient to the whitelist. In another embodiment of the invention, the backend administrator does not directly open the insulin pump function on the patient's app interface, but instead sends a security code to the patient's account. The patient can then use the security code to open the insulin pump function on the app interface when needed or convenient.

[0091] In another embodiment of the present invention, when a doctor determines that a patient needs to use an insulin pump for drug infusion, as shown in Figure 6c, the doctor sends an application to the backend administrator requesting that the patient's account be added to the whitelist, allowing the patient to use the insulin pump function. The backend administrator receives the doctor's whitelist application, adds the patient's account to the whitelist, and simultaneously sends feedback to the doctor confirming the whitelist addition is complete. Furthermore, the doctor notifies the patient that they can apply to use the insulin pump function. After receiving the doctor's notification, the patient sends an application to the backend administrator to enable the insulin pump function. Upon receiving the application, the backend administrator directly opens the insulin pump function in the patient's app interface. In another embodiment of the present invention, after receiving the application, the backend administrator can first verify whether the patient's account is in the whitelist. If the patient's account is confirmed to be in the whitelist, the backend administrator opens the insulin pump function in the patient's app interface. In yet another embodiment, the backend administrator does not directly open the insulin pump function in the patient's app interface, but instead sends a security code to the patient's account. The patient can then use the security code to open the insulin pump function in the app interface when needed or convenient.

[0092] When a doctor determines that a patient no longer needs the insulin pump function, the patient can disable it themselves via the app. The app automatically sends a message to the backend administrator, who then removes the patient's account from the whitelist. Alternatively, the doctor and / or patient can submit a request to the backend administrator to disable the insulin pump function. The administrator will then disable the insulin pump function on the patient's app and remove the patient's account from the whitelist. When the patient needs to re-enable the insulin pump function, the patient's account needs to be re-added to the whitelist using one of the methods shown in Figures 6a-6c.

[0093] It's important to note that when activating the insulin pump function in the app, the insulin pump must be correctly installed on the skin, and the patient's personal information must be paired with the insulin pump's information to achieve pairing and control between the smartphone and the insulin pump. Patient personal information includes name, age, gender, and phone number. Information about the CGM and / or insulin pump includes its identifier. Simultaneously, the smartphone uploads the patient's personal information and the CGM and / or insulin pump identifier to a remote server. The remote server stores the uploaded information and verifies the validity of the CGM and / or insulin pump identifier. If an identifier already exists on the remote server, it sends a notification to the smartphone, reminding the patient that the CGM or insulin pump has been used and needs to be replaced. Once the CGM and / or insulin pump is installed on the patient's skin and successfully activated, it begins operation. The CGM's transmitter sends monitored blood glucose information to the smartphone and further uploads it to the remote server. The insulin pump's control mechanism receives insulin infusion information and controls the infusion mechanism to deliver insulin, while simultaneously sending the infusion status to the smartphone and uploading it to the remote server.

[0094] It should be noted that the CGM and insulin pump in the embodiments of the present invention are developed and manufactured by the same manufacturer, and therefore can be controlled by the same dedicated APP in a smartphone. When a patient needs both CGM and insulin pump at the same time, even if CGM or insulin pumps are manufactured by other manufacturers, they can still be directly controlled by the dedicated APP. This avoids the inconvenience caused to patients by using different APPs to control CGM and insulin pumps separately, and improves the patient experience.

[0095] When a patient's CGM and / or insulin pump needs to be replaced due to reaching its usage period or failure, the unique identifier information of the new CGM and / or insulin pump also needs to be paired and updated with the patient's personal information via a smartphone and then uploaded to a remote server. The patient's personal information can be entered manually, and the identifier information of the CGM and / or insulin pump can also be entered manually or by scanning the QR code, barcode, or NFC tag on the housing or outer packaging of the CGM and / or insulin pump.

[0096] When the CGM has a split structure and the transmitter is reusable, the CGM identifier is set on the transmitter's shell or packaging. When the patient replaces the CGM, only the sensor needs to be replaced, without replacing the transmitter. The CGM identifier also remains unchanged. Therefore, there is no need to update the pairing of the CGM identifier with the patient's personal information via smartphone, nor is it necessary to upload it to a remote server. This reduces the number of steps and improves the patient experience.

[0097] When the insulin pump has a separate structure and the control structure is reusable, the insulin pump identifier is set on the outer shell or packaging of the control structure. When the patient replaces the insulin pump, only the infusion structure needs to be replaced, without replacing the control structure. The insulin pump identifier also remains unchanged. Therefore, there is no need to update the pairing of the insulin pump identifier with the patient's personal information via smartphone, nor is it necessary to upload it to a remote server. This reduces the number of operation steps and improves the patient experience.

[0098] The smartphone and CGM and / or insulin pump, as well as the remote server, communicate wirelessly. This wireless communication can be achieved through, but is not limited to, radio frequency (RF) communication (e.g., RFID, Zigbee, WiFi, infrared, USB, and UWB). Communication protocols and cellular communication, such as Code Division Multiple Access (CDMA) or Global System for Mobile Communications (GSM). Preferably, communication between the smartphone and the remote server is via WiFi and / or cellular communication, and communication between the smartphone and the CGM and / or insulin pump is via... Communication protocol communication.

[0099] When a doctor determines that a patient can activate the automatic mode, the app reads the current blood glucose level monitored by the CGM and the insulin infusion information from the insulin pump. It can then calculate future blood glucose trends and control the insulin pump infusion based on these trends, including increasing, decreasing, or stopping the infusion to influence blood glucose levels, forming an automated closed-loop control. As shown in Figure 7a, the doctor sends a request to the backend administrator to add the patient's account to the whitelist, allowing the patient to use the automatic mode function. The backend administrator receives the whitelist request, adds the patient's account to the whitelist, and sends feedback to the doctor confirming the whitelist addition. Furthermore, the automatic mode function is directly activated in the app interface used by the patient. At this point, the app interface changes from Figure 8a to Figure 8b. The interface in Figure 8b adds an "Auto Mode" button, a function key related to automatic mode, compared to the interface in Figure 8a. In another embodiment of the invention, the backend administrator does not directly activate the automatic mode function in the app interface used by the patient. Instead, a security code is sent to the patient's account, allowing the patient to activate the automatic mode function in the app interface when needed or convenient.

[0100] In another embodiment of the invention, when the doctor determines that the patient can enable the automatic mode, as shown in Figure 7b, the patient can directly send an application to the backend administrator to enable the automatic mode function. After receiving the application, the backend administrator will verify whether the patient's account is in the whitelist. If the patient's account is in the whitelist, the backend administrator will enable the automatic mode function of the APP interface used by the patient, and the APP interface will change from Figure 8a to Figure 8b. If the patient's account is not in the whitelist, a feedback message will be sent to the patient's account, reminding the patient to ask the doctor to send an application to the backend administrator to add the patient to the whitelist. After the doctor sends the application to the backend administrator, the backend administrator can directly enable the automatic mode function of the APP interface used by the patient. If no message is received from the backend administrator within a certain time, such as 1 minute, 2 minutes, or 5 minutes, the doctor can send another application to enable the automatic mode function, or ask the doctor to send an application to add the patient to the whitelist. In another embodiment of the invention, the backend administrator does not directly enable the automatic mode function of the APP interface used by the patient, but instead sends a security code to the patient's account. The patient can enable the automatic mode function of the APP interface using the security code when needed or convenient.

[0101] In another embodiment of the present invention, when a doctor determines that a patient needs to use the automatic mode, as shown in Figure 7c, the doctor sends a request to the backend administrator to add the patient's account to the whitelist, allowing the patient to use the automatic mode function. The backend administrator receives the doctor's request to add the patient's account to the whitelist and simultaneously sends feedback to the doctor confirming the whitelist addition is complete. Furthermore, the doctor notifies the patient that they can apply to use the automatic mode function. After receiving the doctor's notification, the patient sends a request to the backend administrator to enable the automatic mode function. Upon receiving the request, the backend administrator directly opens the automatic mode function in the patient's app interface. In another embodiment of the present invention, after receiving the patient's request to enable the automatic mode function, the backend administrator can first verify whether the patient's account exists in the whitelist. If the patient's account is confirmed to be in the whitelist, the backend administrator opens the automatic mode function in the patient's app interface. In yet another embodiment of the present invention, the backend administrator does not directly open the automatic mode function in the patient's app interface, but instead sends a security code to the patient's account. The patient can then use the security code to open the automatic mode function in the app interface when needed or convenient.

[0102] When a doctor determines that a patient no longer needs automatic mode, the patient can manually disable it on the app. The app will automatically send a message to the backend administrator, who will then remove the patient's account from the whitelist. Alternatively, the doctor and / or patient can submit a request to the backend administrator to disable automatic mode. The administrator will then disable automatic mode on the patient's app and remove the patient's account from the whitelist. When the patient needs to re-enable automatic mode, the patient's account needs to be re-added to the whitelist using one of the methods shown in Figures 7a-7c.

[0103] In this embodiment of the invention, the security code sent by the backend when requesting the insulin pump function and automatic mode function can be any number or combination of numeric characters, alphanumeric characters, and other symbols. It can also be a series of taps, a series of inputs, complex or simple gestures (e.g., swiping or other movements on a touchscreen, drawing images), etc. In some cases, the security code may also include a quiz or a set of questions. The security code sent by the backend each time is random.

[0104] Because dietary habits vary significantly across different regions and age groups, a uniform control plan may result in poor glycemic control for some individuals. Therefore, when the automatic mode function is activated, patients are required to enter a pass code to access different automatic mode interfaces. For individuals with high carbohydrate consumption, the automatic mode interface, as shown in Figure 9a, displays a "large meal" option after entering the pass code. Patients can choose whether to enable the large meal module. After enabling the large meal mode, the infusion page displays two options: "Regular" and "Large Meal," as shown in Figures 9b and 9c. When the patient selects "Regular," the insulin infusion volume corresponding to the regular carbohydrate intake will be administered; when selecting "Large Meal," the insulin infusion volume corresponding to a larger carbohydrate intake will be administered. For individuals with low carbohydrate consumption, the automatic mode interface, as shown in Figure 9d, does not have a large meal option and defaults to "Regular" mode, administering the insulin infusion volume corresponding to the regular carbohydrate intake.

[0105] In some embodiments of the invention, the access code can be a set of small questions, such as "Are you a carbohydrate enthusiast?", "Is your age within the range of A and B?", "What is your gender?", "Where do you live?", "What are your fitness hobbies?", "Do you have any special medical conditions?", "Have you used non-automatic mode before activating automatic mode?", etc. Based on the patient's answers, the system automatically determines whether the patient is a high carbohydrate consumer. In other embodiments of the invention, the access code is provided to the patient in advance by the doctor after diagnosis, or the doctor sends the patient's information regarding high carbohydrate consumption when sending an automatic mode whitelist application to the backend administrator. The backend administrator automatically assigns a corresponding access code to the patient. The access code can be any number or combination of numeric characters, alphanumeric characters, and other symbols, or it can be a series of taps, a series of inputs, complex or simple gestures (e.g., swiping or other movements on a touchscreen, drawing images), etc. The backend administrator can send the access code to the patient when activating automatic mode, or send the access code to the patient at the same time as sending the security code, or send the access code to the patient after confirming that the patient has activated the automatic mode function through the security code. In this embodiment of the invention, both the security code and the access code are generated randomly, and their generation rules can be the same or different. Preferably, the generation rules for the security code and the access code are different to avoid confusion for patients and thus avoid causing them distress.

[0106] In this embodiment of the invention, the system or doctor may make a comprehensive judgment based on multiple factors, including the patient's age, dietary habits, exercise habits, health status, and the results of using the non-automatic mode. When the patient has a record of using the non-automatic mode, the record of using the non-automatic mode shall be used as the main basis for judgment.

[0107] After the patient selects a meal, including "Regular" and "Grand Meal," the automatic mode employs a pre-infusion and supplemental infusion strategy, as shown in Figure 10. During pre-infusion and supplemental infusion, the pre-infusion and supplemental infusion volumes are categorized into different levels for both regular and grand meals, as shown in Table 1 below:

[0108] The insulin infusion volume during pre-infusion is at least related to the actual blood glucose level or rate of change of blood glucose at the time of pre-infusion, and the estimated meal intake. In other embodiments of the invention, the pre-infusion volume may also be related to the intracellular insulin content (IOB). Similarly, the insulin infusion volume during supplemental infusion is at least related to the actual blood glucose level or rate of change of blood glucose at the time of supplemental infusion, and the estimated supplemental meal intake. In other embodiments of the invention, the supplemental infusion volume may also be related to the intracellular IOB. In these embodiments, meal intake refers to the carbohydrate content of the meal.

[0109] The size of the meal intake during pre-infusion and the size of the meal intake during supplementary infusion are independent parameters. The meal intake during pre-infusion of the same level is greater than the meal intake during supplementary infusion, but there is no fixed correspondence between the two. The system can be set according to actual needs. For large meals, the minimum meal intake during pre-infusion is not less than the maximum meal intake for regular meals, and the minimum meal intake during supplementary infusion is not less than the maximum meal intake for regular meals.

[0110] Step 1001: At time T0, the patient selects the meal type and the default pre-infusion insulin volume. The default pre-infusion insulin volume can be any pre-infusion insulin volume corresponding to small, medium, or large meal sizes. Preferably, the default pre-infusion insulin volume is the insulin volume corresponding to small meal sizes. Selecting a small supplemental insulin volume can prevent excessive insulin infusion and reduce the risk of hypoglycemia.

[0111] Step 1002: At time T1, compare the current blood glucose level monitored by CGM with a preset blood glucose threshold, such as 140, 150, 160, 170, 180, 190, 200 mg / mL, etc. If the current blood glucose level is greater than the preset blood glucose threshold, then administer the default supplemental insulin infusion volume; otherwise, do not administer a supplemental infusion volume. The default supplemental infusion volume can be any supplemental insulin infusion volume corresponding to a small, medium, or large meal size. Preferably, the default supplemental insulin infusion volume is the insulin infusion volume corresponding to a small meal size. Selecting a small initial supplemental insulin infusion volume can prevent excessive insulin infusion and reduce the risk of hypoglycemia. Time T1 may be 1 hour, 1.5 hours, 2 hours, 2.5 hours, 3 hours, etc., after time T0. In other embodiments of the present invention, the blood glucose change rate at time T1 can also be used to determine whether to administer a supplemental infusion.

[0112] Step 1003: Within the ΔT0 time interval after time T0, such as 3h, 4h, 5h, if the patient experiences hyperglycemia, the next pre-infusion is upgraded, i.e., a larger insulin infusion volume corresponding to the estimated meal intake is administered, while considering the actual blood glucose level or rate of change of blood glucose, or the internal rate of block (IOB) at the next pre-infusion time; if the patient experiences hypoglycemia, the next pre-infusion is downgraded, i.e., a smaller insulin infusion volume corresponding to the estimated meal intake is administered, while considering the actual blood glucose level or rate of change of blood glucose, or the internal rate of block (IOB) at the next pre-infusion time; if the patient experiences neither hyperglycemia nor hypoglycemia, the next pre-infusion level remains unchanged, i.e., an insulin infusion volume corresponding to the same meal intake is administered, while considering the actual blood glucose level or rate of change of blood glucose, or the internal rate of block (IOB) at the next pre-infusion time.

[0113] It should be noted that, in the embodiments of the present invention, the change of infusion level means the change of the level of estimated meal intake during infusion. At the same time, the actual blood glucose value or blood glucose change rate at the time of infusion, or the in vivo IOB, are also taken into account. Therefore, in the embodiments of the present invention, the change of the estimated meal intake level also means the change of infusion level. That is, the infusion level and the estimated meal intake level can be understood as being consistent.

[0114] Step 1004: Within the ΔT1 time interval after time T1, such as 3h, 4h, 5h, etc., if the patient experiences hyperglycemia, the next supplemental infusion will be upgraded, i.e., a supplemental insulin infusion corresponding to a larger estimated meal intake will be administered, while considering the actual blood glucose level or rate of change of blood glucose, or the internal rate of deficit (IOB) at the time of the next supplemental infusion. If the patient experiences hypoglycemia, the next supplemental infusion will be downgraded, i.e., a supplemental insulin infusion corresponding to a smaller estimated meal intake will be administered, while considering the actual blood glucose level or rate of change of blood glucose, or the internal rate of deficit (IOB) at the time of the next supplemental infusion. If the patient experiences neither hyperglycemia nor hypoglycemia, the next supplemental infusion level will remain unchanged, i.e., a supplemental insulin infusion corresponding to the same estimated meal intake will be administered, while considering the actual blood glucose level or rate of change of blood glucose, or the internal rate of deficit (IOB) at the time of the next supplemental infusion.

[0115] Step 1005: At time T2, the patient selects the type of meal and pre-infusion is performed based on the results of step 1003. This means pre-infusing a larger, smaller, or unchanged amount of insulin corresponding to the estimated meal intake, while also considering the actual blood glucose level or rate of change of blood glucose at time T2, or the intraocular insulin deficit (IOB).

[0116] Step 1006: At time T3, compare the current blood glucose level monitored by CGM with a preset blood glucose threshold, such as 140, 150, 160, 170, 180, 190, 200 mg / mL, etc. If the current blood glucose level is greater than the preset blood glucose threshold, supplementary infusion is performed according to the result of step 1004. That is, a supplementary infusion of insulin corresponding to a larger, smaller, or unchanged supplementary meal intake is performed, while considering the actual blood glucose level or blood glucose change rate at time T3, or the in vivo IOB. If the current blood glucose level is not greater than the preset blood glucose threshold, no supplementary infusion is performed. In other embodiments of the present invention, the blood glucose change rate at time T3 can also be used to determine whether to perform supplementary infusion.

[0117] Step 1007: Within the ΔT0 time interval after time T2, such as 3h, 4h, 5h, if the patient experiences hyperglycemia, the next pre-infusion is upgraded, i.e., a larger insulin infusion volume corresponding to the estimated meal intake is administered, while considering the actual blood glucose level or rate of change of blood glucose, or the internal rate of block (IOB) at the time of the next pre-infusion. If the patient experiences hypoglycemia, the next pre-infusion is downgraded, i.e., a smaller insulin infusion volume corresponding to the estimated meal intake is administered, while considering the actual blood glucose level or rate of change of blood glucose, or the internal rate of block (IOB) at the time of the next pre-infusion. If the patient experiences neither hyperglycemia nor hypoglycemia, the next pre-infusion level remains unchanged, i.e., an insulin infusion volume corresponding to the same meal intake is administered, while considering the actual blood glucose level or rate of change of blood glucose, or the internal rate of block (IOB) at the time of the next pre-infusion.

[0118] Step 1008: Within the ΔT1 timeframe after time T3, such as 3h, 4h, 5h, if the patient experiences hyperglycemia, the next supplemental infusion will be upgraded, i.e., a larger supplemental insulin infusion corresponding to the estimated meal intake will be administered, while considering the actual blood glucose level or rate of change of blood glucose, or the internal rate of blood glucose (IOB) at the time of the next supplemental infusion. If the patient experiences hypoglycemia, the next supplemental infusion will be downgraded, i.e., a smaller supplemental insulin infusion corresponding to the estimated meal intake will be administered, while considering the actual blood glucose level or rate of change of blood glucose, or the IOB at the time of the next supplemental infusion. If the patient experiences neither hyperglycemia nor hypoglycemia, the next supplemental infusion level will remain unchanged, i.e., an insulin infusion corresponding to the same meal intake will be administered, while considering the actual blood glucose level or rate of change of blood glucose, or the IOB at the time of the next supplemental infusion.

[0119] When pre-infusion and supplemental infusion are required at the next moment, repeat steps 1005-1008.

[0120] Generally, to maintain stable blood glucose levels, patients choose the same meal type at time T2 and time T0. That is, if a patient chooses a regular meal at time T0, they will also choose a regular meal at time T2; if they choose a large meal at time T0, they will also choose a large meal at time T2. Therefore, the choice of pre-infusion insulin and supplemental insulin infusion for the next time can depend on the results of the previous choice. However, if the patient's choice of meal for the next time is inconsistent with the previous choice, the patient will revert to the initial default pre-infusion and default supplemental infusion amounts at the next pre-infusion and supplemental infusion time to prevent inaccurate insulin infusion due to changes in meal patterns.

[0121] It's important to note here that if the current pre-infusion dose corresponds to the largest meal intake among the selected meal types, and hyperglycemia occurs within the ΔT0 time interval after time T0, the next pre-infusion dose will not be upgraded; the pre-infusion dose will still correspond to the larger estimated meal intake, while considering the actual blood glucose level or rate of change of blood glucose, or the internal rate of block (IOB) at the next pre-infusion time. Similarly, if the current pre-infusion dose corresponds to the smallest meal intake among the selected meal types, and hypoglycemia occurs within the ΔT0 time interval after time T0, the next pre-infusion dose will not be downgraded; the pre-infusion dose will still correspond to the smaller estimated meal intake, while considering the actual blood glucose level or rate of change of blood glucose, or the internal rate of block (IOB) at the next pre-infusion time.

[0122] Similarly, if the current supplemental infusion corresponds to the largest meal intake among the selected meal types, and hyperglycemia occurs within the ΔT0 time interval after time T0, the next supplemental infusion volume will not be upgraded; instead, the insulin infusion volume corresponding to the estimated large supplemental meal intake will still be administered, while considering the actual blood glucose level or rate of change of blood glucose, or the internal rate of block (IOB) at the time of the next supplemental infusion. Likewise, if the current supplemental infusion corresponds to the smallest meal intake among the selected meal types, and hypoglycemia occurs within the ΔT0 time interval after time T0, the next supplemental infusion volume will not be downgraded; instead, the insulin infusion volume corresponding to the estimated small supplemental meal intake will still be administered, while considering the actual blood glucose level or rate of change of blood glucose, or the internal rate of block (IOB) at the time of the next supplemental infusion.

[0123] In other embodiments of the present invention, the meal intake corresponding to pre-infusion and supplementary infusion is not necessarily graded. That is, the estimated meal intake during pre-infusion may only be the default amount, while the estimated meal intake during supplementary infusion may be of three different levels: large, medium, and small. Therefore, during the pre-infusion stage, each pre-infusion only infuses the insulin infusion amount corresponding to the default meal intake; or the estimated meal intake during pre-infusion may be of three different levels: large, medium, and small, while the estimated meal intake during supplementary infusion may only be the default amount. Therefore, during supplementary infusion, each supplementary infusion only infuses the insulin infusion amount corresponding to the default meal intake.

[0124] Similarly, the estimated food intake settings for pre-infusion and supplementary infusion are not necessarily the same for large meals and regular meals. That is, for each meal mode, the estimated food intake for pre-infusion and supplementary infusion can be selected as graded or not graded (default amount), which can be set according to the actual needs of the patient.

[0125] In this embodiment of the invention, the system is also equipped with a small meal mode, i.e. a snack mode. When the patient selects the snack mode, since the carbohydrate content in the snack is relatively small, the system infuses insulin based on the default estimated meal amount, while taking into account the patient's actual blood glucose level or blood glucose change rate when eating snacks, as well as the IOB in the body. Moreover, the default estimated meal amount in the snack mode is less than the lowest level of estimated meal amount in the regular mode.

[0126] Figure 11a is a schematic diagram of a closed-loop infusion system based on food image recognition according to an embodiment of the present invention. Figure 11b is a schematic diagram of a closed-loop infusion method based on food image recognition according to an embodiment of the present invention. Figure 11c is a schematic diagram of a closed-loop infusion method based on food image recognition according to another embodiment of the present invention.

[0127] In other embodiments of the present invention, quantitative analysis of food is performed to determine the content of nutrients such as carbohydrates, fats and proteins in the food, which can further optimize the insulin infusion amount required during meals and obtain a more accurate insulin infusion amount.

[0128] In this embodiment of the invention, an imaging module 103 is added to the system. The imaging module 103 can be used to acquire food images. These food images are input to a program module 101 or a cloud server 104 for quantitative analysis. This analysis determines the type and weight of the food, for example, 500g of potatoes, 340g of pasta, 220g of beef, and 450g of milk. The nutritional composition of the food is then determined based on its type, and the nutrient content is determined based on the weight of each food item. The cloud server 104 can be a large server located on a public network for storing and computing data.

[0129] Specifically, in some embodiments of the present invention, the imaging module 103 is an image acquisition device independent of the closed-loop artificial pancreas. With the development and popularization of closed-loop artificial pancreas technology, smart devices can be connected to the closed-loop artificial pancreas to assist in the patient's blood glucose detection and insulin infusion. Smart devices, such as mobile phones, tablets, and head-mounted augmented reality devices, generally have camera functions. Patients can use the camera function of the smart device to take pictures of food before meals to obtain food images. Since smart devices can be frequently iterated, their image acquisition software and hardware meet the needs of daily food image acquisition, have excellent performance, and the captured food images are clear, which helps in the quantitative analysis of food images.

[0130] In other embodiments of the present invention, the imaging module 103 may be a sub-module of the closed-loop artificial pancreas. For example, a camera function may be added to the PDM to obtain food images, thus eliminating the need for additional image acquisition equipment and saving patients usage costs.

[0131] Regardless of the device used to photograph the food, it is referred to as imaging module 103 in this embodiment of the invention.

[0132] In this embodiment of the invention, the imaging module 103 is connected to the closed-loop artificial pancreas system via wired or wireless means, and interacts with at least one of the detection module 100, program module 101, and infusion module 102 to transmit food image data to the closed-loop artificial pancreas. Alternatively, after acquiring a food image, the imaging module 103 directly analyzes the data to determine the type and weight of the food, then transmits the food type and weight data to the closed-loop artificial pancreas to determine the nutrient content. The system then combines this data with blood glucose data to determine the insulin infusion volume and administers the insulin. Alternatively, after acquiring a food image, the imaging module 103 analyzes the data to determine the type and weight of the food, then determines the nutrient content based on the food type and weight, and then transmits the nutrient data to the closed-loop artificial pancreas. The system then combines this data with blood glucose data to determine the insulin infusion volume and administer the insulin. Alternatively, after acquiring a food image, the imaging module 103 performs data analysis on the food image to determine the type and weight of the food, then determines the nutrient content based on the type and weight of the food, combines the blood glucose data to determine the insulin infusion amount, and then transmits the insulin infusion amount data to the closed-loop artificial pancreas, and the system infuses insulin according to the insulin infusion amount data.

[0133] In this embodiment of the invention, after acquiring a food image, the imaging module 103 can directly perform quantitative analysis on the food image in the imaging module 103, or it can transmit the food image to a closed-loop artificial pancreas for quantitative analysis by the system, or it can transmit the food image to a cloud server 104 for quantitative analysis by the cloud server 104.

[0134] In this embodiment of the invention, a deep learning-based food image recognition model can be used to quantitatively analyze food images. For example, it can identify features such as color (R, G, B channels), shape, layering, texture, and projection in the food image. Combined with the shooting distance features of the imaging module 103 relative to the food, the type and weight of the food can be determined. To more accurately identify food images, a food image library can be accessed. Each food item in the image can be extracted and compared with images in the food image library to determine the type of food. After determining the type and weight of the food, the data is transmitted to a food nutrition library to determine the nutritional components and their content.

[0135] In this embodiment of the invention, the food nutrient library can record only the nutrient content of various foods, as shown in Table 2.1. After identifying the food type and weight in the food image, the content of each nutrient is calculated by reading the food nutrient content from the food nutrient library. The food nutrient library can also record the nutrient content of a specified weight of food, as shown in Table 2.2. After identifying the food type and weight in the food image, the nutrient content of each food in the food image can be obtained directly by reading the food nutrient library.

[0136] Table 2.1 Food Nutrition Bank

[0137] In this embodiment of the invention, if 500g of potatoes are identified in a food image, and the food nutrient library records the carbohydrate content of potatoes as α, the fat content as β, and the protein content as γ, then the carbohydrates provided by the potatoes in the food image are 500g*α, the fat content as 500g*β, and the protein content as 500g*γ. If 250g of beef is also identified in the same food image, and the food nutrient library records the carbohydrate content of beef as α', the fat content as β', and the protein content as γ', then the carbohydrates provided by the beef in the food image are 250g*α', the fat content as 250g*β', and the protein content as 250g*γ'. Therefore, the total carbohydrates identified in the food image are 500g*α + 250g*α', the total fat content is 500g*β + 250g*β', and the total protein content is 500g*γ + 250g*γ'. Based on these identified total nutrient contents, a simulated blood glucose curve for the patient is predicted, the postprandial insulin infusion volume is calculated, and the corresponding insulin infusion is performed. The above data is for illustrative purposes only. When calculating insulin infusion volume, algorithms such as PID, MPC, and neural networks can be used to predict the patient's simulated blood glucose curve.

[0138] Table 2.2 Food Nutrition Bank

[0139] In this embodiment of the invention, 400g of potatoes and 200g of beef are identified in the food image. The food nutrient database records the carbohydrate content of potatoes as a (mg), fat content as b (mg), and protein content as c (mg), and the carbohydrate content of beef as d (mg), fat content as e (mg), and protein content as f (mg). Therefore, the total carbohydrate content identified in the food image is a+d (mg), the fat content as b+e (mg), and the protein content as c+f (mg). Based on these identified total nutrient contents, a simulated blood glucose curve for the patient is predicted, and the postprandial insulin infusion volume is calculated and administered. The above data is for illustrative purposes only. When calculating the insulin infusion volume, insulin algorithms such as PID, MPC, and neural networks are used to predict the simulated blood glucose curve for the patient.

[0140] In this embodiment of the invention, the food nutrition database can be stored in a smart device, a closed-loop artificial pancreas, or a cloud server 104, without limitation. Due to the limited storage space of the readable storage device, the food nutrition database will not record the nutritional content of all specified weights and types of food. The system can display the closest food item and its corresponding nutritional content for patient confirmation. When the patient or their medical caregiver can provide more precise food nutritional content, it can be input into the system and stored through the smart device or the PDM's interactive interface to update the existing data recorded in the food nutrition database, or to add unrecorded original data to the food nutrition database. The unrecorded original data includes food type, weight, and nutrient content.

[0141] In this embodiment of the invention, the food nutrient bank can record not only the nutritional content of carbohydrates, fats and proteins, but also the content of nutrients such as cholesterol, vitamins, and trace elements, such as vitamin B1, vitamin C, vitamin D, calcium, magnesium, iron, zinc, sodium, and potassium.

[0142] In this embodiment of the invention, after identifying the nutrient content in the food image, simulated blood glucose curves corresponding to each nutrient are calculated based on total carbohydrates, total fats, and total proteins, i.e., carbohydrate blood glucose curve l. α Fat and blood glucose curve β and protein glucose curve l γ Carbohydrate blood glucose curve l α Fat and blood glucose curve β and protein glucose curve l γ This can separately reflect the effects of postprandial carbohydrates, fats, and proteins on a patient's blood glucose levels. When calculating the simulated blood glucose curve, historical reference data and IOB (Independent Blood Scale) can be used as a reference. After obtaining the carbohydrate blood glucose curve... α Fat and blood glucose curve β and protein glucose curve l γ Then, the three curves are fitted to obtain the simulated blood glucose curve l1. The simulated blood glucose curve l1 can be obtained from the carbohydrate blood glucose curve l. α Fat and blood glucose curve β and protein glucose curve l γ Obtained through linear fitting: l1=g*l α +h*l β +i*l γ (1)

[0143] Where g, h, and i are the fitting coefficients of the blood glucose curves for each nutrient.

[0144] The simulated blood glucose curve l1 reflects the theoretical blood glucose change after the nutrient content in the food image is identified by the food image recognition model, the insulin infusion volume is calculated, and the insulin is infused into the patient.

[0145] In this embodiment of the invention, the fitting of the simulated blood glucose curve l1 lasts for a period of time, for example, 2 to 5 hours. During this period, the detection module 100 acquires the patient's actual blood glucose curve l2 in real time. By comparing the simulated blood glucose curve l1 data with the actual blood glucose curve l2 data, the simulated blood glucose curve l1 can be optimized to improve the carbohydrate blood glucose curve l2. α Fat and blood glucose curve β and protein glucose curve l γ This improves the accuracy of food image recognition, allowing the food image recognition model to learn further and adjust the parameters of the food image recognition algorithm and the insulin algorithm, thereby improving the accuracy of food image recognition and insulin infusion calculation. Specifically, after the fitting duration of the simulated blood glucose curve l1 ends, the difference between the simulated blood glucose curve l1 and the actual blood glucose curve l2 during this period is compared to adjust and optimize the carbohydrate blood glucose curve l2. α Fat and blood glucose curve β and protein glucose curve l γ The parameters mean that the total carbohydrate, total fat, and total protein content identified by the food image recognition model before meals can be adjusted and optimized, thereby correcting the insulin algorithm parameters to improve their accuracy and applicability. Similarly, based on the aforementioned comparison differences, the food image recognition algorithm parameters in the food image recognition model can also be corrected, making the next food image recognition result more accurate. The correction procedure for the food image recognition model algorithm parameters and the insulin algorithm parameters is repeatable. This correction procedure can be repeated to continuously iterate the food image recognition model algorithm and the insulin algorithm. This continuous iterative process is the deep learning process of the food image recognition model and the insulin algorithm.

[0146] In this embodiment of the invention, the iterative correction of the food image recognition model algorithm and the insulin algorithm is based on the patient's actual blood glucose curve l2. When infusing insulin, since different patients have different physiological characteristics and lifestyles, even if the food image recognition model identifies the exact same nutrient content and completes the corresponding insulin infusion in the initial state, the actual blood glucose curve l2 detected by the detection module 100 will not be exactly the same. This will result in different correction results for the parameters of the food image recognition model and the parameters of the insulin algorithm. Moreover, as the correction program iterates repeatedly, the parameters of the food image recognition model will become more accurate, and the parameters of the insulin algorithm will gradually become unique to each patient and will no longer be applicable to other patients. After repeated parameter iterations, the insulin algorithm will also change from the generalized insulin algorithm at the factory to the narrow insulin algorithm of each patient.

[0147] In this embodiment of the invention, the food image recognition model and insulin algorithm, after repeated parameter iterations, can be uploaded to the cloud server 104 by the patient's choice. Alternatively, the patient may agree to the system automatically uploading the iterated food image recognition model and insulin algorithm to the cloud server 104. Or, the food image recognition model and insulin algorithm may run on the cloud server 104, and the patient may choose to agree to have the iterated food image recognition model and insulin algorithm made public to the server backend. After obtaining a sufficient number of narrow-sense insulin algorithms, they can be aggregated to form a narrow-sense insulin algorithm cluster. Within the cluster, patients can be categorized into different groups based on identifiable characteristics such as physiological features and lifestyle habits from which the model originated. These groups share common identifiable features. For example, patients can be grouped by age or age range, such as 0-10 years, 10-15 years, 15-18 years, 18-20 years, etc. Alternatively, they can be divided into male and female groups, with female patients further subdivided based on pregnancy status. Groups can also be formed based on patients' habitual daily exercise duration, such as 0-10 minutes, 10-30 minutes, 30-60 minutes, etc. Furthermore, groups can be formed based on a combination of common identifiable features, such as "patients aged 10-15 years with daily exercise duration of 10-30 minutes," "male patients aged 15-18 years," "pregnant patients aged 20-22 years," etc. Identifiable features may also include individual insulin resistance, weight, family history of genetic diseases, medical history, nationality, and occupation. Since patients from each group have one or more common identifiable features, any algorithm within the group can be applied to other patients with the same identifiable features to a certain extent. This provides convenience for patients with these identifiable features. Before meals, these patients can choose whether to use the food image recognition model and narrow insulin algorithm of other patients with the same identifiable features. This has higher accuracy and applicability than directly using the generalized insulin algorithm, and does not require multiple corrections of insulin algorithm parameters and image recognition models. It has higher accuracy in recognizing food images and improves the therapeutic effect of the closed-loop artificial pancreas system.

[0148] In this embodiment of the invention, after using the food image recognition model and narrow-sense insulin algorithm of a certain group of patients, the food image recognition model and narrow-sense insulin algorithm are iterated after one or more meals. The parameters of the iterated food image recognition model and narrow-sense insulin algorithm can also be fed back to the patient's identifiable feature group. For example, a patient with the identifiable feature of "a 16-year-old male patient" uses the food image recognition model to identify food images before meals, and then uses the narrow-sense insulin algorithm to complete insulin infusion. Five hours after meals, the food image recognition model and narrow-sense insulin algorithm used by the patient have completed iterations. The patient chooses to disclose the parameters of the iterated food image recognition model and narrow-sense insulin algorithm to the backend. These parameters can then be fed back to the group of "16-year-old patients", the group of "male patients", and the group of "16-year-old male patients".

[0149] In this embodiment of the invention, identifiable feature groups exist in a parallel and independent manner. For example, "16-year-old male patient" and "16-year-old patient" are two independent groups. Both groups store applicable food image recognition models and narrow-sense insulin algorithm parameters. Within these independent groups, the more identifiable features, the more accurate the food image recognition model and narrow-sense insulin algorithm within that group, and the more suitable its parameters are for the patient. For example, a patient with the identifiable feature "16-year-old male patient" can find a suitable food image recognition model and narrow-sense insulin algorithm in either the "16-year-old patient" or "16-year-old male patient" group. Obviously, the food image recognition model and narrow-sense insulin algorithm found in the "16-year-old male patient" group are more suitable for him because the food image recognition model and narrow-sense insulin algorithm in the "16-year-old patient" group have also been learned from patient data from "female patients," which does not conform to the patient's physiological state. This will obviously affect the insulin infusion calculation for "male patients."

[0150] In other embodiments of the present invention, identifiable feature groups are stored as subsets in the system or cloud server 104. For example, "16-year-old male patients" is a subset of "16-year-old patients". When searching for the "16-year-old male patients" group, it is necessary to first search for the "16-year-old patients" group, and then search for the "male patients" group within the "16-year-old patients" group. Obviously, among the identifiable features, age and gender are two independent features, both of which can be used as subsets. When searching for the "16-year-old male patients" group, it is also possible to first search for the "male patients" group, and then search for the "16-year-old patients" group within the "male patients" group. Both of the above methods can ultimately point to the "16-year-old male patients" group, that is, "16-year-old patients" and "male patients" can be subsets of each other.

[0151] In this embodiment of the invention, the subset of identifiable features is geared towards individual patients or patient groups, and its catalog is editable. Manufacturers cannot exhaustively store all identifiable features in the system or cloud server 104 for patients to choose from; therefore, patients may not be able to retrieve applicable identifiable features. In this case, patients can edit the subset of identifiable features, adding, modifying, or deleting subsets, or request the server to add, modify, or delete subsets of identifiable features.

[0152] In this embodiment of the invention, due to the objective determinability of the nutritional content in food images, the parameters of the food image recognition model will tend to be consistent after multiple iterations. The food image recognition model parameters provided by each patient can be used to improve the food image recognition model and the food nutrient database. Improving the food image recognition model through deep learning using big data can enhance the accuracy of food image recognition, which is beneficial for the development of closed-loop artificial pancreas systems.

[0153] Referring to Figures 11a and 11b, in this embodiment of the invention, in step 2001, when the patient uses the closed-loop artificial pancreas system with food image recognition, they first need to take a picture of the food to be eaten using the imaging module 103 of the intelligent device. Considering that the food may be scattered in multiple utensils when the patient is eating, the patient can take one or more pictures of the food in each utensil. When taking multiple pictures, different angles can be selected to identify any food that may be obscured. The food image recognition model can identify the food that the patient has photographed repeatedly by judging the food's color, shadow, shape, size, and other elements in the food image, and remove the redundant food that has been photographed repeatedly when calculating the nutritional content of the food.

[0154] In step 2002, after the patient takes a picture of the food, if the food image recognition model is set on the smart device, the food image can be directly transmitted to the food image recognition application (hereinafter referred to as the application) where the food image recognition model is located. The application can be downloaded to the smart device by the patient or their guardian, or the application is a subroutine of the closed-loop artificial pancreas program, and the recognition of the food image can be completed in the closed-loop artificial pancreas program.

[0155] In step 2003, after the application completes the recognition of the food image, it can obtain the nutrient content (mainly including carbohydrate, fat and protein content) in the food image and display it to the patient through an interactive interface. The displayed content includes the recognized food type, weight, corresponding nutrient content and total nutrient content. The interactive interface can also prompt the patient to confirm the displayed content. If the patient confirms the food type, weight, corresponding nutrient content and total nutrient content obtained by the food image recognition model, it can be used to calculate the insulin infusion volume. If the patient believes that there is a significant deviation in the food type, weight, corresponding nutrient content and total nutrient content obtained by the food image recognition model, the patient can choose to re-recognize the food image or re-photograph the food and then re-recognize it until the recognition result of the food image recognition model is confirmed by the patient.

[0156] In step 2004, the application transmits the total nutrient content to the closed-loop artificial pancreas. Based on the total nutrient content, the insulin algorithm is invoked to calculate and administer the insulin infusion. The postprandial insulin infusion includes the basal dose and the bolus dose. Simultaneously, the system simulates a blood glucose curve (l1) for a period after the patient's meal based on the calculated insulin infusion. In this step, it is assumed that the insulin algorithm is stored in the closed-loop artificial pancreas. The insulin algorithm can also be stored in the application. After the application recognizes the food image, it can directly calculate the postprandial insulin infusion locally, then transmit the data to the closed-loop artificial pancreas for infusion. In this step, the insulin algorithm invoked can be the one stored in the application or the closed-loop artificial pancreas, or it can be an algorithm located in a narrow-sense insulin algorithm cluster. Invoking an algorithm from the narrow-sense insulin algorithm cluster requires the patient to first identify their identifiable features and then retrieve a suitable insulin algorithm based on those features. The more identifiable features a patient has, the more suitable the retrieved insulin algorithm will be for them, resulting in a more accurate postprandial insulin infusion.

[0157] In step 2005, while the postprandial insulin infusion begins, the detection module 100 detects and records the patient's actual postprandial blood glucose data to form the patient's actual blood glucose curve l2, which lasts for a period of time, such as 2 to 5 hours. This period of time coincides with the duration of the simulated blood glucose curve l1.

[0158] In step 2006, after the detection module 100 records blood glucose data for a preset time, the system compares the data of the simulated blood glucose curve l1 with the data of the actual blood glucose curve l2. Based on the comparison results, the insulin algorithm parameters are adjusted so that the simulated blood glucose curve l1 and the actual blood glucose curve l2 gradually become consistent. The parameter-adjusted narrow-sense insulin algorithm will be more suitable for the patient. Therefore, the parameter-adjusted narrow-sense insulin algorithm can be called the narrow-sense insulin algorithm. At the same time, on the one hand, the system sends the comparison results back to the application, which can adjust the parameters of the food image recognition model to optimize the food image recognition model. The parameter-adjusted insulin algorithm and the food image recognition model will be used for food image recognition and insulin calculation during the next meal. On the other hand, with the patient's consent, the parameter-adjusted insulin algorithm can be aggregated into the narrow-sense insulin algorithm cluster and classified into corresponding groups according to the patient's identifiable characteristics, so that the patient or other patients with the same identifiable characteristics can call it before meals.

[0159] In some embodiments of the present invention, due to the limited computing power and storage capacity of intelligent devices or closed-loop artificial pancreas, they may not be able to support complex image recognition and insulin infusion calculations, may not be able to store enough narrow-sense insulin algorithms, and may not be able to store a sufficiently rich food nutrient database. Considering this practical problem, a cloud server 104 can be connected to the closed-loop artificial pancreas system. The cloud server 104 can establish communication with the intelligent device or closed-loop artificial pancreas via wired or wireless means to achieve data interaction. After connecting to the cloud server 104, the cloud server 104 can provide powerful computing power and storage capacity support for the closed-loop artificial pancreas and intelligent devices. Image recognition models, narrow-sense insulin algorithm clusters, or food nutrient databases can be stored in the cloud server 104, and food image recognition or insulin infusion calculations can be completed in the cloud server 104, or the comparison between simulated blood glucose data and actual blood glucose data can be completed, or even the correction of insulin algorithm or image recognition model parameters can be completed. In this way, food recognition and insulin infusion calculation during meals can be completed without using a high-performance closed-loop artificial pancreas and intelligent devices, saving patients usage costs.

[0160] After connecting to the cloud server 104, the method of using the closed-loop artificial pancreas system will change. Referring to Figure 11c, the specific steps are as follows:

[0161] In step 3001, the patient uses a smart device to take a picture of the food, and the food image recognition model identifies the food image.

[0162] In step 3002, the food image recognition model is set in the cloud server 104. After taking a picture of the food, the patient uploads the food image to the cloud server 104 through a smart device and completes the food image recognition.

[0163] In step 3003, the cloud server 104 transmits the identified food type, weight, corresponding nutrient content, and total nutrient content data back to the smart device and displays them to the patient through an interactive interface. If the patient confirms that the food type, weight, corresponding nutrient content, and total nutrient content obtained by the food image recognition model are accurate, they can be used to calculate the postprandial insulin infusion volume. If the patient believes that there is a significant deviation in the food type, weight, corresponding nutrient content, and total nutrient content obtained by the food image recognition model, the patient can choose to re-identify the food image or re-photograph the food and then re-identify it, until the recognition result of the food image recognition model is confirmed by the patient.

[0164] In step 3004, the patient invokes an insulin algorithm or retrieves an insulin algorithm with the same identifiable features from the narrow insulin algorithm cluster. Based on the determined insulin algorithm and nutrient content, the cloud server 104 calculates the patient's postprandial insulin infusion volume and simulates the patient's postprandial blood glucose changes to obtain simulated blood glucose data, forming a simulated blood glucose curve l1. The cloud server 104 sends the postprandial insulin infusion volume data to the closed-loop artificial pancreas and completes the infusion. The postprandial insulin infusion volume includes the basal dose and the bolus dose.

[0165] In step 3005, while insulin infusion begins, the detection module 100 detects and records the patient's actual blood glucose data to form an actual blood glucose curve l2, which lasts for a period of time, such as 2 to 5 hours. This period of time coincides with the duration of the simulated blood glucose curve l1, and the actual blood glucose data during this period is transmitted to the cloud server 104.

[0166] In step 3006, the cloud server 104 compares the simulated blood glucose curve l1 data with the actual blood glucose curve l2 data, and adjusts the insulin algorithm parameters based on the comparison results, so that the simulated blood glucose curve l1 and the actual blood glucose curve l2 gradually become consistent. The insulin algorithm with adjusted parameters will be more suitable for the patient, so it can be called the narrow insulin algorithm. At the same time, the cloud server 104 uses the comparison results and the adjustment results of the insulin algorithm parameters to adjust the parameters of the food image recognition model to optimize the food image recognition model. On the one hand, the narrow insulin algorithm and the food image recognition model will be used for food image recognition and insulin calculation during the next meal. On the other hand, with the patient's consent, the narrow insulin algorithm can be aggregated into the narrow insulin algorithm cluster, classified into the corresponding group according to the patient's identifiable characteristics, and replace the insulin algorithm before parameter adjustment, so that the patient or other patients with the same identifiable characteristics can call it before meals.

[0167] In this embodiment of the invention, the calculation of insulin infusion volume can be completed in a closed-loop artificial pancreas, a smart device, or a cloud server 104, and the specific implementation process will not be described in detail.

[0168] In this embodiment of the invention, the recognition of food images can be completed in a closed-loop artificial pancreas, a smart device, or a cloud server 104, and the specific implementation process will not be described in detail.

[0169] In this embodiment of the invention, the simulation calculation of the simulated blood glucose curve l1 can be completed in a closed-loop artificial pancreas, a smart device, or a cloud server 104, and the specific implementation process will not be described in detail.

[0170] Referring to Figure 12, which is a schematic diagram of another closed-loop infusion method based on food image recognition according to an embodiment of the present invention, the insulin algorithm can be integrated into the food image recognition model. Once the food image recognition model identifies the type and weight of the food, it is not necessary to identify the nutritional content; the postprandial insulin infusion volume can be directly determined through the insulin algorithm, as described below. The model resulting from the normalization of the food image recognition model and the insulin algorithm is referred to as the large model in this invention.

[0171] In this embodiment of the invention, the insulin algorithm can be divided into two methods: logical operation and lookup table, which are described in detail below.

[0172] In this embodiment of the invention, the large model can be stored in a smart device, a closed-loop artificial pancreas system, or a cloud server 104. After receiving food image input, it can directly output the postprandial insulin infusion amount, and the postprandial insulin infusion is completed by the closed-loop artificial pancreas.

[0173] Taking the storage of a large model in a smart device as an example, in this embodiment of the invention, in step 4001, the large model still needs to use the imaging module 103 of the smart device to acquire food images. The steps for acquiring food images are the same as those described above and will not be repeated here.

[0174] In step 4002, the food image is transmitted to a food image recognition application, where a large model identifies the type and weight of the food in the image.

[0175] In step 4003, after the large model identifies the type and weight of food in the food image, it is displayed to the patient through the interactive interface of the smart device, prompting for confirmation. If the patient confirms the type and weight of food displayed on the interactive interface, the postprandial insulin infusion dose can be output through the preset insulin algorithm. If the patient believes that the identified type and weight of food are significantly inaccurate, the patient can choose to re-identify the food image or re-take a picture of the food for identification, until the identification result of the large model is confirmed by the patient.

[0176] In step 4004, after the patient confirms the type and weight of the food, the large model calculates the patient's postprandial insulin infusion amount based on the type and weight of the food using a preset insulin algorithm. The preset insulin algorithm can be one or more of the following logical operations: neural network, PID, MPC, etc. It can calculate the postprandial insulin infusion amount based on the type and weight of the food. After recognizing the food image, the large model calculates the postprandial insulin infusion amount for each type of food separately, and then adds up the postprandial insulin infusion amounts for all foods to obtain the total postprandial insulin infusion amount. For example, the large model recognizes 500g of potatoes, 250g of pasta, and 250g of beef from the food image provided by the patient. Using the preset insulin algorithm, it calculates the postprandial insulin infusion amount corresponding to 500g of potatoes as I1, 250g of pasta as I2, and 250g of beef as I3. Therefore, the patient's total postprandial insulin infusion amount is I1 + I2 + I3. The above is for illustrative purposes only.

[0177] The preset insulin algorithm can also be a lookup table. Based on the lookup table, the postprandial insulin infusion amount is retrieved according to the type and weight of the food. The postprandial insulin infusion amount for each food and its corresponding weight is stored in the large model in the form of a lookup table, as shown in Table 3.1. For example, the large model identifies 300g of potatoes, 200g of pasta, and 200g of beef from the food image provided by the patient. By retrieving the lookup table, the postprandial insulin infusion amount corresponding to 300g of potatoes is I4, the postprandial insulin infusion amount corresponding to 200g of pasta is I5, and the postprandial insulin infusion amount corresponding to 200g of beef is I6. Then, the patient's total postprandial insulin infusion amount is I4+I5+I6. The above is only an illustrative description.

[0178] Table 3.1 Food-Insulin Lookup Table

[0179] Table 3.1 is one way to express a lookup table, which includes at least the "Food Type," "Food Weight," and "Insulin Infusion Volume" columns. More detailed lookup tables may also include additional food information, such as "Food Origin," "Cooking Method," and "Food Storage Time." After taking a picture of the food, patients can select relevant additional information on the smart device's interface, which helps the food image recognition model better identify the food image. This additional information is crucial for understanding the nutrient content of food. This information may be difficult for the food image recognition model to recognize, which can affect the quantitative analysis of nutrients in food. Therefore, before the food image is recognized, patients need to input or select relevant information as additional information.

[0180] In large models, whether postprandial insulin infusion is obtained through logical operations or lookup tables, commonly used empirical parameters are initially used. The same type and weight of food may have different effects on blood glucose in different patients due to factors such as meal time, patient's digestive level, and cooking method. Insulin algorithms cannot exhaustively consider all influencing factors, so insulin algorithms can be modified.

[0181] After obtaining the patient's postprandial insulin infusion volume, the large model will simulate the patient's postprandial blood glucose data based on the postprandial insulin infusion volume, forming a simulated blood glucose curve l3. This simulation process lasts for a period of time, such as 2 to 5 hours.

[0182] In step 4005, the application transmits the patient's postprandial insulin infusion data to the closed-loop artificial pancreas, and the system begins postprandial insulin infusion. At the same time, the system records the actual blood glucose data detected by the detection module 100, forming an actual blood glucose curve l4. The duration of the actual blood glucose curve l4 coincides with the duration of the simulated blood glucose curve l3.

[0183] In step 4006, the system transmits the actual blood glucose data to the smart device. The large model compares the differences between the simulated blood glucose curve l3 data and the actual blood glucose curve l4 data, and corrects the simulated blood glucose curve until it matches the actual blood glucose curve. The comparison results are used to correct the preset insulin logic operation parameters or lookup table data in the large model, and can even be used to correct the image recognition model parameters.

[0184] In this embodiment of the invention, the insulin logic operation parameters or lookup table data will gradually adapt to the patient's dietary habits after each correction. Therefore, the correction process of the insulin algorithm parameters in this scheme is a learning process. During the learning process, the insulin algorithm becomes more and more suitable for the patient's physiological characteristics and lifestyle. As a result, the calculated postprandial insulin infusion volume will be more accurate, which will help the patient's diabetes treatment.

[0185] In this embodiment of the invention, when revising the lookup table, the data in the "Insulin Infusion Volume" column can be modified, and data in the "Type" and "Weight" columns can be added, modified, or deleted. The revised lookup table can still be aggregated and grouped according to the patient's identifiable characteristics for other patients to access. As mentioned above, the revised lookup table can be referred to as a narrow-sense lookup table to distinguish it from the original lookup table.

[0186] In some embodiments of the present invention, the food image recognition model and the insulin algorithm can be separate and stored in different modules. For example, the food image recognition model is stored in a smart device, and the insulin algorithm is stored in one of the detection module 100, program module 101, or infusion module 102 of the closed-loop artificial pancreas system. After the smart device completes food image recognition, it transmits the food type and weight data to the closed-loop artificial pancreas, whereby the system calculates or queries the insulin dosage and completes the insulin infusion. Alternatively, the food image recognition model can be stored in a cloud server, and the insulin algorithm can be stored in a smart device. Or, for another example, the food image recognition model can be stored in a cloud server, and the insulin algorithm can be stored in one of the detection module 100, program module 101, or infusion module 102 of the closed-loop artificial pancreas system.

[0187] In this embodiment of the invention, the order of some of the above steps can be interchanged. For example, the step of calculating simulated blood glucose data can be set before or after the step of obtaining actual blood glucose data. This does not affect the calculation and infusion of postprandial insulin for patients, nor does it affect the comparison between simulated blood glucose data and actual blood glucose data to correct the narrow insulin algorithm parameters.

[0188] Figure 13a is a schematic diagram of a personalized closed-loop infusion system based on a cloud server according to an embodiment of the present invention, and Figure 13b is a schematic diagram of a personalized closed-loop infusion method according to an embodiment of the present invention.

[0189] Referring to Figures 13a and 13b, in this embodiment of the invention, leveraging the powerful storage and computing capabilities of the cloud server 104, relevant information about the patient group is stored in the cloud server 104. This includes patient blood glucose data records, insulin infusion data records, meal data records, exercise data records, sleep data records, etc. This information can be uploaded to the cloud server 104 by the patient or their caregiver. Similarly, these data records are categorized into groups according to the patient's identifiable characteristics. Data records within a group point to their corresponding patient, and each patient in the group wears a closed-loop artificial pancreas system with a corresponding insulin algorithm; that is, the data records within a group correspond to the insulin algorithm.

[0190] In this embodiment of the invention, the patients to whom the data records within the same group point to have similar lifestyles or physiological characteristics, and their blood glucose data records, insulin infusion records, meal records, exercise records, sleep records, and other related information can be referenced and compared with each other. The cloud server 104 can also use data mining or machine learning algorithms to perform in-depth analysis of the stored data records. These algorithms can help discover patterns, correlations, and trends in the data, thereby extracting valuable information. For example, in the group of "16-year-old male patients," cloud server 104, through analysis of data records, discovered that during the period from 12:00 to 1:00 PM on a certain day, two patients were asleep. Patient A's blood glucose data remained stable and within the safe blood glucose threshold, while Patient B's blood glucose data fluctuated significantly and had the potential to exceed the safe blood glucose threshold. Therefore, cloud server 104 can determine that Patient A's insulin algorithm parameters are more suitable for blood glucose control during the 12:00 to 1:00 PM period for this group of "16-year-old male patients." Cloud server 104 will call Patient A's insulin algorithm parameters and push them to Patient B to correct Patient B's insulin algorithm parameters, so that the next time Patient B is asleep during the 12:00 to 1:00 PM period, their blood glucose can be better controlled. Alternatively, cloud server 104 can also use its built-in parameter fine-tuning algorithm to fine-tune Patient B's insulin algorithm parameters by referring to Patient A's insulin algorithm parameters, and then push the fine-tuned parameters to Patient B to correct Patient B's insulin algorithm. The above is only an illustrative description. When comparing patients' blood glucose data, it is also necessary to consider the patient's activities before and after falling asleep. For example, cloud server 104 records that patient A had a period of calm rest before falling asleep, while patient B ate before falling asleep. Obviously, patient A's insulin algorithm parameters are no longer fully applicable to patient B, and cloud server 104 will not push patient A's insulin algorithm parameters to patient B. In this case, cloud server 104 needs to find the blood glucose data of patient C, who ate before falling asleep, in the same group, and compare it with patient B's blood glucose data. Based on the comparison results, the insulin algorithm parameters of the patient with better blood glucose control are pushed to another patient, or the insulin algorithm parameters are fine-tuned and calculated before being pushed to another patient.

[0191] In this embodiment of the invention, when the cloud server 104 pushes insulin algorithm parameters to the patient, it also needs to be confirmed by the patient or their guardian before it can be used to correct the patient's insulin algorithm.

[0192] In this embodiment of the invention, the cloud server 104 can actively or passively push insulin algorithm parameters to the patient, and this can be set by the patient's preference. In the passive setting, the patient sends an instruction to the cloud server 104 to push insulin algorithm parameters; in response to the instruction, the cloud server 104 sends the required insulin algorithm parameters to the patient's closed-loop artificial pancreas system.

[0193] In this embodiment of the invention, regardless of whether it is patient A or patient B, their insulin algorithm parameters can be uploaded or downloaded to the cloud server 104 via the network periodically or in real time. When uploading insulin algorithm parameters, later parameters can overwrite earlier insulin algorithm parameters stored in the cloud server 104, or both later and earlier parameters can be stored in the cloud server 104, forming a complete insulin algorithm parameter record. For example, a patient can choose to upload their insulin algorithm parameters to the cloud server 104 at 8:00 AM every day, and simultaneously instruct the cloud server 104 to download the applicable insulin algorithm parameters and update the original insulin algorithm. As another example, a patient can also choose to upload their insulin algorithm parameters to the cloud server 104 every hour, and simultaneously instruct the cloud server 104 to download the applicable insulin algorithm parameters and correct the original insulin algorithm. Of course, uploading and downloading insulin algorithm parameters do not necessarily need to be performed simultaneously; patients can also choose to upload only or download only the insulin algorithm parameters.

[0194] In this embodiment of the invention, due to the large number of patients using the closed-loop artificial pancreas system, the instantaneous data volume generated when patients upload or download insulin algorithm parameters is enormous. The cloud server 104 needs to rapidly group and analyze this large amount of data to meet the patients' data upload and download demands. Based on this usage scenario, the cloud server 104 can also utilize distributed computing frameworks, such as Hadoop and Spark, to perform multi-threaded parallel processing of big data to improve data processing efficiency. Those skilled in the art will understand that any step swaps that do not affect the implementation of this solution, including both the solution before and after the step swaps, should be included within the scope of protection of this invention.

[0195] Those skilled in the art will understand that regardless of whether the food image recognition model and the insulin algorithm are unified or separated, and no matter how the storage location changes, they should all be included within the scope of protection of this invention.

[0196] In summary, this invention discloses a personalized closed-loop infusion system and method. It uploads patient-related information to a cloud server and groups this information according to the patient's identifiable characteristics. Within the same group, it compares the relevant information of different patients, such as blood glucose data records and insulin infusion data records. Under the premise of similar or identical patient information, the insulin algorithm parameters of patients with better blood glucose control can be used to correct the insulin algorithm of patients with poorer blood glucose control. The closed-loop artificial pancreas system no longer needs to learn the patient's physiological characteristics and lifestyle habits over a long period of time, and can quickly obtain a personalized insulin algorithm suitable for the patient, which is beneficial to the treatment of diabetes.

[0197] While specific embodiments of the invention have been described in detail by way of examples, those skilled in the art should understand that the above examples are for illustrative purposes only and are not intended to limit the scope of the invention. Those skilled in the art should understand that modifications can be made to the above embodiments without departing from the scope and spirit of the invention. The scope of the invention is defined by the appended claims.

Claims

1. A closed-loop artificial pancreas system, characterized in that: include The detection module is used to acquire the patient's actual blood glucose data; An insulin algorithm determines the patient's insulin infusion volume based at least on the actual blood glucose data. The infusion module completes insulin infusion based on the stated insulin infusion volume; and A cloud server is used to receive relevant patient information and classify the relevant information into groups according to identifiable features; The cloud server is also used to analyze the patient's relevant information, determine the correction parameters of the insulin algorithm based on the analysis results, and correct the insulin algorithm based on the correction parameters.

2. The closed-loop artificial pancreas system according to claim 1, characterized in that, The patient's relevant information includes at least blood glucose data records, insulin infusion records, meal records, exercise records, and sleep records.

3. The closed-loop artificial pancreas system according to claim 1, characterized in that, The identifiable features are associated with the patient’s physiological state or lifestyle.

4. The closed-loop artificial pancreas system according to claim 1, characterized in that, The groups are independent of each other.

5. The closed-loop artificial pancreas system according to claim 1, characterized in that, The groups that have at least one of the same identifiable features are subsets of each other.

6. The closed-loop artificial pancreas system according to claim 1, characterized in that, The insulin algorithm parameters for different patients within the same group can be mutually modified.

7. The closed-loop artificial pancreas system according to claim 6, characterized in that, The correction of the insulin algorithm parameters is performed actively or passively by the cloud server.

8. The closed-loop artificial pancreas system according to claim 7, characterized in that, Before modifying the insulin algorithm parameters, patient confirmation is required.

9. The closed-loop artificial pancreas system according to claim 1, characterized in that, It also includes a program module, which is used to control the infusion module to complete the insulin infusion.

10. The closed-loop artificial pancreas system according to claim 9, characterized in that, The insulin algorithm is located in the detection module, the program module, the infusion module, or the cloud server.

11. A closed-loop insulin infusion method, comprising: a detection module, an infusion module, a cloud server, and an insulin algorithm, characterized in that, Including the following steps: I. The detection module is used to obtain actual blood glucose data, and the cloud server receives relevant information uploaded by the patient; II. The cloud server groups the patient's relevant information according to the patient's identifiable characteristics; III. The cloud server analyzes the relevant information of patients in the same group; IV. Based on the analysis results, determine the correction parameters of the insulin algorithm, and correct the insulin algorithm based on the correction parameters; V. Based on the modified insulin algorithm and the actual blood glucose data, the insulin infusion volume is calculated, and the infusion module completes the insulin infusion.

12. The closed-loop insulin infusion method according to claim 11, characterized in that, Step I also includes uploading the patient's insulin algorithm parameters to the cloud server.

13. The closed-loop insulin infusion method according to claim 11, characterized in that, In step III, the relevant patient information analyzed includes at least blood glucose data records.

14. The closed-loop insulin infusion method according to claim 13, characterized in that, In step IV, based on the analysis results of the blood glucose data records, the method further includes determining the insulin algorithm parameters of the patients with better blood glucose control as the correction parameters, and pushing the correction parameters to other patients.

15. The closed-loop insulin infusion method according to claim 14, characterized in that, It also includes the patient sending an instruction to the cloud server, and in response to the instruction, the cloud server pushing the correction parameters to the patient.

16. The closed-loop insulin infusion method according to claim 13, characterized in that, In step IV, based on the analysis results of the recorded blood glucose data, the cloud server also calculates the correction parameters based on the insulin algorithm parameters of the patients with better blood glucose control, and pushes the correction parameters to other patients.

17. The closed-loop insulin infusion method according to claim 11, characterized in that, Step IV also includes the step of the patient downloading the correction parameters to the cloud server.

18. The closed-loop insulin infusion method according to claim 11, characterized in that, Step IV also includes a step where the patient confirms that the insulin algorithm has been corrected using the corrected parameters.

Images