Drug dosage recommendation method based on scene recognition, electronic equipment and system

Abstract

The invention provides a scene recognition-based drug dosage recommendation method, electronic equipment and a scene recognition-based drug dosage recommendation system. The method comprises the following steps of: obtaining behavior information of a target object in a current scene; according to the behavior information of the target object, determining a drug reference dosage for adjusting the blood glucose level of the target object; and obtaining drug dosage recommendation of the target object according to the drug reference dosage. The behavior information of the target object in the current scene is identified, then the drug reference dosage is determined according to the identified behavior information, and the drug dosage recommendation of the target object is determined according to the drug reference dosage, so that the drug dosage recommendation is associated with the behavior information of the target object in the current scene; the influence of the external environment on the blood sugar of the diabetic patient is fully considered, and the matching degree between the drug dosage recommendation and the current scene where the target object is located is improved.

Description

Technical Field

[0001] The present application belongs to the technical field of blood sugar control, and in particular, relates to a method, device, electronic equipment system and storage medium for recommending drug dosage based on scene recognition. Background Art

[0002] Nowadays, with the rapid increase of diabetes patients worldwide, diabetes has become the third most serious chronic disease threatening human health after tumors and cardiovascular diseases. Current medical technology cannot cure diabetes. For most diabetic patients, the only way to achieve the purpose of treatment is to inject insulin, that is, inject insulin into diabetic patients by imitating the secretion method of normal human insulin, so as to stabilize the blood sugar level in diabetic patients.

[0003] At present, for diabetic patients, a special insulin pump is generally used to inject insulin regularly and quantitatively to help restore insulin-induced diabetes to normal levels. However, due to the influence of many factors, the blood sugar level in diabetic patients is not fixed, but changes dynamically. In the case of dynamic changes in blood sugar levels, if a quantitative insulin injection dose is used, the blood sugar control effect of diabetic patients is poor. When using an insulin pump to inject insulin by themselves, blood sugar fluctuates and is difficult to control, increasing the risk of hyperglycemia and other complications. How to accurately provide a recommended insulin dosage has become an urgent problem to be solved in this field. Summary of the invention

[0004] The embodiments of the present application provide a method, electronic device, and system for recommending drug dosage based on scene recognition, which can determine the reference dosage of drugs for adjusting the blood sugar level of the target object based on the behavior information of the target object in the current scene, and obtain the drug dosage recommendation of the target object based on the drug reference dosage. The influence of the current scene of the target object on the drug dosage recommendation is fully considered, thereby obtaining a more accurate drug dosage recommendation.

[0005] In a first aspect, an embodiment of the present application provides a method for recommending drug dosage based on scene recognition, the method comprising: obtaining behavior information of a target object in a current scene; determining a reference drug dosage for regulating the blood sugar level of the target object based on the behavior information of the target object; and obtaining a drug dosage recommendation for the target object based on the reference drug dosage.

[0006] The drug dosage recommendation method based on scene recognition in the first aspect can identify the behavior information of the target object in the current scene, and then determine the reference drug dosage based on the identified behavior information, and determine the drug dosage recommendation for the target object based on the reference drug dosage, so that the drug dosage recommendation is associated with the behavior information of the target object in the current scene, fully considering the impact of the external environment on the blood sugar of diabetic patients, and improving the matching degree of the drug dosage recommendation with the current scene where the target object is located.

[0007] In a possible implementation of the first aspect, obtaining the behavior information of the target object in the current scene includes: obtaining sound scene data, the sound scene data including sound data in the current scene; performing foreground-background sound separation on the sound scene data to obtain background sound; and determining the behavior information of the target object based on the background sound. In this embodiment, the background sound is first separated from the sound scene data through the foreground-background sound separation technology, and then the behavior information is identified for the background sound, which can improve the accuracy of behavior information identification.

[0008] In a possible implementation of the first aspect, determining the behavior information of the target object based on the background sound includes: obtaining voiceprint features of the background sound; inputting the voiceprint features into a behavior recognition model to obtain the behavior information of the target object, the behavior information including: multiple behavior state categories of the target object and the probability of each behavior state category, each behavior state category representing a type of behavior state of the target object. In this embodiment, after extracting the voiceprint features of the background sound, the voiceprint features are recognized by the behavior recognition model to obtain the behavior information, so that the voiceprint features input into the behavior recognition model can well reflect the characteristics of the background sound, thereby improving the accuracy of the behavior information output by the behavior recognition model.

[0009] In a possible implementation of the first aspect, obtaining the behavior information of the target object in the current scene also includes: obtaining inertial sensor data and navigation positioning sensor data of the target object in the current scene; updating the probability of each behavior state category according to the inertial sensor data; and updating the probability of each behavior state category after the update for a second time according to the navigation positioning sensor data; or obtaining the behavior information of the target object in the current scene also includes: updating the probability of each behavior state category according to the navigation positioning sensor data; and updating the probability of each behavior state category after the update for a second time according to the inertial sensor data. In this embodiment, by updating the probability of the behavior state category in the behavior information through multiple sensor data, the complexity of the scene can be better coped with, and the accuracy and stability of the behavior information can be further improved.

[0010] In a possible implementation of the first aspect, the inertial sensor data includes acceleration information, and the probability of each behavior state category is updated according to the inertial sensor data, including: estimating the number of steps of the target object in the next preset time period according to the acceleration information; determining the exercise intensity of the target object in the next preset time period according to the number of steps; and updating the probability of each behavior state category according to the exercise intensity. In this embodiment, the number of steps of the target object in the next preset time period is estimated using the acceleration information, and the exercise intensity of the target object in the next preset time period is determined according to the estimated number of steps, and then the probability is updated according to the exercise intensity. The method is simple and easy to implement, and fully considers the impact of exercise intensity on blood sugar, thereby improving the accuracy of the probability.

[0011] In a possible implementation manner of the first aspect, the exercise intensity is obtained according to the following formula:

[0012]

[0013] Where: I ex represents the exercise intensity of the target object in the next preset time period, steps represents the number of steps of the target object in the next preset time period, and stepsMax represents the maximum number of steps of the target object in the preset time period.

[0014] In a possible implementation manner of the first aspect, the multiple behavior state categories include n behavior state categories, where n is a positive integer greater than 1; each behavior state category is a motion state or a non-motion state, and the probability of the i-th behavior state category is updated according to the following formula:

[0015]

[0016] f(l ex )=l ex *ΔP1

[0017] Where: P i ' represents the probability of the updated i-th behavior state category, P i represents the probability of the i-th behavior state category before updating, ΔP1 represents the first threshold set in advance, and I ex Indicates the target object's exercise intensity in the next preset period of time.

[0018] In a possible implementation of the first aspect, the navigation and positioning sensor data includes: positioning information and a signal strength value at a current moment; and updating the probability of each behavior state category according to the navigation and positioning sensor data, including: if the signal strength value is greater than a preset strength threshold, updating the probability of each behavior state category according to a first comparison result and a second comparison result, the first comparison result being a comparison result obtained by comparing the positioning information with historical positioning information, and the second comparison result being a comparison result obtained by comparing the first behavior state category with the historical behavior information; wherein: the first behavior state category is a behavior state category corresponding to a maximum probability among multiple behavior state categories, the historical positioning information includes multiple first positioning information of the target object, each first positioning information corresponds to a moment before the current moment; the historical behavior information includes multiple first behavior information of the target object, each first behavior information corresponds to a moment before the current moment; and if the signal strength value is less than or equal to the preset strength threshold, updating the probability of each behavior state category according to the second comparison result. In this embodiment, when the signal strength is greater than a preset threshold, it indicates that the accuracy of the positioning information obtained at this time is relatively high. Therefore, when updating the probability of each behavior state category, the positioning information can be taken into account, and the update is performed from the perspectives of positioning information and behavior information. In addition, historical positioning information and historical behavior information are taken into account during the update. The historical positioning information and historical behavior information reflect the behavior patterns of the target object. By introducing historical positioning information and historical behavior information, the accuracy of the updated probability can be improved.

[0019] In a possible implementation of the first aspect, the multiple behavior state categories include n behavior state categories, where n is an integer greater than 1; the multiple first positioning information include m groups of first positioning information, each group of first positioning information includes first positioning information corresponding to multiple moments in the same day, and m is an integer greater than 1; the multiple first behavior information includes p groups of first behavior information, each group of first behavior information includes multiple first behavior information corresponding to multiple moments in the same day, and p is an integer greater than 1; based on the first comparison result and the second comparison result, updating the probability of each behavior state category, including: determining k target positioning information from the m groups of first positioning information, and each target positioning information corresponding to The time of the day corresponds to the current time, and the probability of the target positioning information appearing in the m groups of first positioning information is greater than a preset probability threshold; from the p groups of first behavior information, the target behavior information is determined, the time of the day corresponding to the target behavior information corresponds to the current time, and in the p groups of first behavior information, the probability of the behavior state category in the target behavior information is the largest; if the positioning information successfully matches any one of the k target positioning information, and the first behavior state category is the same as the target behavior state category, and the target behavior state category is the behavior state category in the target behavior information, then the probability of the i-th behavior state category in the n behavior state categories is updated according to the following formula:

[0020]

[0021] Where P i ′ represents the probability of the updated i-th behavior state category, P i represents the probability of the i-th behavior state category before updating, and ΔP2 represents a preset second threshold.

[0022] In a possible implementation of the first aspect, the positioning information includes a first coordinate, the first target positioning information includes a second coordinate, the first target positioning information is any one of the k target positioning information, and the method further includes: determining a distance value between the first coordinate and the second coordinate; if the distance value is less than a preset distance threshold, determining that the positioning information successfully matches the first target positioning information; if the distance value is greater than or equal to the preset distance threshold, determining that the positioning information fails to match the second target positioning information. In this embodiment, whether the positioning information and the target positioning information match is determined based on the coordinate distance, and the method is simple and easy to implement.

[0023] In a possible implementation of the first aspect, the multiple behavior state categories include n behavior state categories, where n is an integer greater than 1; the multiple first behavior information includes p groups of first behavior information, where p is an integer greater than 1, and each group of first behavior information includes multiple first behavior information corresponding to multiple moments in the same day; based on the second comparison result, updating the probability of each behavior state category, including: determining target behavior information from the p groups of first behavior information, the moment of the day corresponding to the target behavior information corresponds to the current moment, and in the first behavior information, the probability of the behavior state category in the target behavior information is the largest; if the first behavior state category is the same as the target behavior state category, and the target behavior state category is the behavior state category in the target behavior information, then the probability of the i-th behavior state category among the n behavior state categories is updated according to the following formula:

[0024]

[0025] Where P i ′ represents the probability of the updated i-th behavior state category, P i represents the probability of the i-th behavior state category before updating, and ΔP3 represents a preset third threshold.

[0026] In a possible implementation of the first aspect, the method further includes: obtaining a blood glucose measurement value of the target object; obtaining a drug dosage recommendation for the target object based on a drug reference dosage, including: obtaining a drug dosage recommendation for the target object based on the drug reference dosage and the blood glucose measurement value of the target object. In this embodiment, the blood glucose measurement value of the target object is introduced when determining the drug dosage recommendation, so that the safety and accuracy of the obtained drug dosage recommendation can be further guaranteed.

[0027] In a second aspect, an embodiment of the present application provides a drug recommendation device, which includes: a functional unit for executing each step of the method in any embodiment of the first aspect above.

[0028] In a third aspect, an embodiment of the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method in any one of the embodiments of the first aspect above is implemented.

[0029] In a fourth aspect, an embodiment of the present application provides a drug infusion system, the system comprising: an electronic device and a drug infusion device in any embodiment of the third aspect above, wherein: the electronic device is used to obtain behavior information of a target object in a current scene; based on the behavior information of the target object, a reference dosage of a drug for regulating the blood glucose level of the target object is determined; based on the reference dosage of the drug, a drug dosage recommendation for the target object is obtained, and the drug dosage recommendation is sent to the drug infusion device, and the drug infusion device is used to infuse the drug into the target object according to the drug dosage recommendation.

[0030] In a possible implementation of the fourth aspect, the system also includes: a blood glucose measuring device, wherein: the blood glucose measuring device is used to monitor the blood glucose of the target object and obtain the blood glucose measurement value of the target object, and the electronic device is also used to obtain the blood glucose measurement value from the blood glucose monitoring device, and obtain a drug dosage recommendation for the target object based on the drug reference dosage and the blood glucose measurement value.

[0031] In a fifth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method described in any embodiment of the first aspect above are implemented.

[0032] In a sixth aspect, an embodiment of the present application provides a computer program product, which, when executed on a server, enables the server to execute the steps of the method described in any one of the embodiments of the first aspect above.

[0033] In a seventh aspect, an embodiment of the present application provides a chip, comprising: a processor, configured to call and run a computer program from a memory, so that an electronic device equipped with the chip executes the steps of the method described in any embodiment of the first aspect above.

[0034] It can be understood that the beneficial effects of the second to seventh aspects mentioned above can be found in the relevant description of the first aspect mentioned above, and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0036] Figure 1 is a structural schematic diagram of a drug infusion system provided in one embodiment of the present application;

[0037] Figure 2It is a flowchart of a method for recommending drug dosage based on scene recognition provided by an embodiment of the present application;

[0038] Figure 3 It is a flow chart of updating the probability of each behavior state category according to the inertial sensor data in the drug dosage recommendation method based on scene recognition provided by an embodiment of the present application;

[0039] Figure 4 In the drug dosage recommendation method based on scene recognition provided in one embodiment of the present application, according to the navigation positioning sensor data;

[0040] Figure 5 It is a structural schematic diagram of a drug dosage recommendation device based on scene recognition provided in an embodiment of the present application;

[0041] Figure 6 It is a structural schematic diagram of an electronic device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0042] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0043] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof.

[0044] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0045] As used in the specification and appended claims of this application, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce [described condition or event] is detected" or "in response to detecting [described condition or event]", depending on the context.

[0046] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0047] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0048] At present, for some special diseases (such as diabetes), the administration time and dosage of drugs (such as insulin) are affected by various factors. Among them, the scene in which the patient is located has a certain influence on the administration time and dosage. For example, when using an insulin pump to inject insulin to diabetic patients, the insulin injection dosage generally uses a quantitative dosage, or sets different dosages according to time periods. If the impact of the user's scene on the blood sugar of diabetic patients is not considered, blood sugar control is not ideal, blood sugar fluctuates and is difficult to control, and the risk of hyperglycemia, hypoglycemia and other complications is increased. Therefore, how to accurately identify the impact of the external environment on the blood sugar of diabetic patients and provide a recommended insulin dosage plan has become an urgent problem to be solved in this field.

[0049] In order to solve the above technical problems, the present application provides a drug dosage recommendation method based on scene recognition, a drug recommendation device, an electronic device, a drug infusion system and a computer-readable storage medium. The drug dosage recommendation method based on scene recognition includes: first obtaining the behavior information of the target object in the current scene; then determining the reference dosage of the drug used to adjust the blood sugar value of the target object according to the behavior information of the target object; and finally obtaining the drug dosage recommendation of the target object according to the reference dosage of the drug. The blood sugar of diabetic patients is affected by the external environment such as eating and exercising, and cannot be regulated by the insulin secreted by themselves in time, resulting in the imbalance of the blood sugar environment in the body, affecting physical health. Therefore, the drug dosage recommendation method based on scene recognition in the embodiment of the present application can identify the behavior information of the target object in the current scene, and then determine the reference dosage of the drug according to the identified behavior information, and determine the drug dosage recommendation of the target object according to the reference dosage of the drug, so that the drug dosage recommendation is associated with the behavior information of the target object in the current scene, fully considering the impact of the external environment on the blood sugar of diabetic patients, and improving the matching degree of the drug dosage recommendation with the current scene where the target object is located.

[0050] The following is an exemplary description of the drug dosage recommendation method based on scene recognition, drug recommendation device, electronic device, drug infusion system and computer-readable storage medium provided in the present application in combination with specific embodiments.

[0051] See also Figure 1 FIG. 1 is a schematic diagram of the structure of a drug infusion system provided in one embodiment of the present application. Figure 1 As shown, the drug infusion system in the embodiment of the present application includes: an electronic device 100 and a drug infusion device 200. There is a communication connection between the electronic device 100 and the drug infusion device 200. The communication connection established between the two can be a wired connection or a wireless connection, for example, the wireless connection can be a Bluetooth connection, or a WIFI connection, etc., which is not limited in the embodiment of the present application. Among them:

[0052] The electronic device 100 obtains the behavior information of the target object in the current scene; determines the reference dosage of the drug used to adjust the blood sugar level of the target object according to the behavior information of the target object; obtains the recommended dosage of the drug for the target object according to the reference dosage of the drug, and sends the recommended dosage of the drug to the drug infusion device. The drug infusion device 200 infuses the drug to the target object according to the recommended dosage of the drug.

[0053] The drug infusion system in the embodiment of the present application uses an electronic device to obtain the behavior information of the target object in the current scene, and determines the reference dosage of the drug based on the behavior information, thereby determining the recommended dosage of the drug based on the reference dosage of the drug, and the drug infusion device 200 infuses the drug to the target object based on the recommended dosage of the drug. As a result, the recommended dosage of the drug based on which the drug infusion device performs drug infusion is associated with the behavior information of the target object in the current scene, so that the drug infusion fully considers the impact of the current scene where the target object is located on the blood sugar level of the diabetic patient, thereby ensuring the effect of blood sugar control. Figure 1 As shown, in another embodiment, the drug infusion system may further include: a blood glucose measuring device 300. The electronic device 100 and the blood glucose measuring device 300 are connected in communication, and optionally, the blood glucose measuring device 300 may also be connected in communication with the drug infusion device 200, wherein:

[0054] The blood glucose measuring device 300 performs blood glucose monitoring on the target object to obtain the target object's blood glucose measurement value. The electronic device 100 obtains the target object's behavior information in the current scene; obtains the target object's blood glucose measurement value from the blood glucose measuring device 300; determines the reference dosage of the drug used to adjust the target object's blood glucose value based on the target object's behavior information; obtains the target object's drug dosage recommendation based on the drug reference dosage and the blood glucose measurement value; and then sends the drug dosage recommendation to the drug infusion device. The drug infusion device 200 infuses the drug to the target object based on the drug dosage recommendation.

[0055] It is worth noting that in the embodiment of the present application, when the electronic device 100 itself has a blood sugar monitoring function, Figure 1 The blood sugar measuring device 300 shown may be omitted.

[0056] In the drug infusion system in the embodiment of the present application, the electronic device obtains the target object's blood glucose measurement value through the blood glucose measuring device 300, and determines the reference dosage of the drug used to adjust the target object's blood glucose value based on the target object's behavior information in the current scene, and finally obtains the target object's drug dosage recommendation by combining the reference dosage of the drug and the blood glucose measurement value; the drug infusion device 200 infuses the drug to the target object according to the drug dosage recommendation. In the system of this embodiment, the drug dosage recommendation based on which the drug infusion device performs drug infusion is associated with the target object's behavior information in the current scene, so that the drug infusion fully considers the impact of the current scene where the target object is located on the blood glucose value of the diabetic patient, and ensures the effect of blood glucose control; and because the drug dosage recommendation obtained in the system also considers the target object's blood glucose measurement value, the safety of the drug infusion system is improved.

[0057] It should be understood that the electronic device 100 in the embodiment of the present application can be a device independent of the drug infusion device 200 and the blood glucose measuring device 300, or the electronic device 100 can also be integrated on the drug infusion device 200, or the electronic device 100 can also be integrated on the blood glucose measuring device 300, and the present application does not impose any restrictions on this.

[0058] For example, the electronic device 100 as an independent device can be a wearable device, a mobile phone, a tablet computer, a computing device, etc. The wearable devices include but are not limited to: watches, wristbands, glasses, headbands, shoes, socks, patch-type wearable devices, smart clothing, accessories, etc., and this application does not impose any restrictions on this.

[0059] In some embodiments, the electronic device 100, the drug infusion device 200 and the blood glucose measuring device 300 in the drug infusion system may also be integrated into one device, as long as the functions of the electronic device 100, the drug infusion device 200 and the blood glucose measuring device 300 can be realized, which will not be elaborated here.

[0060] In some embodiments, the drug infusion device 200 in the embodiments of the present application may be an insulin pump worn on the body of the target object, and the insulin pump stores drugs that can regulate the blood sugar of the target object, for example, the drugs may include insulin for lowering blood sugar, and of course in some embodiments, the drugs in the insulin pump may also include glucagon for raising blood sugar. The present application does not limit the specific drugs stored in the drug infusion device, as long as the drugs are clinically usable blood sugar regulating drugs.

[0061] It is understandable that the drug infusion device 200 in the embodiment of the present application may also be other devices that can infuse drugs into a target object, and this application will not elaborate or enumerate on this.

[0062] In some embodiments, the blood glucose measuring device 300 in the embodiment of the present application can be a continuous blood glucose monitoring device (CGM, Continuous Glucose Monitoring), which implants a glucose sensor into the body of the target object, thereby realizing real-time monitoring of the glucose concentration in the subcutaneous tissue fluid of the target object.

[0063] It can be understood that the electronic device 100 obtains the blood glucose measurement value of the target object in the current scene through the blood glucose measuring device 300. Specifically, the blood glucose measuring device 300 sends the blood glucose measurement value to the electronic device 100, or the electronic device 100 obtains the blood glucose measurement value from the blood glucose monitoring device 300, or the blood glucose monitoring device 300 sends the blood glucose measurement value to the transfer device, and the electronic device 100 obtains the blood glucose measurement value from the transfer device.

[0064] In some embodiments, the electronic device 100 obtains the behavior information of the target object in the current scene. Specifically, it can be: collecting relevant sensor data of the current scene where the target object is located through a sensor, and then identifying the sensor data to obtain the behavior information of the target object in the current scene.

[0065] In some embodiments, the sensor used to collect relevant sensor data of the current scene where the target object is located can be set independently of the electronic device 100, the drug infusion device 200 or the blood glucose measurement device 300. For example, all relevant sensors can be integrated into a scene data collection device 400. The scene data collection device 400 can be used as a part of the drug infusion system or as an external device of the drug infusion system to provide sensor data to the electronic device 100. In both cases, the scene data collection device 400 can be connected to the electronic device 100 for communication.

[0066] For example, the scene data acquisition device 400 can be a wearable device, a mobile phone, etc. equipped with a sensor. The wearable device includes but is not limited to: watches, wristbands, glasses, headbands, shoes, socks, patch-type wearable devices, smart clothing, accessories, etc., and this application does not impose any restrictions on this.

[0067] In some other embodiments, the sensor for collecting relevant sensor data of the current scene where the target object is located can be integrated in the electronic device 100, the drug infusion device 200 and / or the blood glucose measuring device 300. For example, when there is only one sensor, the sensor can be set in the electronic device 100, the drug infusion device 200 or the blood glucose measuring device 300; when there are multiple sensors, the multiple sensors can be set in any one or more of the electronic device 100, the drug infusion device 200 or the blood glucose measuring device 300, that is, the multiple sensors can be set on the same device or separately on multiple devices; the embodiment of the present application does not specifically limit the specific setting method of the sensor.

[0068] For example, when the sensor for collecting relevant sensor data of the current scene where the target object is located is set in the electronic device 100, the electronic device 100 can be a wearable device, a mobile phone, etc. equipped with a sensor. The wearable device includes but is not limited to: watches, wristbands, glasses, headbands, shoes, socks, patch-type wearable devices, smart clothing, accessories, etc., and this application does not impose any restrictions on this.

[0069] In some embodiments, the sensor used to collect relevant sensor data of the current scene where the target object is located may include one or more of a sound sensor, an inertial sensor, and a navigation and positioning sensor, wherein the sound sensor can be used to collect sound data in the current scene where the target object is located; the inertial sensor can be used to collect inertial sensor data, for example, the inertial sensor data may include acceleration information and / or speed information of the target object; the navigation and positioning sensor can be used to collect navigation and positioning sensor data, for example, the navigation and positioning sensor may include the positioning information and signal strength value of the target object at the current moment.

[0070] It is understandable that the communication connection in the embodiment of the present application may be a wireless communication connection, for example, the wireless communication connection may be implemented through wireless communication technologies such as Bluetooth (BT) technology, wireless-fidelity (WiFi) technology or near field communication (NFC) technology. Of course, the communication connection in the embodiment of the present application may also be a wired communication connection, and the present application does not limit or elaborate on this.

[0071] After introducing the drug infusion system, Figure 1 The drug infusion system shown is an exemplary illustration of a drug dosage recommendation method based on scene recognition based on an electronic device in an embodiment of the present application.

[0072] Figure 2 The flowchart of the method for recommending drug dosage based on scene recognition provided in one embodiment of the present application is shown in FIG. The method for recommending drug dosage based on scene recognition is applied to the electronic device 100 in the above-mentioned drug infusion system embodiment. The method for recommending drug dosage based on scene recognition includes: S201 to S203, and each step is described below.

[0073] S201, obtaining behavior information of a target object in a current scene.

[0074] It should be understood that the target subject is a subject who needs drug infusion, for example, any diabetic patient who needs to regulate blood sugar by drug infusion. The diabetes suffered by the diabetic patient may be type 1 diabetes, type 2 diabetes, other specific diabetes or gestational diabetes, and this application does not limit this.

[0075] It should also be understood that the current scene refers to the environment or scene that the target object is in at the current moment, or a period of time up to the current moment. The behavior information represents the behavior and state of the user target object. Different behaviors and states will have different effects on the target object's blood sugar level.

[0076] In some embodiments, the behavior information of the target object can also be determined by image recognition. For example, an image of the target object in the current scene can be obtained, and the image can be recognized by a pre-trained behavior recognition model to obtain the behavior information of the target object. The behavior recognition model can be a common deep learning model for image processing, such as RCNN (Region-CNN, candidate region-convolutional neural network), Fast R-CNN (Fast Region-CNN, fast-candidate region-convolutional neural network), Faster R-CNN (Faster Region-CNN, faster-candidate region-convolutional neural network) or YOLO (you only look once, target detection model); the training method of the behavior recognition model can adopt a conventional deep learning model training method, for example, the training process can include: loading a training data set, building a network model, setting training parameters, starting training, model saving and evaluation, etc., which will not be elaborated in this application.

[0077] In some other embodiments, the specific process of obtaining behavior information includes: obtaining sensor data of the target object in the current scene, the sensor data including data obtained by using a sensor to collect signals of the current scene where the target object is located; and determining the behavior information of the target object using the sensor data.

[0078] Exemplarily, the sensor data includes one or more of sound scene data, inertial sensor data, or navigation and positioning sensor data, wherein the sound scene data is obtained by a sound sensor.

[0079] In some embodiments, the specific process of obtaining behavior information includes: obtaining sound scene data, the sound scene data includes sound data in the current scene; performing foreground-background sound separation on the sound scene data to obtain background sound; and determining the behavior information of the target object based on the background sound. The sound scene data collected by the sound sensor usually includes both ambient sound and voice signals. Sometimes the voice signal may dominate, and the accuracy of directly using the sound scene data to identify the behavior information may not be high enough. Therefore, the foreground-background sound separation technology can be used to first separate the background sound from the sound scene data, and then the behavior information can be identified for the background sound, which can improve the accuracy of behavior information identification.

[0080] It is understandable that the sound scene data can be obtained by collecting signals from the current scene where the target object is located through a sound sensor. The sound sensor can be a microphone. The specific setting method of the sound sensor can refer to the description in the drug infusion system embodiment and will not be repeated here.

[0081] Exemplarily, the method for foreground-background sound separation can be a probabilistic method based on a hidden Markov model (HMM), or a deep learning method based on convolutional neural networks (CNN) or recurrent neural networks (RNN), etc., and the present application does not impose any restrictions on this.

[0082] In some embodiments, determining the behavior information of the target object according to the background sound includes: obtaining voiceprint features of the background sound; inputting the voiceprint features into a behavior recognition model to obtain behavior information, wherein the behavior information includes: multiple behavior state categories of the target object and the probability of each behavior state category, and each behavior state category represents a type of behavior state of the target object. After extracting the voiceprint features of the background sound, the voiceprint features are recognized by the behavior recognition model to obtain the behavior information, so that the voiceprint features input into the behavior recognition model can well reflect the characteristics of the background sound, thereby improving the accuracy of the behavior information output by the behavior recognition model.

[0083] It should be understood that obtaining the voiceprint features of the background sound specifically involves extracting the voiceprint features of the background sound. The specific voiceprint feature extraction method may be an i-vector or x-vector method, etc., which will not be elaborated in this application.

[0084] For example, the behavior recognition model can be a machine learning model, such as support vector machine (SVM), decision tree, random forest or K-Nearest Neighbors, etc. The machine learning model can be trained by using preprocessed audio data and feature vectors, and the trained model can be evaluated and optimized to finally obtain a usable behavior recognition model, which is not described in detail in this application.

[0085] For another example, the behavior recognition model can also be a deep learning model, such as: convolutional neural network (CNN), recurrent neural network (RNN), LSTM (long short-term memory network), gated recurrent unit (GRU), Transformer, etc. The deep learning model can be trained and optimized by using the original waveform of the audio data or the preprocessed features as input, and finally a usable behavior recognition model can be obtained, which is not described in detail in this application.

[0086] In some other embodiments, determining the behavior information of the target object based on the background sound may include: inputting the background sound into a first behavior recognition model to obtain behavior information, the behavior information including: multiple behavior state categories of the target object and the probability of each behavior state category, each behavior state category characterizing a type of behavior state of the target object, wherein the first behavior recognition model is preferably a deep learning model, such as a convolutional neural network (CNN), a recurrent neural network (RNN), an LSTM (long short-term memory network), a gated recurrent unit (GRU), a Transformer, etc., which will not be elaborated in this application. In this embodiment, there is no need to extract the voiceprint features of the background sound, and the background sound is directly input into the first behavior recognition model to obtain the behavior information, and the data processing process is simpler and more convenient.

[0087] In some other embodiments, determining the behavior information of the target object according to the background sound includes: performing a spectrum analysis on the background sound to obtain a spectrum analysis result; comparing the spectrum analysis result with a plurality of preset spectrum information, and determining the target spectrum information closest to the spectrum analysis result from the plurality of spectrum information; determining the behavior information corresponding to the target spectrum information as the behavior information of the target object in the current scene, and the behavior information includes a behavior state category. For example, the behavior state category may be any one of: "sleep", "work", "low-intensity exercise", "high-intensity exercise" or "meal".

[0088] In some embodiments, after determining the behavior information of the target object based on the background sound, it is also necessary to update the probability of each behavior state category in the behavior information through other sensor data. For example, the other sensor data may include inertial sensor data or navigation positioning sensor data. Considering the complexity of the environment, by using other sensor data to update the probability of each behavior state category in the behavior state information, the accuracy and stability of behavior information recognition can be improved.

[0089] In some embodiments, obtaining the behavior information of the target object in the current scene further includes: updating the probability of each behavior state category according to the navigation positioning sensor data; and performing a second update on the updated probability of each behavior state category according to the inertial sensor data. The accuracy of the probability of each behavior state category is further improved through the two updates.

[0090] In some other embodiments, obtaining the behavior information of the target object in the current scene also includes: obtaining the inertial sensor data and navigation positioning sensor data of the target object in the current scene; updating the probability of each behavior state category according to the inertial sensor data; and updating the probability of each behavior state category after the update for a second time according to the navigation positioning sensor data. In this embodiment, the accuracy of the probability of each behavior state category is further improved by two updates; and because the signal strength of the inertial sensor data is generally strong, and the signal strength of the navigation positioning sensor data may be different due to the different positions of the target object, the update result of the inertial sensor with stable signal strength is more reliable than the navigation positioning sensor data with uncertain signal strength. Therefore, this embodiment first updates the inertial sensor data with guaranteed signal strength, and then updates the navigation positioning sensor data with uncertain signal strength, which can improve the accuracy of the update result.

[0091] S202: Determine a reference dosage of a drug for regulating the blood sugar level of the target subject according to the behavior information of the target subject.

[0092] It should be understood that after determining the behavior information of the target object, the reference dosage of the medicine is determined according to the behavior information, so that the reference dosage of the medicine is associated with the behavior information of the user, thereby improving the accuracy of the reference dosage of the medicine.

[0093] In some embodiments, a fixed reference dosage of medicine can be determined for each behavior state category, and different behavior state categories correspond to different reference dosages of medicine. After determining the behavior information of the target object, if the behavior information includes a behavior state category, the reference dosage of medicine corresponding to the behavior state category is determined as the reference dosage of medicine for the target object; if the behavior information includes multiple behavior state categories and the probability of each behavior state category, the reference dosage of medicine for the behavior state category corresponding to the maximum probability can be determined as the reference dosage of medicine for the target object, or the reference dosage of medicine for each behavior state category can be multiplied by the probability value corresponding to the behavior state category, and the multiple products obtained are summed to obtain the reference dosage of medicine for the target object. In this embodiment, by determining a fixed reference dosage of medicine for each behavior state category, the reference dosage of medicine exclusively for the target object is determined according to the behavior information of the target object. The method is simple and easy to implement.

[0094] In some other embodiments, when the behavior information includes multiple behavior state categories and the probability of each behavior state category, the reference dosage of the drug is determined by determining the preset coefficient corresponding to each behavior state category in the multiple behavior state categories; and according to the preset coefficient corresponding to each behavior state category and the probability of each behavior state category. By determining the reference dosage of the drug by the preset coefficient and the probability of each behavior state category, each behavior state category is fully considered, thereby improving the accuracy of the determined reference dosage of the drug.

[0095] In some embodiments, the blood sugar regulating drug includes insulin, the preset coefficient includes a basal rate coefficient, and the reference dosage of the drug is determined according to the preset coefficient corresponding to each behavior state category and the probability of each behavior state category, including: determining the basal rate corresponding to each behavior state category according to the basal rate coefficient corresponding to each behavior state category and the preset basal rate; multiplying the basal rate corresponding to each behavior state category, the probability corresponding to each behavior state category, and the preset time interval to obtain the insulin infusion amount corresponding to each behavior state category; summing up multiple insulin infusion amounts to obtain a theoretical insulin infusion amount; if the theoretical insulin infusion amount is greater than or equal to 0, determining the theoretical insulin infusion amount as the reference dosage of the drug; if the theoretical insulin infusion amount is less than 0, determining the reference dosage of the drug as 0. The preset basal rate may be a standard basal rate. In this embodiment, a basal rate is determined for each behavior state category by using the basal rate coefficient corresponding to each behavior state category and the preset basal rate, so that the insulin infusion amount is determined by the basal rate of each behavior state category and the probability corresponding to each behavior category, and multiple insulin infusion amounts are summed to obtain the theoretical insulin infusion amount; when the theoretical insulin infusion amount is greater than 0, it means that the target subject needs to inject insulin to lower blood sugar; when the theoretical insulin infusion amount is equal to 0, it means that the blood sugar level of the target subject is in a normal state and does not need to be adjusted, so the insulin injection amount is 0; when the theoretical insulin infusion amount is less than 0, it means that the target subject has low blood sugar, and insulin cannot be injected at this time, and the drug reference amount is determined to be 0. In the case where the blood sugar regulating drug includes insulin, different drug reference amounts are determined according to the positive and negative of the theoretical insulin injection amount to ensure that the drug reference amount is a positive number greater than or equal to 0, avoid the situation where the drug reference amount is less than 0 and cannot be achieved, and improve the feasibility of the solution.

[0096] The following is an example of the process of determining the reference dosage of a drug, taking the case where the behavior information includes n behavior state categories of the target object and the probability of each behavior state category, and the drug for regulating blood sugar is insulin. As shown in Table 1, it is assumed that the target object has multiple behavior categories including n, namely category 1, category 2, ..., category n, and the probability corresponding to each category is P1, P2, ..., P n, as shown in Table 1, the probability value represents the probability that the target object belongs to the corresponding behavior state category in the current scene (or the current time t).

[0097] Table 1

[0098] Behavioral state categories Category 1 Category 2 Category 3 …… Category Probability <![CDATA[P1]]> <![CDATA[P2]]> <![CDATA[P3]]> …… <![CDATA[P n ]]>

[0099] In some embodiments, the drug may include insulin, and determining the reference dosage of the drug includes the following steps (a) to (c):

[0100] Step (a), set a standard base rate.

[0101] It is understandable that the standard basal rate can also be called the baseline basal rate, that is, the amount of insulin injected per hour. You can refer to the basal rate setting method in the "Guidelines for Insulin Pump Therapy in China (2021 Edition)" to set it, including 1-stage method or multi-stage method setting. For example, for adult patients with T1DM (type 1 diabetes), according to the 1-stage method, the standard basal rate is determined according to formula (1):

[0102]

[0103] In formula (1): Basal std represents the standard basal rate, W represents the patient's weight, α and β are two coefficients, where: β∈[0.4,0.5], α∈[0.4,0.5].

[0104] It should be understood that the unit of body weight is kg, i.e. kilogram; the unit of standard basal rate is IU / h, i.e. unit / hour, for example 3ml:300IU of insulin, each milliliter is 100IU, and 10 units are equivalent to 0.1 ml.

[0105] It is understandable that the setting of the standard basis may also adopt any possible method in the prior art, which is not enumerated in this application.

[0106] Step (b), determining the base rate of each behavior state category in the multiple behavior state categories.

[0107] Exemplarily, the base rate of the i-th behavior state category among the n behavior state categories at the current time t is determined according to formula (2):

[0108] B i (t) = Basal std ×a i (t) (2)

[0109] In formula (2): B i (t) represents the basic rate of the i-th behavior state category, Basal std represents the standard base rate, ai (t) represents the base rate coefficient of the ith behavior state category.

[0110] It should be understood that the calculation method of the basal rate coefficient of different behavior state categories is different. For ease of understanding, the following uses multiple behavior state categories including "sleep", "work", "low-intensity exercise", "high-intensity exercise", "meal", and "other" as examples to illustrate the calculation method of the basal rate coefficient.

[0111] Among them, the behavior state category of "other" can be used to handle various abnormal situations that do not belong to the first five behavior state categories, thereby ensuring that no matter what scene the target object is in, the behavior recognition model will have the recognition results of the behavior state category and the corresponding probability output. The basic rate coefficient of "other" is set to 1. The probabilities and basic rate coefficients of the six behavior state categories are shown in Table 2.

[0112] Table 2

[0113] Behavioral state categories Sleep Office Low-intensity exercise High-intensity exercise meal other Probability <![CDATA[P1]]> <![CDATA[P2]]> <![CDATA[P3]]> <![CDATA[P4]]> <![CDATA[P5]]> <![CDATA[P6]]> Base rate coefficient <![CDATA[a1]]> <![CDATA[a2]]> <![CDATA[a3]]> <![CDATA[a4]]> <![CDATA[a5]]> 1

[0114] The basic rate coefficients of different behavior status categories in Table 2 can be determined according to formula (3):

[0115]

[0116] In formula (3), C1 and C2 are constants, and 0<C1≤1, 0<C2≤1; Rate carb Indicates the carbon water absorption rate: Rate carb It is defined as the ratio of total carbohydrate intake to meal duration, that is: ICR stands for Carbohydrate Sensitivity Ratio, which represents the number of grams of carbohydrates per unit of insulin in the target subject; Basal std Indicates the standard base rate; Index exercise Indicates the intensity of exercise.

[0117] It should be understood that C1 and C2 can be set based on experience, for example, they can be set with reference to the Chinese Insulin Pump Therapy Guidelines (2021 Edition). For example, a1=C1=0.6, a2=C2=1.

[0118] Exemplarily, the motion intensity can be determined based on the data collected by the inertial sensor. For example, the motion intensity can be equal to the ratio of the number of steps counted in a preset time period starting or ending at the current moment to the maximum number of steps in the preset time period. The maximum number of steps in the preset time period is generally predefined. The preset time period refers to a preset time length, that is, the ratio of the number of steps of the target object in the time length corresponding to the current moment to the maximum number of steps in the preset time length is used as the motion intensity of the target object in the current scene.

[0119] In some other embodiments, the exercise intensity can also be calculated based on parameters such as heart rate, oxygen uptake, energy metabolism rate, etc., and this application does not make specific limitations on this.

[0120] In some other embodiments, the basal rate coefficient of the "other" behavior state category can also be obtained according to the blood glucose measurement value algorithm. For example, the basal rate coefficient of the "other" behavior state category can be determined according to formula (4):

[0121]

[0122] In formula (4): function Represents the n-dimensional historical blood sugar vector The calculated insulin injection amount at the current time t; Basal std Indicates the standard basal rate.

[0123] Exemplarily, given the independent variable blood glucose measurement value G and the dependent variable insulin injection amount I at each time t, a corresponding functional relationship I=f(G, t) can be established.

[0124] For example, the function It can also be a function based on proportional-integral-derivative (PID). Specifically, the function It can be determined according to equations (5) to (7):

[0125]

[0126] e(t)=bg(t)-bg(t) Target (6)

[0127] e(t-1)=bg(t-1)-bg(t-1) Target (7)

[0128] Where: Kp represents the proportional coefficient, Kd represents the differential coefficient, bg(t) represents the blood glucose measurement value at time t, bg(t) Targetrepresents the target blood glucose value at time t, bg(t-1) represents the blood glucose measurement value at time t-1, and bg(t-1) Target Indicates the target blood sugar level at time t-1.

[0129] It is understandable that the function It can also be a function based on model predictive control (MPC) or a function based on neural network training, which will not be elaborated in this application.

[0130] Step (c), determining a reference dosage of the drug according to a preset coefficient corresponding to each behavioral state category and the probability of each behavioral state category.

[0131] In some embodiments, the reference dosage of the drug can be determined according to formula (8) and formula (9), as follows:

[0132] First, according to formula (8), the reference dosage of the target object at the current moment is determined:

[0133]

[0134] In formula (8): M e (t) represents the theoretical insulin infusion volume, B i (t) represents the base rate of the ith behavior state category, p i (t) represents the probability corresponding to the i-th behavior state category, and ΔT represents the preset time interval.

[0135] It is understandable that the preset time interval ΔT may be the time interval between two adjacent insulin injections, for example, 5 minutes; P1+P2+…+P n =1, that is The sum of the probabilities of the n behavior state categories is equal to 1.

[0136] Then, the reference dosage of the drug is determined according to the theoretical insulin infusion amount. Specifically, the reference dosage of the drug can be determined according to formula (9):

[0137] M(t)=max(0,M e (t)) (9)

[0138] In formula (9): M(t) represents the reference dosage of the drug, M e (t) represents the theoretical insulin infusion volume.

[0139] It can be understood that formula (9) means: if the theoretical insulin infusion amount is greater than or equal to 0, the theoretical insulin infusion amount is determined as the reference drug dosage; if the theoretical insulin infusion amount is less than 0, the reference drug dosage is determined to be 0.

[0140] In some other embodiments, drugs for regulating blood sugar may include insulin and glucagon, and the preset coefficients include basal rate coefficients. According to the preset coefficients corresponding to each behavior state category and the probability of each behavior state category, a reference drug dosage is determined, including: determining the basal rate corresponding to each behavior state category according to the basal rate coefficient corresponding to each behavior state category and a preset basal rate (for example, a standard basal rate); multiplying the basal rate corresponding to each behavior state category, the probability corresponding to each behavior state category, and the preset time interval to obtain the insulin infusion amount corresponding to each behavior state category; summing up multiple insulin infusion amounts to obtain a theoretical insulin infusion amount; if the theoretical insulin infusion amount is greater than or equal to 0, determining the theoretical insulin infusion amount as the reference drug dosage; if the theoretical insulin infusion amount is less than 0, converting the theoretical insulin infusion amount into a glucagon infusion amount corresponding to the blood sugar-raising effect, and determining the glucagon infusion amount as the reference drug dosage.

[0141] It should be understood that in this embodiment, when the calculated theoretical insulin infusion amount is less than 0, it means that the target object needs to increase blood sugar at this time, so there is no need for insulin infusion. Since the conditions for infusing glucagon are met in this embodiment, the theoretical insulin infusion amount can be converted into the injection amount of glucagon, and the reference dosage of the drug is adapted to the actual behavior state of the target object in the current scenario.

[0142] S203, obtaining a drug dosage recommendation for the target subject based on the drug reference dosage.

[0143] It is understandable that the drug dosage recommendation can also be called the recommended drug dosage. The drug dosage recommendation can be generated based on the drug reference dosage, and the drug reference dosage is determined based on the behavior information of the target object in the current scene. Therefore, the obtained drug dosage recommendation is related to the behavior information of the target object in the current scene.

[0144] In some embodiments, after determining the reference drug dosage, the reference drug dosage can be used as a drug dosage recommendation for the target object, so that the obtained drug dosage recommendation is associated with the behavior information of the target object in the current scenario, thereby improving the accuracy of the drug dosage recommendation.

[0145] In some other embodiments, the method for recommending drug dosage based on scene recognition further includes: obtaining the target subject's blood glucose measurement value; and obtaining the target subject's drug dosage recommendation based on the drug reference dosage includes: obtaining the target subject's drug dosage recommendation based on the drug reference dosage and the blood glucose measurement value. In this embodiment, the target subject's blood glucose measurement value is introduced when determining the drug dosage recommendation, so as to further ensure the safety and accuracy of the obtained drug dosage recommendation.

[0146] In one embodiment, the recommended drug dosage for the target object can be determined according to formula (10):

[0147]

[0148] In formula (10), M(t) represents the reference dosage of the drug, G(t) represents the blood glucose measurement value of the target subject at the current moment, and G LGS The blood glucose threshold that indicates the need to stop insulin infusion, M r (t) is the recommended dosage of the drug.

[0149] In this embodiment, the reference drug dosage is adjusted and set by comparing the blood glucose threshold for stopping insulin infusion with the blood glucose measurement value, and the blood glucose measurement value is integrated into the determination of the recommended drug dosage to ensure the safety of the recommended drug dosage.

[0150] In another embodiment, the recommended drug dosage for the target object can be determined according to formula (11):

[0151]

[0152] In formula (11), M(t) represents the reference dosage of the drug, G L represents the target blood sugar lower limit of insulin infusion, G(t) represents the blood sugar measurement value of the target object at the current moment, G LGS The blood glucose threshold that indicates the need to stop insulin infusion, M r (t) is the recommended dosage of the drug.

[0153] In this embodiment, the reference drug dosage is adjusted and set by using the blood glucose threshold for stopping insulin infusion, the blood glucose measurement value, and the target blood glucose lower limit for insulin infusion, and the blood glucose measurement value is integrated into the determination of the recommended drug dosage to ensure the safety of the recommended drug dosage.

[0154] For ease of understanding, the updating process of the behavior status category in the behavior information in step S202 is exemplarily described below with reference to the accompanying drawings.

[0155] See also Figure 3 In a method for recommending drug dosage based on scene recognition provided in one embodiment of the present application, a flowchart for updating the probability of each behavioral state category according to the inertial sensor data, wherein the inertial sensor data includes acceleration information, and the probability update process includes: S301 to S303, and each step is described below.

[0156] S301, estimating the number of steps of the target object in the next preset time period according to the acceleration information.

[0157] In the embodiment of the present application, the preset time period can be a time period of a fixed time span, for example, it can be 1 minute, 5 minutes or 10 minutes, etc., and the present application does not limit the specific length. The next preset time period can be a preset time period after the current time. For example, if the current time is 12:00 and the preset time period length is 10 minutes, then the next preset time period refers to the time period from 12:00 to 12:10.

[0158] It should be understood that the speed of the target object in the next preset time period can be estimated based on the acceleration information, and the number of steps of the target object in the next preset time period can be estimated based on the speed information.

[0159] For example, the number of steps of the target object can be obtained by multiplying the speed by the time length corresponding to the preset period and dividing it by the stride of the target object. Of course, those skilled in the art can also use other calculation methods to obtain the number of steps, which is not limited here.

[0160] S302: Determine the exercise intensity of the target object in the next preset time period according to the number of steps.

[0161] It is understandable that the number of steps is closely related to the exercise intensity, so the exercise intensity of the target object in the next preset time period can be determined by estimating the number of steps of the target object in the next preset time period.

[0162] Exemplarily, the exercise intensity can be obtained according to formula (12):

[0163]

[0164] In formula (12): I ex represents the exercise intensity of the target object in the next preset time period, steps represents the number of steps of the target object in the next preset time period, and stepsMax represents the maximum number of steps of the target object in the preset time period.

[0165] Of course, the exercise intensity can also be determined using other formulas or methods related to the number of steps, which will not be elaborated in this application.

[0166] S303: Update the probability of each behavior state category according to the exercise intensity.

[0167] In the embodiment of the present application, it is assumed that the multiple behavior state categories include n behavior state categories, n is an integer greater than 1, and each behavior state category is a motion state or a non-motion state; after determining the exercise intensity, the probability of each behavior state category can be updated according to formula (13) and formula (14):

[0168]

[0169] f(I ex )=l ex *ΔP1 (14)

[0170] In formula (13) and formula (14): P i ' represents the probability of the i-th behavior state category after updating, P i represents the probability of the i-th behavior state category before updating, ΔP1 represents a preset first threshold and 0<ΔP1<1, I ex Indicates the exercise intensity of the target object in the next preset time period.

[0171] It can be understood that i is an integer and 1≤i≤n, and the first threshold can be set based on experience, which is not elaborated in this application.

[0172] In some embodiments, the value range of ΔP1 is 0.1-0.3, that is, 0.1≤ΔP1≤0.3.

[0173] It can be understood that by classifying each behavior state as a motion state or a non-motion state, the inertial sensor is more accurate in detecting the motion state, but relatively poor in detecting the non-motion state; if the behavior state class belongs to the motion state, the probability of the behavior state class is increased by the calculation of formula (13) and formula (14); if the behavior state class belongs to the non-motion state, the probability of the behavior state class is reduced. In this way, the probability of the corresponding behavior state class is closer to the actual situation.

[0174] It should be understood that there are many other ways to update the probability of each behavior state category based on inertial sensor data, which are not enumerated here.

[0175] See also Figure 4 In the drug dosage recommendation method based on scene recognition provided in one embodiment of the present application, a flowchart is provided for updating the probability of each behavior state category according to the navigation positioning sensor data, wherein the navigation positioning sensor data includes: the positioning information and signal strength value at the current moment, and the updating process includes: S401 to S403, and each step is described below.

[0176] S401, comparing the signal strength value with a preset strength threshold.

[0177] In the embodiment of the present application, the preset strength threshold can be set based on experience. For example, the preset strength threshold can be determined based on the relationship between the accuracy of the navigation and positioning sensor data and the signal strength. This application will not elaborate on this.

[0178] S402, if the signal strength value is greater than the preset strength threshold, the probability of each behavior state category is updated according to the first comparison result and the second comparison result, the first comparison result is the comparison result obtained by comparing the positioning information with the historical positioning information, and the second comparison result is the comparison result obtained by comparing the first behavior state category with the historical behavior information.

[0179] Among them, the first behavior state category is the behavior state category corresponding to the maximum probability among multiple behavior state categories, the historical positioning information includes multiple first positioning information of the target object, and each first positioning information corresponds to a moment before the current moment; the historical behavior information includes multiple first behavior information of the target object, and each first behavior information corresponds to a moment before the current moment.

[0180] It can be understood that when the signal strength is greater than the preset threshold, it means that the accuracy of the positioning information obtained at this time is relatively high. Therefore, when updating the probability of each behavior state category, the positioning information can be taken into account and updated from the two perspectives of positioning information and behavior information. In addition, historical positioning information and historical behavior information are taken into account during the update. The historical positioning information and historical behavior information reflect the behavioral patterns of the target object. By introducing historical positioning information and historical behavior information, the accuracy of the updated probability can be improved.

[0181] Exemplarily, the multiple first positioning information includes m groups of first positioning information, each group of first positioning information includes first positioning information corresponding to multiple moments in the same day, and m is an integer greater than 1; the multiple first behavior information includes p groups of first behavior information, each group of first behavior information includes multiple first behavior information corresponding to multiple moments in the same day, and p is an integer greater than 1; according to the first comparison result and the second comparison result, updating the probability of each behavior state category includes the following steps (1) to (3):

[0182] Step (1) determines k target positioning information from m groups of first positioning information, the time of day corresponding to each target positioning information corresponds to the current time, and the probability of the target positioning information appearing in the m groups of first positioning information is greater than a preset probability threshold.

[0183] Exemplarily, the m groups of first positioning information included in the historical positioning information can be an array of positioning information, referred to as the historical positioning array, which can be represented in the form of a matrix of N rows and M columns, where each row represents positioning information at multiple times in a day, and each column represents positioning information at the same time on different days. The time corresponding to each positioning information in the historical positioning array is a historical moment before the current moment corresponding to the current scene. That is, the historical positioning array includes positioning information for M days, and the positioning information for each day includes positioning information for N times. In this case, m is the number of days, that is, m=M.

[0184] In some embodiments, each first positioning information in the positioning information and historical positioning array may include a longitude value and a latitude value corresponding to each, or may also include three-dimensional coordinate information corresponding to each.

[0185] For example, assuming that the current time is 12:00, the historical positioning array is a matrix of 24 rows and 100 columns, that is, the historical positioning array includes 100 days, 24 positioning information per day; assuming that each positioning information corresponds to the hour of the day. Then the positioning information corresponding to 12:00 is the positioning information at 12:00 every day, that is, there are a total of 100 positioning information; among the 100 positioning information, there will be a lot of identical or similar positioning information, and the identical or similar positioning information will be classified into one category, and the probability of occurrence of each type of positioning information in the 100 positioning information is determined, and the top k types of positioning information with the largest probability are determined, and the specific positioning data corresponding to each type of positioning information is a target positioning information. Alternatively, after determining the probability of occurrence of each type of positioning information in the 100 positioning information, k types of positioning information with a probability greater than a preset probability threshold are determined, and the specific positioning data corresponding to each type of positioning information is a target positioning information. This embodiment of the present application will not be elaborated on this.

[0186] Step (2) determines the target behavior information from the p groups of first behavior information, the time of the day corresponding to the target behavior information corresponds to the current time, and in the p groups of first behavior information, the probability of the behavior state category in the target behavior information is the largest.

[0187] Exemplarily, the historical behavior information includes p groups of first behavior information, which can also be an array of behavior information, referred to as the historical behavior array. When the historical behavior array corresponds to the historical positioning array, the historical behavior data can be represented in the form of a matrix with N rows and M columns, where each row represents the behavior information at multiple moments in a day, and each column represents the behavior information at the same moment on different days. The moment corresponding to each behavior information in the historical behavior array is the historical moment before the current moment corresponding to the current scene. That is, the historical behavior array includes M days of behavior information, and each day's behavior information includes N moments of behavior information. In this case, p is the number of days, that is, p=M.

[0188] In some embodiments, the historical behavior information includes p groups of first behavior information, and the historical positioning information includes m groups of first positioning information, where p = m, that is, the number of behavior information in the historical behavior information is the same as the number of positioning information in the historical positioning information, and each first positioning information corresponds to one first behavior information, and the corresponding first behavior information and first positioning information correspond to the same time of the same day.

[0189] In some embodiments, each first behavior information in the historical behavior array may respectively include a behavior state category corresponding to each of them. Each first behavior information corresponds to a behavior state category, which represents the behavior state of the target object at the corresponding moment. For example, the behavior state category may be: sleeping, eating, working, low-intensity exercise, high-intensity exercise or any other.

[0190] For example, assuming that the current time is 12:00, the historical behavior array is a matrix with 24 rows and 100 columns, that is, the historical behavior array includes 100 days, 24 historical information per day; assuming that each historical information corresponds to the hour of the day. Then the historical information corresponding to 12:00 is the historical information at 12:00 every day, that is, there are 100 historical information in total; among the 100 historical information, there will be a lot of identical historical information, and the identical historical information will be classified into one category, and the probability of occurrence of each category of information in the 100 historical information will be determined, and the category of behavior information with the largest probability will be determined, and the behavior state category corresponding to this category of behavior information is the target behavior state category Sm, and this behavior information is the target behavior information.

[0191] Step (3) If the positioning information successfully matches any of the k target positioning information, and the first behavior state category is the same as the target behavior state category, and the target behavior state category is the behavior state category in the target behavior information, then the probability of the i-th behavior state category in the n behavior state categories is updated according to formula (14):

[0192]

[0193] In formula (15), P i ′ represents the probability of the updated i-th behavior state category, P i represents the probability of the i-th behavior state category before updating, ΔP2 represents a preset second threshold, and 0<ΔP2<1.

[0194] It can be understood that i is an integer and 1≤i≤n, the second threshold can be set based on experience; the preset probability threshold can also be set based on experience, for example: the preset probability threshold can be determined based on statistical data, which is not elaborated in this application.

[0195] In some embodiments, the value range of ΔP2 is 0.1-0.3, that is, 0.1≤ΔP2≤0.3.

[0196] It is understandable that matching the positioning information with any target positioning information can be performed using a probabilistic method.

[0197] Exemplarily, the positioning information includes a first coordinate, the first target positioning information includes a second coordinate, the first target positioning information is any one of k target positioning information, and the matching process includes: first determining a distance value between the first coordinate and the second coordinate; if the distance value is less than a preset distance threshold, determining that the positioning information matches the first target positioning information successfully; if the distance value is greater than or equal to the preset distance threshold, determining that the positioning information matches the second target positioning information unsuccessfully.

[0198] It is understandable that the distance value between the first coordinate and the second coordinate can be calculated using any mature algorithm, such as: Euclidean distance (i.e., L2 norm), Manhattan distance, Minkowski distance, cosine similarity distance, dynamic time warping, etc., which are not specifically limited or enumerated here.

[0199] For example, the location information of the target object in the current scene is Lt, the k target location information are (L1, L2..., Lk), the first behavior state category is Sv, and the target behavior state category is Sm. The condition for formula (15) to be valid is: Lt matches any one of (L1, L2..., Lk), and Sv is the same as Sm.

[0200] In some embodiments, if Lt cannot match any of (L1, L2, . . . , Lk), or Sv is different from Sm, the probability of any behavior state category among the n behavior state categories is not updated.

[0201] S403: If the signal strength value is less than or equal to the preset strength threshold, the probability of each behavior state category is updated according to the second comparison result.

[0202] It is understandable that when the signal strength is less than or equal to the preset threshold, it means that the accuracy of the positioning information obtained at this time is low. Therefore, when updating the probability of each behavior state category, the positioning information can be ignored, and the probability can be updated directly through the comparison result of the behavior information and the historical behavior information, avoiding the influence of the less accurate positioning information on the update effect, thereby ensuring the accuracy of the updated probability.

[0203] Exemplarily, the multiple behavior state categories include n behavior state categories, where n is an integer greater than 1; the multiple first behavior information includes p groups of first behavior information, where p is an integer greater than 1, and each group of first behavior information includes multiple first behavior information corresponding to multiple moments in the same day; and updating the probability of each behavior state category according to the second comparison result includes the following steps (1) and (2):

[0204] Step (1), determine the target behavior information from the p groups of first behavior information, the time of the day corresponding to the target behavior information corresponds to the current time, and in the first behavior information, the probability of the behavior state category in the target behavior information is the largest.

[0205] Step (2): If the first behavior state category is the same as the target behavior state category, and the target behavior state category is the behavior state category in the target behavior information, then the probability of the i-th behavior state category among the n behavior state categories is updated according to formula (15):

[0206]

[0207] In formula (16), P i ′ represents the probability of the updated i-th behavior state category, P i represents the probability of the i-th behavior state category before updating, ΔP3 represents a preset third threshold, and 0<ΔP3<1.

[0208] It can be understood that i is an integer and 1≤i≤n, and the third threshold can be set based on experience; illustratively, ΔP3<ΔP2, for example:

[0209] In some embodiments, if the first behavior state category is different from the target behavior state category, the probability of any behavior state category among the n behavior state categories is not updated.

[0210] In some embodiments, after the probability of any one of the n behavior state categories is updated, the maximum probability Pm can be determined from the updated multiple probabilities, and the behavior state category corresponding to the maximum probability Pm is Sv'; if Pm is greater than the preset probability Pth, Sv' is determined as a historical behavior information and saved in the historical behavior array, and the positioning information is saved in the historical positioning array; if Pm is less than or equal to the preset probability Pth, the historical behavior array and the historical positioning data are not updated. In this embodiment, when the maximum probability in the updated probability is greater than the probability threshold, it means that the updated probability value is high, so the corresponding behavior state category Sv' and positioning information have a higher credibility, so the behavior state category Sv' and positioning information are saved for subsequent probability updates, so that the historical behavior array and the historical positioning array can be updated in time, so the historical behavior array and the historical positioning array can better reflect the behavior habits of the target object and ensure the accuracy of the probability update.

[0211] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0212] Corresponding to the drug dosage recommendation method based on scene recognition in the above embodiment, an embodiment of the present application also provides a drug dosage recommendation device based on scene recognition, which includes a unit for executing each step in the drug dosage recommendation method based on scene recognition in any of the above embodiments.

[0213] The following is an exemplary description of the drug dosage recommendation device based on scene recognition in the embodiment of the present application in conjunction with the accompanying drawings. For the sake of convenience of explanation, only the parts related to the embodiment of the present application are shown.

[0214] Reference Figure 5 The drug dosage recommendation device 500 based on scene recognition includes: a first acquisition unit 510, a determination unit 520, and a drug dosage recommendation unit 530, wherein:

[0215] A first acquisition unit 510 is used to acquire behavior information of a target object in a current scene;

[0216] A determination unit 520, configured to determine a reference dosage of a drug for regulating the blood sugar level of the target subject according to the behavior information of the target subject;

[0217] The drug dosage recommendation unit 530 is used to obtain a drug dosage recommendation for a target subject based on a drug reference dosage.

[0218] In some embodiments, the drug dosage recommendation device 500 based on scene recognition further includes: a second acquisition unit, wherein:

[0219] a second acquisition unit, configured to acquire a blood glucose measurement value of a target subject;

[0220] The drug dosage recommendation unit 530 is used to obtain a drug dosage recommendation for a target subject based on a drug reference dosage and a blood glucose measurement value.

[0221] It should be understood that the specific process of each unit in the scene recognition-based drug dosage recommendation device 500 executing the corresponding steps in the above-mentioned scene recognition-based drug dosage recommendation method can be referred to the description related to the scene recognition-based drug dosage recommendation method in the previous text. For the sake of brevity, it will not be repeated here.

[0222] See also Figure 6 FIG. 1 is a schematic diagram of the structure of an electronic device provided in one embodiment of the present application. Figure 6 As shown, the electronic device 6 in the embodiment of the present application includes at least one processor 600 ( Figure 6 Only one processor is shown), a memory 601, and a computer program 602 stored in the memory 601 and executable on the at least one processor 600, wherein the processor 600 implements any of the steps in the above-mentioned method for recommending drug dosage based on scene recognition when executing the computer program 602.

[0223] Figure 6 This is merely an example of the electronic device 6 and does not constitute a limitation on the electronic device 6 . The electronic device 6 may include more or fewer components than those shown in the figure, or a combination of certain components, or different components.

[0224] The processor 600 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0225] In some embodiments, the memory 601 may be an internal storage unit of the electronic device 6, such as a hard disk or memory of the electronic device 6. In other embodiments, the memory 601 may also be an external storage device of the electronic device 6, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the electronic device 6. Further, the memory 601 may also include both an internal storage unit of the electronic device 6 and an external storage device. The memory 601 is used to store an operating system, an application program, a boot loader (Boot Loader), data, and other programs, such as the program code of the computer program, etc. The memory 601 may also be used to temporarily store data that has been output or is to be output.

[0226] Those skilled in the art will understand that Figure 6 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0227] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0228] An embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the above-mentioned method embodiments can be implemented.

[0229] An embodiment of the present application provides a computer program product. When the computer program product runs on a mobile terminal, the mobile terminal can implement the steps in the above-mentioned method embodiments when executing the computer program product.

[0230] The embodiment of the present application also provides a chip located in an electronic device, the chip comprising: a processing unit and a communication unit, the processing unit, for example, may be a processor, the communication unit, for example, may be an input / output interface, a pin or a circuit, etc. The processing unit may execute computer instructions to enable the electronic device to execute any of the drug dosage recommendation methods based on scene recognition provided in the above-mentioned embodiment of the present application.

[0231] Optionally, the computer instructions are stored in a storage unit.

[0232] Optionally, the storage unit is a storage unit within the chip, such as a register, a cache, etc. The storage unit can also be a storage unit located outside the chip within the terminal, such as a ROM or other types of static storage devices that can store static information and instructions, random RAM, etc. Among them, the processor mentioned in any of the above can be a CPU, a microprocessor, an ASIC, or one or more integrated circuits for controlling the execution of the program of the above-mentioned feedback information transmission method. The processing unit and the storage unit can be decoupled and respectively set on different physical devices, and connected by wired or wireless means to implement the respective functions of the processing unit and the storage unit to support the system chip to implement the various functions in the above-mentioned embodiments. Alternatively, the processing unit and the memory can also be coupled on the same device.

[0233] Among them, the electronic device, computer-readable storage medium, computer program product or chip provided in this embodiment are all used to execute the corresponding methods provided above. Therefore, the beneficial effects that can be achieved can refer to the beneficial effects in the corresponding methods provided above and will not be repeated here.

[0234] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of the above-mentioned various method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium can at least include: any entity or device that can carry the computer program code to the projection device / terminal device, recording medium, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, RandomAccess Memory), electric carrier signal, telecommunication signal and software distribution medium. For example, a USB flash drive, a mobile hard disk, a magnetic disk or an optical disk. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electric carrier signals and telecommunication signals.

[0235] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0236] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0237] In the embodiments provided in the present application, it should be understood that the disclosed devices / network equipment and methods can be implemented in other ways. For example, the device / network equipment embodiments described above are merely schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0238] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0239] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for recommending drug dosage based on scene recognition, characterized in that: The method comprises: Get the behavior information of the target object in the current scene; Determining a reference dosage of a drug for regulating the blood sugar level of the target subject according to the behavior information of the target subject; A drug dosage recommendation for the target object is obtained based on the drug reference dosage.

2. The method according to claim 1, characterized in that The step of obtaining the behavior information of the target object in the current scene includes: Acquiring sound scene data, wherein the sound scene data includes sound data in the current scene; Performing foreground-background sound separation on the sound scene data to obtain background sound; The behavior information of the target object is determined according to the background sound.

3. The method according to claim 2, characterized in that The determining of the behavior information of the target object according to the background sound comprises: Acquiring the voiceprint feature of the background sound; The voiceprint feature is input into a behavior recognition model to obtain behavior information of the target object, wherein the behavior information includes: multiple behavior state categories of the target object and the probability of each behavior state category, and each behavior state category represents a type of behavior state of the target object.

4. The method according to claim 2, characterized in that: The step of obtaining the behavior information of the target object in the current scene further includes: Acquire inertial sensor data and navigation positioning sensor data of the target object in the current scene; updating the probability of each behavior state category according to the inertial sensor data; performing a second update on the updated probability of each behavior state category according to the navigation positioning sensor data; or, updating the probability of each behavior state category according to the navigation positioning sensor data; The updated probability of each behavior state category is updated for the second time according to the inertial sensor data.

5. The method according to claim 4, characterized in that The inertial sensor data includes acceleration information, and updating the probability of each behavior state category according to the inertial sensor data includes: estimating the number of steps taken by the target object in the next preset time period according to the acceleration information; Determining the exercise intensity of the target object in the next preset time period according to the number of steps; The probability of each behavior state category is updated according to the exercise intensity.

6. The method according to claim 5, characterized in that The exercise intensity is obtained according to the following formula: Where: I ex represents the exercise intensity of the target object in the next preset time period, steps represents the number of steps of the target object in the next preset time period, and stepsMax represents the maximum number of steps of the target object in the preset time period.

7. The method according to claim 6, characterized in that The multiple behavior state categories include n behavior state categories, where n is a positive integer greater than 1; each behavior state category is a motion state or a non-motion state, and the probability of the i-th behavior state category is updated according to the following formula: f(I ex )=I ex *ΔP1 Where: P i ' represents the probability of the i-th behavior state category after updating, P i represents the probability of the i-th behavior state category before updating, ΔP1 represents the first threshold value set in advance, and I ex Indicates the exercise intensity of the target object in the next preset time period.

8. The method according to claim 4, characterized in that The navigation and positioning sensor data includes: positioning information and signal strength value at the current moment. According to the navigation and positioning sensor data, the probability of each behavior state category is updated, including: If the signal strength value is greater than a preset strength threshold, the probability of each behavior state category is updated according to a first comparison result and a second comparison result, wherein the first comparison result is a comparison result obtained by comparing the positioning information with historical positioning information, and the second comparison result is a comparison result obtained by comparing the first behavior state category with historical behavior information; Wherein: the first behavior state category is the behavior state category corresponding to the maximum probability among the multiple behavior state categories, the historical positioning information includes multiple first positioning information of the target object, each first positioning information corresponds to a moment before the current moment; the historical behavior information includes multiple first behavior information of the target object, each first behavior information corresponds to a moment before the current moment; If the signal strength value is less than or equal to a preset strength threshold, the probability of each behavior state category is updated according to the second comparison result.

9. The method according to claim 8, characterized in that The multiple behavior state categories include n behavior state categories, where n is an integer greater than 1; the multiple first positioning information include m groups of first positioning information, each group of first positioning information includes first positioning information corresponding to multiple moments in the same day, where m is an integer greater than 1; the multiple first behavior information includes p groups of first behavior information, each group of first behavior information includes multiple first behavior information corresponding to multiple moments in the same day, where p is an integer greater than 1; The updating of the probability of each behavior state category according to the first comparison result and the second comparison result includes: Determine k target positioning information from the m groups of first positioning information, the time of day corresponding to each target positioning information corresponds to the current time, and the probability of the target positioning information appearing in the m groups of first positioning information is greater than a preset probability threshold; Determine target behavior information from the p groups of first behavior information, the time of day corresponding to the target behavior information corresponds to the current time, and in the p groups of first behavior information, the probability of the behavior state category in the target behavior information is the largest; If the positioning information successfully matches any one of the k target positioning information, and the first behavior state category is the same as the target behavior state category, and the target behavior state category is the behavior state category in the target behavior information, then the probability of the i-th behavior state category in the n behavior state categories is updated according to the following formula: Where P i ′ represents the probability of the i-th behavior state category after updating, P i represents the probability of the i-th behavior state category before updating, and ΔP2 represents a preset second threshold.

10. The method according to claim 9, characterized in that The positioning information includes a first coordinate, the first target positioning information includes a second coordinate, the first target positioning information is any one of the k target positioning information, and the method further includes: determining a distance value between the first coordinate and the second coordinate; If the distance value is less than a preset distance threshold, it is determined that the positioning information matches the first target positioning information successfully; If the distance value is greater than or equal to a preset distance threshold, it is determined that the positioning information fails to match the second target positioning information.

11. The method according to claim 8, characterized in that The multiple behavior state categories include n behavior state categories, where n is an integer greater than 1; the multiple first behavior information includes p groups of first behavior information, where p is an integer greater than 1, and each group of first behavior information includes multiple first behavior information corresponding to multiple moments in the same day; The updating of the probability of each behavior state category according to the second comparison result includes: Determine target behavior information from the p groups of first behavior information, the time of day corresponding to the target behavior information corresponds to the current time, and in the first behavior information, the probability of the behavior state category in the target behavior information is the largest; If the first behavior state category is the same as the target behavior state category, and the target behavior state category is the behavior state category in the target behavior information, then the probability of the i-th behavior state category among the n behavior state categories is updated according to the following formula: Where P i ′ represents the probability of the i-th behavior state category after updating, P i represents the probability of the i-th behavior state category before updating, and ΔP3 represents a preset third threshold.

12. The method according to any one of claims 1 to 11, characterized in that: The method further comprises: Obtaining a blood sugar measurement value of the target subject; The step of obtaining a recommended drug dosage for the target subject according to the reference drug dosage includes: A drug dosage recommendation for the target object is obtained based on the drug reference dosage and the blood glucose measurement value of the target object.

13. A drug recommendation device, characterized in that: The device comprises functional units for executing the steps of the method according to any one of claims 1 to 12.

14. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 12 are implemented.

15. A drug infusion system, characterized in that: The system comprises: the electronic device and the drug infusion device according to claim 14, wherein: The electronic device is used to obtain the behavior information of the target object in the current scene; determine the reference dosage of the drug used to adjust the blood sugar level of the target object based on the behavior information of the target object; obtain the recommended drug dosage for the target object based on the reference drug dosage, and send the recommended drug dosage to the drug infusion device, and the drug infusion device is used to infuse the drug to the target object according to the recommended drug dosage.

16. The system according to claim 15, characterized in that The system further comprises: a blood sugar measuring device, wherein: The blood glucose measuring device is used to monitor the blood glucose of a target object and obtain the blood glucose measurement value of the target object. The electronic device is also used to obtain the blood glucose measurement value from the blood glucose monitoring device and obtain a recommended drug dosage for the target object based on the drug reference dosage and the blood glucose measurement value.

17. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 12 are implemented.

Images