System for estimating an unexpected meal consumed

Abstract

System for estimating an unexpected meal consumed. A method for estimating the size of an unexpected meal consumed by a user, being applied by a system, this system comprising: - a storage module recording the user's blood glucose values, - a forecasting module determining an expected change in blood glucose at a given time, - an estimation module estimating the size of the unexpected meal consumed based on: - a difference in change between a measured change in blood glucose between a blood glucose value at a given time and a blood glucose value at a previous time, and the expected change in blood glucose between a blood glucose value at a given time and a blood glucose value at a previous time; - at least one user parameter linked to and personalized according to the user. Fig. 4

Description

Title of the invention: System for estimating an unexpected meal consumed

[0001] DOMAIN OF INVENTION

[0002] The invention relates to the field of closed-loop or open-loop systems for managing a user's blood glucose levels. In particular, the invention relates to the field of estimating the amount of food consumed by a user.

[0003] TECHNOLOGICAL CONTEXT

[0004] Closed-loop systems for managing or controlling blood glucose in people with type 1 diabetes (TID), i.e., artificial pancreas (AP) systems, always rely on prior actions, such as meal notifications, to achieve effective control or management. Open-loop systems for managing or controlling blood glucose in people with type 1 diabetes (TID), for example, smart pens combined with a continuous glucose monitoring system, also always rely on prior actions, such as meal notifications, to achieve effective control or management.

[0005] Numerous studies have indicated that a large number of meal boluses are missed, particularly in adolescents during insulin pump therapy. For example, many studies have reported a link between glycated hemoglobin (HbAlc) levels and missed meal boluses [1-3], reporting an average increase of 4 mmol / mol (0.3%) in HbAlc over a 2-week period due to missed meal boluses.

[0006] Increases in HbAlc increase the risk of long-term complications, such as retinopathies, nephropathies, neuropathies, heart disease and strokes.

[0007] In healthy subjects, hypoglycemia and hyperglycemia are countered by a physiological control system including pancreatic hormones such as insulin, glucagon, and amylin. In the case of DTI, this control system is impaired, which is particularly noticeable after a meal, with a significant postprandial increase in blood glucose. Meals are difficult for blood glucose management systems to compensate for due to the delayed onset of current rapid-acting insulin formulations and the 5- to 15-minute discrepancy inherent in the CGS measurement lag between blood glucose values ​​and interstitial glucose values ​​measured in the interstitial space. For this reason, these systems always require prior actions by the user, such as notifications of meals. But, as mentioned previously, unexpected meals can occur and have significant negative consequences.

[0008] Some systems have therefore been created to detect the occurrence of an unplanned meal. Some systems are based on optimal filtering or state observation. Others are based on the detection of anomalies in the flow of blood glucose values ​​combined with machine learning techniques. Fuzzy systems or moving horizon estimation systems have also been proposed.

[0009] However, detection represents only one step towards obtaining a fully autonomous blood glucose management system. Indeed, an estimation of the detected unplanned meal is necessary to ensure reliable blood glucose management. Current meal detection systems and methods do not estimate the size of the detected meal, or do so with an additional delay. The delay for an estimate is approximately 20 minutes.

[0010] This delay is due to the fact that the detection is based on the integration of the glycemic rise and on the estimation of going below the curve, instead of relying solely on the glycemic rise.

[0011] For some systems and methods, compensation is achieved through a PD controller, which implies further delays.

[0012] Therefore, a better method, especially a faster method, for estimating unplanned meals should be implemented. Summary of the invention

[0013] The invention relates to a method for estimating the size of an unexpected meal ingested by a user, the method being applied by a system, this system comprising:

[0014] - a storage module, this storage module recording a time series of user blood glucose values ​​stored in a blood glucose meter,

[0015] - a forecasting module, this forecasting module determining an expected variation blood glucose levels at a given moment,

[0016] - an estimation module, this estimation module estimating the size of the unplanned meal ingested based on:

[0017] - a difference in variation between

[0018] a variation in blood glucose measured between a blood glucose value at a given instant and a blood glucose value at a point in the time series preceding the given instant,

[0019] blood glucose values ​​being collected by a blood glucose acquisition system or recorded in blood glucose storage;

[0020] and

[0021] the expected change in blood glucose between a blood glucose value at a time given and a blood glucose value at a point in the time series preceding the given moment;

[0022] - at least one user parameter linked to and customized according to the user.

[0023] The size of the unplanned meal ingested is calculated at the given time To, that is, as soon as the ingestion of an unplanned meal has been detected. Indeed, the data necessary for estimating the unplanned meal are already available. This availability allows the method to be applied from time To. This method is therefore faster than known methods that require additional data after To for integration.

[0024] Thanks to this method, the estimation begins with the detection process or as soon as the detection has been confirmed.

[0025] Preferably, the forecasting module determines the expected variation in blood glucose at a given time based at a minimum on:

[0026] - the user's insulin sensitivity factor, and / or

[0027] - instantaneous consumption of insulin by the user.

[0028] Preferably, the user-related parameter linked to and customized according to the user is based on the user's weight.

[0029] Preferably, the system also includes a first estimation security module, this first estimation security module determining

[0030] - if the measurement of the difference in blood glucose variation determined by the module if the estimation exceeds a threshold value of difference of variation and setting the difference of variation to a value equal to the threshold value of difference of variation if the difference of variation determined by the estimation module exceeds the threshold value of difference of variation.

[0031] Preferably, the system also includes a second estimation safety module, this second estimation safety module determining whether the size of the unplanned meal ingested estimated by the estimation module exceeds a meal estimation threshold value and setting the size to a value equal to the meal estimation threshold value if the size of the unplanned meal ingested estimated by the estimation module exceeds the meal estimation threshold value.

[0032] The first and second safety modules of the estimation can be combined into a single safety module of the estimation which determines:

[0033] - if the measurement of the difference in blood glucose variation determined by the module if the estimation exceeds a threshold value for the difference in variation, and the difference in variation is set to a value equal to the threshold value for the difference in variation if the difference in variation determined by the estimation module exceeds the threshold value for the difference in variation; and

[0034] - if the size of the unplanned meal ingested, as estimated by the estimation module, exceeds a meal estimation threshold value and size parameter on a value equal to the value meal estimation threshold if the size of the unplanned meal ingested estimated by the estimation module exceeds the meal estimation threshold value.

[0035] The variation threshold serves as a safety measure for the user. Setting a variation threshold—that is, a maximum value beyond which the calculated variation is capped if it exceeds this maximum value—allows the artificial pancreas to calculate a maximum insulin dose for the unexpected meal based on this maximum variation threshold. This maximum insulin dose for the unexpected meal cannot be too high thanks to the variation threshold. Since the maximum insulin dose for the unexpected meal cannot be too high, the user is not at risk of experiencing hypoglycemia as a result.

[0036] This is a first action or security measure.

[0037] The meal estimation threshold has a similar objective: the maximum insulin dose for the unexpected meal cannot be too high thanks to the threshold applied to the estimated size. Since the maximum insulin dose for the unexpected meal cannot be too high, the user is not at risk of experiencing hypoglycemia as a result.

[0038] This is a second action or security measure.

[0039] Preferably, the meal estimation threshold value is calculated based on a threshold parameter determined for the given time.

[0040] Preferably, the system further comprises a correction module,

[0041] this correction module corrects the size of the unforeseen meal ingested estimated by the estimation module by applying at least a correction coefficient,

[0042] this minimum correction coefficient being predetermined according to the user's profile based at a minimum on the user's age, weight, sex and / or insulin requirements.

[0043] Preferably, the system also includes an aggressiveness module, this aggressiveness module calculating a personalized aggressiveness estimate of the meal size by applying an aggressiveness factor to the size of the unforeseen ingested meal estimated by the estimation module.

[0044] In another aspect, the invention relates to a system for estimating the size of an unexpected meal ingested by a user, which includes:

[0045] - a storage module adapted to record a time series of values ​​of user blood glucose in a blood glucose storage system - a forecasting module adapted to determine an expected variation in blood glucose at a given time,

[0046] - an estimation module adapted to estimate the size of the unplanned meal ingested while based on:

[0047] - a difference in variation between - a variation in blood glucose measured between a blood glucose value at a given moment and a blood glucose value at a point in the time series preceding that moment, blood glucose values ​​are collected by a blood glucose acquisition system or recorded in blood glucose storage;

[0048] and - the expected change in blood glucose between a blood glucose value at a given moment and a blood glucose value at a point in the time series preceding the given moment; - at least one user setting linked to and customized according to the user.

[0049] In a preferred embodiment, the system also includes a first safety module for estimation adapted for

[0050] - determine if the difference in variation determined by the estimation module exceeds a threshold value of difference of variation and set the difference of variation to a value equal to the threshold value of difference of variation if the difference of variation determined by the estimation module exceeds the threshold value of difference of variation.

[0051] In a preferred embodiment, the system also includes a second safety module for estimation adapted to determine whether the size of the unplanned meal ingested estimated by the estimation module exceeds a meal estimation threshold value and to set the size to a value equal to the meal estimation threshold value if the size of the unplanned meal ingested estimated by the estimation module exceeds the meal estimation threshold value.

[0052] In a preferred embodiment, the first and second safety modules of the estimation are combined into a single safety module of the estimation adapted to determine:

[0053] - if the measurement of the difference in blood glucose variation determined by the module If the estimation value exceeds a threshold value for the difference in variation, the difference in variation should be set to a value equal to the threshold value for the difference in variation if the difference in variation determined by the estimation module exceeds the threshold value for the difference in variation; and

[0054] - if the size of the unplanned meal ingested, as estimated by the estimation module, exceeds a meal estimation threshold value and set the size to a value equal to the meal estimation threshold value if the size of the unforeseen meal ingested estimated by the estimation module exceeds the meal estimation threshold value.

[0055] In a preferred embodiment, the meal estimation threshold value is calculated based on a threshold parameter determined for a given time.

[0056] In a preferred embodiment, the system also includes a module of a suitable correction to adjust the size of the unexpected meal ingested, as estimated by the estimation module, by applying at least a correction coefficient,

[0057] this minimum correction coefficient being predetermined according to the user's profile based at a minimum on the user's age, weight, sex and / or insulin requirements.

[0058] In a preferred embodiment, the system also includes an aggressiveness module adapted to calculate a personalized aggressiveness estimate of the meal size by applying an aggressiveness factor to the size of the unforeseen ingested meal estimated by the estimation module.

[0059] In another aspect, the invention relates to a computer program for estimating the size of an unexpected meal ingested by a user, this computer program being adapted, when executed on a processor, to ask the processor to apply the method of the invention.

[0060] The computer program is adapted, when executed on a processor, to request the processor to apply each step of the method of the invention.

[0061] BRIEF DESCRIPTION OF THE SCHEMATICS

[0062] Embodiments of the invention are described below, in relation to the following diagrams:

[0063] [Fig.1] and [Fig.2] illustrate a possible algorithmic organization (and some variations) of the estimation method.

[0064] [Fig.1] corresponds to the first part of the algorithmic organization and [Fig.2] corresponds to the second part.

[0065] [Fig.3] and [Fig.4] illustrate one possible embodiment of an artificial pancreas 1 applying an estimation method.

[0066] [Fig. 5] illustrates a possible embodiment of an artificial pancreas 1 comprising the modules required to estimate the amount of insulin needed to compensate for an unexpected meal ingested by a user.

[0067] In the diagrams, the same reference symbols indicate identical or similar objects.

[0068] DEFINITIONS

[0069] Insulin unit

[0070] In this detailed description and in accordance with the WHO Expert Committee on Biological Standardization, one international insulin unit (1 U) is defined as the "biological equivalent" of 34.7 micrograms (pg) of pure crystalline insulin. This unit is the relevant unit for discussing the amount of insulin to be injected into a user, and it cannot be converted into the International System of Units because the conversion would depend on the type of insulin used. To facilitate the presentation of the present invention, it is important that the quantities insulin are expressed in a system relevant to the invention, readers, and the scientific community.

[0071] Closed-loop or open-loop system

[0072] A closed-loop control system is a set of mechanical or electronic devices that automatically regulate a process variable to a desired state or setpoint without human intervention. Closed-loop control systems differ from open-loop control systems, which require manual intervention.

[0073] Insulin Sensitivity Factor (ISF)

[0074] An insulin sensitivity factor (ISF) or correction factor describes how much one unit of rapid-acting or regular insulin lowers blood glucose. For example:

[0075] An FSI of 1 means: 1 unit of insulin lowers blood glucose by 1 mmol / L

[0076] An FSI of 2 means: 1 unit of insulin lowers blood glucose by 2 mmol / L

[0077] An FSI of 3 means: 1 unit of insulin lowers blood glucose by 3 mmol / L

[0078] Insulin / carbohydrate ratio

[0079] The insulin / carbohydrate ratio corresponds to the number of grams of carbohydrates covered by 1 unit of rapid-acting insulin. DETAILED DESCRIPTION

[0080] General presentation

[0081] An artificial pancreas 1 is a system that automatically regulates the insulin supply of a diabetic user based on their history of blood glucose, meals and insulin.

[0082] An artificial pancreas 1 is often a model-based artificial pancreas with different modules. These modules can be computerized. From a very general point of view, four modules can be distinguished within the artificial pancreas 1:

[0083] - a data acquisition system 2,

[0084] - a data processing system 3 comprising or in communication with a a blood glucose prediction module 31 and an insulin dose calculation module (and potentially also a glucagon dose calculation module) 32. These two modules can be combined in a data processing system 3 or simply communicate with each other,

[0085] - an active system 4 adapted to deliver a dose of insulin or glucagon

[0086] - a controller system 5.

[0087] The data processing system 3 includes a clock adapted to determine the time and duration of periods.

[0088] The data acquisition system 2 is responsible for managing the various sensors worn by the user or in communication with the artificial pancreas 1. It can manage or communicate with a continuous glucose monitoring (CGM) system. The objective of the data processing system 3 is to process the various data collected by the data acquisition system 2 and to transmit a dose of insulin (or a dose of glucagon) to the active system 4.

[0089] The objective of the blood glucose prediction module 31 is to predict future blood glucose based on current blood glucose as well as on physiological variables or parameters capable of influencing the evolution of blood glucose.

[0090] The blood glucose prediction module 31 is adapted to make a prediction concerning a time value. This prediction is a prediction data value. Here, the time value is the blood glucose level.

[0091] To determine the predicted data value, the blood glucose prediction module 31 can be adapted to use a predictive model. The blood glucose prediction module 31 can be based on a physiological predictive model or on a deep learning or machine learning predictive model.

[0092] For example, the variables of the blood glucose prediction module 31 may be a current value of active insulin in the user's body, ingested carbohydrates (CHO), or meal history, and the physiological parameters used by the blood glucose prediction module 31 may be the insulin sensitivity factor (ISF) or the carbohydrate / insulin ratio. The blood glucose prediction module 31 is responsible for predicting a hypoglycemic or hyperglycemic state in the user.

[0093] This predictive model is adapted to determine or calculate a prediction P, that is, a predicted value P. This prediction is the result of the model from an algorithmic point of view. It can be called the predictive result of the model. This prediction concerns a physical characteristic of a temporal system, which varies over time, that is, a temporal parameter (or characteristic) representative of the temporal system. Here, the temporal variable or parameter is blood glucose.

[0094] Furthermore, this prediction relates to a predetermined future time.

[0095] By convention, starting from an initial instant To, a prediction P with a prior duration At is a temporal prediction Px made at the initial instant To for a future instant Tt = To + At, where At corresponds to a duration.

[0096] Here, the blood glucose prediction module 31 calculates or provides a blood glucose prediction PGly with a predetermined constant prior interval, for example, one, two, five, or ten minutes. The PGly prediction can apply to any upcoming duration AX. For example, it can correspond to the next five, ten, fifteen, or thirty minutes. It is also possible for several predictions to be made at the same initial time.

[0097] The objective of the insulin dose calculation module (and potentially also of glucagon) 32 is to determine whether a dose of insulin (or glucagon) is required based on the current blood glucose and the predicted blood glucose calculated by the blood glucose prediction module 31 to maintain or achieve a target blood glucose value or target blood glucose range and, if so, to calculate the appropriate dose of insulin (or glucagon) to maintain or achieve the target blood glucose value or target blood glucose range.

[0098] The active system 4 is responsible for administering insulin (or glucagon) based on the insulin (or glucagon) dose calculated by the insulin dose calculation module 32. The active system 4 includes an infusion system. For example, the insulin dose calculation module 32 can be located away from the active system 4 and adapted to communicate with it by any suitable means, whether wireless (radio frequency, such as Bluetooth) or wired.

[0099] The controller system 5 is responsible for the coordination and control of the different modules of the artificial pancreas 1.

[0100] In particular, the present invention relates to meal history and especially to unplanned meals ingested.

[0101] As described above, meal history is taken into account by the blood glucose prediction module 31. A recently ingested meal can also be considered, but, as previously mentioned, user input is then required. If the user does not indicate that a meal has been ingested and the amount of carbohydrates (CHO) ingested, the prediction made by the blood glucose prediction module 31 will likely be incorrect.

[0102] The failure to detect and estimate an unplanned meal results in a rise in blood glucose, which is only taken into account when detected by the CGS system. Subsequently, the insulin dose calculation module 32 calculates an insulin dose based on the rise in blood glucose due to the unplanned meal.

[0103] This process may be too slow to prevent a hyperglycemic state in the user.

[0104] It is therefore desirable to improve the performance of the artificial pancreas 1 and, more specifically, to improve the accuracy and responsiveness of the prediction model in order to better estimate insulin requirements and reduce the risk of hyperglycemia.

[0105] Solution - estimation of the unforeseen meal and compensation for that meal

[0106] Estimation method

[0107] Components of the general system applying the estimation method

[0108] The components required by the general system, i.e. the artificial pancreas 1, applying the estimation method are those which have been described previously in the general presentation:

[0109] - a data acquisition system 2;

[0110] - a data processing system 3 adapted for use with the types of detailed modules further on;

[0111] - a controller system 5.

[0112] Modules for the estimation method

[0113] Different modules can be used by the data processing system 3 of the general system, i.e. the artificial pancreas 1:

[0114] - a blood glucose acquisition module 21 adapted to receive a signal from Blood glucose (Gly) readings are sent by a blood glucose acquisition system and recorded in a 301 storage module.

[0115] - a 302 forecasting module adapted to determine a variation in blood glucose expected at a given moment by determining an expected blood glucose level and comparing it with the current blood glucose level,

[0116] - an estimation module (CHO module) 11, a correction module (module CHOcorr) 12, an aggressiveness module (CHOagr module) 14, an estimation security module (CHO sécu) 13, all these modules being adapted to perform mathematical operations with the values ​​given by the signals sent by the other modules of the artificial pancreas 1 or with the values ​​of different parameters recorded in the memory of the data processing system 3. In particular, the previous modules are adapted to perform additions, subtractions, divisions and multiplications.

[0117] These modules are computerized.

[0118] The blood glucose acquisition module 21 is adapted to receive a blood glucose signal (Gly) sent by a blood glucose acquisition system at a predetermined interval. For example, this interval may be one, two, five minutes, or any duration between one and ten minutes.

[0119] Estimation method - general proposal

[0120] It is assumed that a meal has already been detected.

[0121] The purpose of the method is to estimate the size of a meal that is either unplanned or planned without notification of the carbohydrate quantity. The suggested method is to convert a blood glucose deviation into a meal, or a carbohydrate estimate (CHO).

[0122] Preferably, the conversion takes place immediately after the detection of the discrepancy, for an instantaneous estimation.

[0123] The data processing system 3 also has in its memory a set of variables concerning the user or different temporal data fields: an FSI parameter which can be re-evaluated regularly, the history of insulin administered to the user, the user's active insulin determined by the history of insulin administered.

[0124] These variables may vary over time depending on updates to the set parameters. However, at the time the method is applied, these variables are fixed and determined.

[0125] The proposal of the invention is, at a given point in time during the blood glucose monitoring series, to compare the latest blood glucose measurements with previous blood glucose measurements and to detect a discrepancy between the latest blood glucose measurements and the expected blood glucose measurements. The proposal then uses this discrepancy to estimate the size of the unexpected meal that caused the discrepancy.

[0126] It is assumed that the storage module has recorded a time series of blood glucose measurements sent over time by the blood glucose data acquisition module 21.

[0127] Based on these blood glucose measurements, potentially a notified ingested meal, and all the different parameters concerning the user, the prediction module 302 determines an expected variation in the user's blood glucose.

[0128] This expected variation in blood glucose can be determined using different approaches.

[0129] For example, the forecasting module 302 determines the expected variation in blood glucose at a given time based at a minimum on:

[0130] - the user's insulin sensitivity factor, and

[0131] - instantaneous consumption of insulin by the user.

[0132] In another example, the change in blood glucose can be determined by calculating an expected blood glucose level at a future time and the expected difference between the expected blood glucose level and the first subsequent blood glucose measurement. Then, this difference can be compared to the difference between the current blood glucose level and the first subsequent blood glucose measurement (i.e., two consecutive blood glucose measurements).

[0133] This expected blood glucose level can be determined for any time in the future. For example, the expected blood glucose level can be determined for a future time 5, 10, 15, 20, 25, 30 minutes later, or 5 to 30 minutes later. The future time is the given time at which the method is applied.

[0134] Based on this expected blood glucose at a future time, i.e. the given time, the forecasting module 302 compares the expected blood glucose to the current blood glucose at the future time, i.e. the given time, and determines the expected variation.

[0135] By extension, the expected variation can be determined for a future time located 5, 10, 15, 20, 25, 30 minutes later, 5 to 30 minutes later.

[0136] Based on this expected variation, the proposal is to detect a discrepancy between the current blood glucose variation and the expected blood glucose variation at the given time.

[0137] Thanks to this feature, the method makes it possible to estimate the unplanned meal as soon as it is detected.

[0138] This deviation can be determined by a dedicated deviation module, the 302 forecasting module or estimation module 11.

[0139] In other words, the gap is a difference in variation between:

[0140] - a variation in blood glucose measured between a blood glucose value at time given and a blood glucose value at a point in the time series preceding the given moment, and

[0141] - the expected variation in blood glucose at the given time.

[0142] Then, based on this variation, at least one blood glucose value recorded for the time preceding the given moment and at least one user parameter linked to and customized according to the user, the estimation module 11 estimates the size of the unforeseen meal ingested.

[0143] Preferably, the user-related parameter linked to and customized according to the user is based on the user's weight as will be described more precisely in a detailed embodiment.

[0144] Based on this prediction, the blood glucose management system determines the amount of insulin to administer to the user. An overestimation can thus lead to an excessive amount of insulin, which could result in a hypoglycemic state for the user. To avoid such a dangerous state for the user, the method proposes using a first safety module of estimation 13a and a second safety module of estimation 13b, or a safety module of estimation 13 combining the first and second safety modules of estimations 13a and 13b.

[0145] The safety modules of estimation 13a and 13b are designed to prevent an overestimation of the size of the unplanned meal. To this end, the safety modules of estimation 13a and 13b monitor two thresholds:

[0146] - the first safety module of estimate 13a monitors a variation threshold applying to the difference in variation,

[0147] - the second safety module of estimation 13b monitors an estimation threshold meal allowances apply to the estimated size of the unexpected meal.

[0148] The safety modules of estimation 13a and 13b monitor the determined difference in variation and the estimated size, and define the determined values ​​of the difference in variation and the estimated size at the threshold values.

[0149] These threshold values ​​can be determined for the user.

[0150] Thus, the security module of estimate 13a determines:

[0151] - if the difference in variation determined by the estimation module 11 exceeds one threshold of variation and defines the difference in variation at that threshold in this case. Similarly, the safety module of estimate 13b determines:

[0152] - if the size of the unplanned meal ingested estimated by estimation module 11 exceeds a meal estimation threshold and it defines the size at the level of this threshold in this case.

[0153]

[0154] The security modules of estimates 13a and 13b can be combined into a single security module of estimates 13 that applies the same monitoring as the security modules of estimates 13a and 13b. It can also apply only the monitoring of the security module of estimates 13a or only the monitoring of the security module of estimates 13b.

[0155] The meal estimation threshold can be calculated based on a threshold parameter determined for a given time.

[0156] In addition to these security measures, the method can also use a correction module 12 and an aggressiveness module 14 to personalize the estimation according to the user.

[0157] The correction module 12 is configured to correct the estimated size of the unplanned ingested meal estimated by the estimation module 11 by applying a minimum correction coefficient, the minimum correction coefficient being predetermined according to a user profile based at a minimum on the user's age, weight, sex and / or insulin requirements.

[0158] This correction module 12 allows the estimation to be customized according to the user thanks to the minimum correction coefficient.

[0159] In a preferred embodiment, two correction coefficients are used by the correction module 12 to correct the estimated size with a linear correction.

[0160] The aggressiveness module 14 is configured to calculate a personalized aggressiveness estimate of the meal size by applying an aggressiveness factor to the size of the unforeseen ingested meal estimated by the estimation module 11.

[0161] The aggressiveness factor can also be customized according to and by the user.

[0162] These two personalization steps applied by the correction module 12 and the aggressiveness module 14 can be associated in all possible ways: first the correction measure then the aggressiveness measure, or vice versa.

[0163] Detailed embodiment

[0164] In this embodiment, the estimation module 11 uses a blood glucose impact module to determine the user's expected blood glucose variation.

[0165] The blood glucose impact module calculates the impact of blood glucose. The impact of blood glucose corresponds to the expected blood glucose level at a given time. It can be considered as the instantaneous expected change in blood glucose.

[0166] For example, the impact of blood glucose (BGI) is calculated with the following formula:

[0167] BGI(t) - activity _msuline(t) x FSI(t) x A /

[0168] where:

[0169] insulin_activity(t) is determined by the active insulin at time t,

[0170] FSI(t) is the user's insulin sensitivity factor [mg / dL / U] at time t, and

[0171] At is the interval between given times.

[0172] Insulin activity can be defined as the instantaneous disappearance or consumption of insulin.

[0173] If insulin activity is measured as negative, its value is equal to zero.

[0174] A model describes the gradual disappearance of insulin after a bolus injection, This pattern is typically bell-shaped (slow disappearance shortly after injection, followed by a peak in insulin absorption, then slow consumption long after injection). The variable IOB(t) represents the total amount of insulin available, and insulin activity is the negative of its derivative, the instantaneous disappearance of insulin. This insulin activity is derived from the model and depends on the timing and amount of insulin injected. When insulin is injected, the variable IOB(t) increases sharply, so the negative of its derivative is artificially negative for a short time, and its value is set to zero.

[0175] This zeroing for negative insulin activity also represents a safety step or measure, since insulin activity less than 0 has no physical meaning, so it is set to 0.

[0176] Therefore:

[0177] BGl(T^ = activity_insulme(T^ xx ^T0-(TQ - Arjj

[0178] Next, the estimation module 11 of the data processing system 3 applies the following formulas at the given time:

[0179] - a difference G(T0) between a blood glucose value at T0 from the signal of blood glucose and a blood glucose value at the time preceding To in the time series Gly(T0_Af):

[0180] G(Tq) = GlyÇT^ - Gly^ - A t)

[0181] - a difference D(T0) between the blood glucose value G(T0) and the impact value of the blood glucose at T0 BGI(T0): [0i82] = - BGI(Tq)

[0183] - an estimated value of the CHO of the unplanned meal ingested by calculating the formula next: [0!84] CHOGQ = D(TQ x RappSuc(TQ) x 100

[0185] where RappSuc(t) is a predetermined sweetening ratio [g / g / L] based on the user's weight.

[0186] The sweetening ratio depends on weight and is known from diabetes research for a portion of carbohydrates of 20 g CHO.

[0187] In this embodiment, a sugar ratio formula allows determining accurately calculate a sweetening ratio for a person weighing between 20 and 220 kg, and then use this sweetening ratio to determine the value of the CHO of the unplanned meal ingested.

[0188] The limit values ​​for the sweetener ratio are 5 and 90 for example.

[0189] Since the ingestion of the meal has already been detected as assumed, it is likely that G(T0) is positive.

[0190] However, since the detected meal was not planned, it is also likely that G(T0) is greater than BGI(T0) and therefore that D(T0) is positive.

[0191] With a positive D(T0), the CHO(T0) of the unplanned ingested meal are estimated by the formula mentioned above.

[0192] Next, as described later, the estimated size is used to manage the impending rise in blood glucose. It can be used to control the administration of insulin and / or glucagon in order to manage the user's blood glucose.

[0193] Additional modules

[0194] Some improvements will now be described. These improvements can be associated with and are applied by the following additional modules: the estimation security module 13 (13a; 13b), the correction module 12 and the aggressiveness module 14.

[0195] In one embodiment, the safety module of estimation 13a implements a safety measure related to hypoglycemia.

[0196] Indeed, this hypoglycemia safety measure serves to prevent an excessive response from the insulin dose calculation module 32 after the estimation of the unexpected ingested meal. The insulin dose calculated with a high estimated CHO amount can lead to hypoglycemia if the CHO estimate was inaccurate.

[0197] To implement this hypoglycemia safety measure, a first safety step compares the difference G(T0) to a predetermined safety threshold. This predetermined safety threshold can be named UMM_DEETA_G_MAX_SEOPE. Thus, the first security step determines if

[0198] a) G(r^ > UMM_DELTA_Gu,asLm

[0199] Or

[0200] b)G(7-J <UMM_DELTA_GUAXslorE

[0201] If condition a) is verified, then the first security step parameter G(T0) on UMM_DEETA_G_MAX_SEOPE.

[0202] In other words, the difference G(T0) is limited by a higher value which avoids an excessive insulin administration response by the insulin dose calculation module 32.

[0203] The predetermined safety threshold UMM_DELTA_G_MAX_SLOPE is preferably between two and six mg / dL / min and, ideally, set at two, three, four, five or six mg / dL / min.

[0204] The safety module of estimation 13b applies a second safety step: a maximum size measurement. According to this measurement, the estimated size is capped by a maximum size CARBSMAX. Even if the discrepancy leads to the determination of an estimated size greater than the maximum size CARBSMAX, the second safety step sets the estimated size value to CARBSMAX.

[0205] For example, during the second safety step, the safety module of estimation 13 calculates the estimate of the safety ingested meal against the CHOsécu hypoglycemia by applying the following formula:

[0206] CHOsécuÇT^ = min(CHO(T^ CARBSMAX- CHOtotal)

[0207] where

[0208] CHOtotal is the sum of the estimated CHO, CHOcorr or CHOagr since the last meal detection, and

[0209] CARBSMAX can have two predetermined values ​​UMM_MAX_EST1M_CHO_PRANDIAL and UMM_MAX_ESTIM_CHO_OUT_PRANDIAL depending on the time of day:

[0210] -CARBSMAX - UMM^MAX^ESTIM^CHO^PRANDIAL for mealtimes,

[0211] - CARBSMAX = UMM_MAX_ESTIM_CHO_OUT_PRANDIAL during other periods.

[0212] For example, meal periods correspond to the periods [11 a.m.; 2 p.m.] and [6 p.m.; 10 p.m.],

[0213] For example, the predetermined value of UMM_MAX_ESTIM_CHO_PRANDIAL is between 30 and 70 and

[0214] For example, the predetermined value of UMM_MAX_ESTIM_CHO_OUT_PRANDIAL is between 10 and 50.

[0215] The safety modules of estimates 13a and 13b can be combined into a single safety module of estimates 13 or constitute two separate modules. If the safety modules of estimates 13a and 13b are separate, it is possible to use only one of the two.

[0216] The correction module 12 aims to personalize the estimation according to the user.

[0217] To this end, the correction module 12 corrects the estimated size based on at least one correction factor.

[0218] For example, correction module 12 calculates an improved CHOcorr estimate by applying a linear correction model of the meal estimate to the CHO of the ingested meal with the following formula: [02i9] CHOcorrÇT^ = UMM_A_CORRx CH(XT^ + UMM_B_CORR

[0220] where

[0221] The UMM_A_CORR and UMM_B_CORR coefficients are predetermined according to the user's profile based at a minimum on the user's age, weight and / or sex.

[0222] For example, for adolescents of average weight, UMM_A_CORR = 2.5 and UMM_B_CORR = 1.

[0223] UMM_B_C0RR can be equal to zero. In this case, the correction module 12 uses only one correction factor.

[0224] In other words, it allows simple personalization based on the user by customizing the parameters according to a classification of the population.

[0225] In one embodiment, the estimation of CHO(T0) of the ingested meal can be parameterized or personalized according to the user (and potentially by the user) with a user-related aggressiveness factor represented by [3. This user-related aggressiveness factor must be compared to the standard aggressiveness factor [30.

[0226] For example, [3 can be customized by the user in the range [0,1 ; 2] and preferably in the range [0,3 ; 1,3].

[0227] For example, [30 is a standard aggressiveness factor whose value is in the range [0.1; 2] and is preferably equal to 0.7.

[0228] The aggressiveness module 14 calculates a personalized aggressiveness estimate CHOagr by applying an aggressiveness factor to the CHO(T0) estimated with the following formula: 102291 CHOagr CH(^ = CHOT^ x (

[0230] As mentioned previously, the embodiments can be combined. In other words, the aggressiveness module 14 can calculate a personalized aggressiveness estimate CHOagr by applying an aggressiveness factor to the improved estimate CHOcorr(To) with the following formula: 102311 CHOagrCH(^ = CHOcor^ x (

[0232] These different embodiments are illustrated by the different variations in [Fig.2] in the relationships between the different modules.

[0233] As mentioned previously, the embodiments can be combined. Therefore, the safety module of estimation 13 can apply the second safety step to the improved CHOcorr estimation obtained from the estimated CHOs, on The personalized aggressiveness estimate CHOagr is obtained from the estimated CHO values, and even from the personalized aggressiveness estimate CHOagr obtained from the improved estimate CHOcorr. These combinations correspond to the following formulas:

[0234] CHOsécu(T^ - min[CHOcorr(T^ CARBSMAX- CHOtotaï}

[0235] CHOsafe(T^ = min(CHOagr (JCARBSMAX- CHOtotaÙ

[0236] CHOsecuiT^ = min(CHOagrCHO (T CARBSMAX- CHOtotaÙ

[0237] These different embodiments are also illustrated by the different variations in [Fig.2] in the relationships between the different modules.

[0238] Second embodiment

[0239] This embodiment differs only with respect to how the expected change in blood glucose at the given time To is determined.

[0240] In the previous embodiment, the impact of blood glucose corresponds to the expected instantaneous blood glucose variation at a given time.

[0241] In this other embodiment, the expected blood glucose level is determined by the blood glucose prediction module 31.

[0242] As described previously, the blood glucose prediction module 31 can calculate a PGly AX prediction in advance with a constant interval. The reliability of this prediction can be tested over the AX duration. It is assumed that the reliability of the prediction has been evaluated for use by the data processing system 3.

[0243] In this alternative embodiment, the blood glucose prediction module 31 produces a PGly prediction every five minutes, thirty minutes in advance. Thus, at any given time To, the blood glucose prediction module 31 has already calculated a PGly(T0) prediction for time To.

[0244] This prediction PGly(T0) could have been made at any previous time as long as its reliability is evaluated.

[0245] Meanwhile, as described in the first embodiment, the blood glucose acquisition module 21 adapted to receive a blood glucose signal Gly sent by a blood glucose acquisition system receives the signal transmitted by the data acquisition system 2.

[0246] Therefore, for the same instant, the data processing system 3 has in its memory the prediction PGly made for that instant and the current blood glucose value Gly.

[0247] In the second embodiment, the difference D(T0) is calculated using the following formula:

[0248] _ PGly(T^

[0249] Given this difference between the first embodiment and the second embodiment, the CHO(T0) are calculated by the estimation module 11 by applying the following formulas

[0250] CHO(T() = D(T() x RappSuc(TQ) x 100 [025i] CHO(Tq) = (G(T0) - PGly^T^ x RappSudT^ x 100

[0252] Advantages of the estimation method

[0253] The estimation method resolves the disadvantages of the prior art mentioned above.

[0254] As a reminder, the concentration of blood glucose over time in a subject is considered a temporal system with a temporal inertia of approximately 30 minutes. In other words, there is a 30-minute delay between an insulin injection and the corresponding effect on blood glucose.

[0255] In particular, the method allows for an immediate estimation of the unplanned meal since the estimation starts as soon as the detection of an unplanned meal is confirmed.

[0256] This immediate estimate is possible by taking into account the difference D, which is based on the current blood glucose value Gly(T0) and past values.

[0257] Given that the time inertia is approximately 30 minutes, tolerating a delay for the algorithmic management of the unforeseen meal inevitably induces a rise in blood glucose which may have negative consequences for the user's health.

[0258] In addition, by immediately estimating the unplanned meal ingested, the method allows for an improvement in the prediction of future blood glucose and therefore a better calculation of the insulin dose needed to reach or maintain blood glucose in the target range or at the level of a target value.

[0259] In other words, the method allows for a virtuous algorithmic circle.

[0260] As detailed previously, the method allows the imminent rise to be managed from the start, here for example in 5 minutes, and therefore allows for smooth blood glucose management.

[0261] Compensation

[0262] Compensation represents the management of blood glucose by the artificial pancreas 1.

[0263] The compensation is applied by a system including an active system or a remote server, that is to say a server in communication with the user's system.

[0264] Components

[0265] The method and system of the invention can be used to improve blood glucose management in people with type 1 diabetes (T1D).

[0266] The components required by the general system, i.e. the artificial pancreas 1, The methods for applying compensation are those described previously in the general presentation: - a data acquisition system 2; - a data processing system 3 adapted for use with the types of modules detailed later; - a controller system 5.

[0267] However, the artificial pancreas 1 requires an additional component compared to the components used for the estimation method: an active system 4, which includes a perfusion system.

[0268] It is also possible that the artificial pancreas 1 has an interface system allowing the user to view visual outputs of the data processing system 3 or to enter notifications into the data processing system 3.

[0269] For example, the data processing system 3 can be located away from the active system 4 and adapted to communicate with it by any suitable means, whether wireless (radio frequency, such as Bluetooth) or wired.

[0270] The method for estimating the size of an unexpected meal can be applied regularly by the blood glucose management system. This estimation is preferably performed every five minutes.

[0271] Next, the estimated size, whether or not it is enhanced by the additional modules detailed above, is used to manage the user's blood glucose.

[0272] Based on the estimated size of the unplanned meal, the blood glucose management system determines or estimates an insulin bolus to counteract the expected rise in blood glucose due to the unplanned meal.

[0273] System modules

[0274] The compensation can be applied by a system comprising different modules used by the data processing system 3. The system can be a terminal or a remote server in communication with a terminal.

[0275] These modules are as follows:

[0276] - a 101 detection module adapted or configured to detect an unexpected meal as soon as the departure,

[0277] - an estimation module 102 adapted or configured to estimate meal size unexpected event detected by detection module 101,

[0278] - a compensation module 103 adapted or configured to calculate the quantity of insulin to compensate for the detected and estimated unexpected meal.

[0279] The 101 detection module can detect an unplanned meal by applying any method.

[0280] In one embodiment, detection is based on a difference between the expected change in blood glucose and the actual change in blood glucose. The difference in change comparing a change in blood glucose measured between a blood glucose value at the given time and a blood glucose value at a time in the time series prior to the given time used by the estimation module 102 involved in the estimation method can be used by the detection module 101 involved in compensation.

[0281] The estimation module 102 can be one of the following modules already presented: an estimation module (CHO module) 11, a correction module (CHOcorr module) 12, an aggressiveness module (CHOagr module) 14, an estimation security module (CHO sécu) 13.

[0282] As described previously, all these modules are adapted to estimate the size of the unplanned meal based on:

[0283] - a difference in variation between:

[0284] - a variation in blood glucose measured between a blood glucose value at time given and a blood glucose value at a point in the time series preceding the given moment, and

[0285] - the expected variation in blood glucose at the given time.

[0286] and

[0287] - at least one user setting linked to and customized according to the user.

[0288] The compensation module 103 is adapted or configured to calculate or estimate an amount of insulin based on the estimated size of the unforeseen meal.

[0289] Any method can be used to estimate or calculate the amount of insulin needed to lower blood glucose after it has risen due to an unexpected meal or, even better, to anticipate the rise in blood glucose due to an unexpected meal.

[0290] In particular embodiments, the system may also include a meal size safety module 23 adapted to determine whether the estimated size of the unexpected meal exceeds a meal estimation threshold value and to set the estimated size to a value equal to the meal estimation threshold value if the estimated size of the unexpected meal exceeds the meal estimation threshold value. This meal size safety module 23 may be the estimation safety module 13.

[0291] In another particular embodiment, the system further includes a condition module 15 adapted to monitor compensation, and in particular the calculation step. The condition module 15 is adapted to check whether at least one condition is met, one of the minimum conditions being: an IOB(t) variable, representing the variation over time of the user's active insulin quantity, is less than a predetermined value and / or a total quantity of insulin calculated over a predetermined past period is less than a predetermined value.

[0292] If this minimum condition is met, then the calculation step is applied by the compensation module 103.

[0293] In a particular embodiment, the quantity of insulin calculated by the compensation module 103 may comprise one or more of the following components:

[0294] - a residual component based on at least one previous estimated size of a meal unforeseen previous event and a predetermined ingestion profile linked to and personalized according to the user.

[0295] - a meal component calculated based on a predetermined meal ratio and size estimated cost of the unexpected meal.

[0296] - a net IOB component calculated based on the user's net active insulin.

[0297] The meal component is calculated based on a predetermined meal ratio and the estimated size of the unexpected meal. For example, the following formula can be used:

[0298] Meal component = estimated meal size x meal ratio

[0299] The meal component is intended to compensate only for the unplanned meal.

[0300] The residual component is intended to compensate for a residual part of a previous unforeseen meal that would not have been fully compensated by a previous quantity of insulin calculated by the compensation module 103.

[0301] First, the compensation module 103 must assess the extent to which the estimated last unplanned meal ingested has diffused into the user's bloodstream since the last estimate. For this purpose, a user-specific diffusion curve is used, linking a quantity (CHO) of an ingested meal to the corresponding diffusion time in the blood. According to the diffusion curve, the entirety of an ingested meal is diffused into the bloodstream after a user-specific time D.

[0302] The diffusion curve can be customized according to the user based on calibration measurements such as the user-specific duration D, but it can also be a user-type diffusion curve chosen from a number of user types or categories.

[0303] Based on the diffusion curve, the compensation module 103 can determine that a residual portion of the estimated last unplanned meal was not fully compensated by the last calculated amount of insulin. For example, the following formula can be used:

[0304] Residual meal (t) = Initial estimated meal size (t - At) - Diffusion size (At)

[0305] From the start of the estimation of the unexpected meal size, all residual meal quantities for the user-specific duration D are calculated based on the diffusion curve with a different dt for each residual meal. Then, based on all the estimated residual quantities, the compensation module 103 estimates the residual component to compensate for the sum of the estimated residual quantities.

[0306] The start of the estimation of the size of the unplanned meal slides with a time interval dt.

[0307] The net IOB component calculated by the compensation module 103 reflects the previously injected insulin that is still active in the body and must be taken into account to avoid any potential insulin overdose. It is related to the baseline active insulin corresponding to the user's reference basal insulin.

[0308] A positive value for the net IOB implies active insulin above the baseline, while a negative value implies active insulin below the baseline. This value of the net IOB component is subtracted from the amount of insulin calculated by the compensation module 103 only in the case of a positive or zero value, in order to allow the compensation module 103 to reduce the calculated amount of insulin.

[0309] The net IOB component is therefore a zero or negative component of the quantity of insulin calculated by the compensation module 103.

[0310] Furthermore, in the preceding particular embodiment, the quantity of insulin calculated by the compensation module 103 may also include another component defined as a blood glucose component.

[0311] The blood glucose component aims to maintain the user's blood glucose within a target user blood glucose range or at a target user blood glucose level by reference to the current blood glucose measured (or received) before the detection of the unplanned ingested meal.

[0312] The compensation module assesses whether the current blood glucose level before the detection of the unplanned meal is below a predetermined hypoglycemia limit (for example, between 70 and 85 mg / dL) or above a user-predetermined target blood glucose level.

[0313] In this case, the compensation module 103 estimates the glucose component of the insulin quantity based on the current glucose level before detection of the unexpected meal, the user's predetermined target glucose level (or glucose limits) and a determined compensation ratio.

[0314] For example, the glycemic component of the quantity of insulin can be estimated using the following formula:

[0315] Blood glucose component = (pre-blood glucose) x compensation ratio

[0316] If the current blood glucose level before the detection of the unexpected meal is below the limit of predetermined hypoglycemia, then the previous formula induces a negative value of the glucose component, which means that the amount of insulin is reduced according to the negative value of the glucose component.

[0317] If the current blood glucose level before the detection of the unplanned meal is above a user-predetermined target blood glucose level, then the preceding formula induces a positive value for the glucose component, meaning that the amount of insulin is increased to reach the target blood glucose level and compensate for the unplanned meal. (with the meal component).

[0318] If the current blood glucose level before detection of the unplanned meal is not within predetermined hypoglycemia limits (for example, between 70 and 85 mg / dL) and is not above a user-predetermined target blood glucose level, then the blood glucose component is equal to zero.

[0319] Compensation steps

[0320] From an algorithmic point of view, the compensation receives as input a quantity of ingested carbohydrates (CHO) and transmits as output a quantity of insulin to compensate for the ingested carbohydrates.

[0321] The method starts as soon as the unplanned meal ingested has been estimated by the estimation module 102, which can apply any embodiment of the estimation method.

[0322] As described previously, the method for estimating the size of an unexpected meal can be applied regularly by the blood glucose management system. This estimation is preferably performed every five minutes.

[0323] Next, the compensation module 103 applies a step of calculating the quantity of insulin based at least on the estimated size of the unforeseen meal.

[0324] Any method may be used to estimate or calculate the amount of insulin.

[0325] However, the compensation can be supervised by the safety module of estimation 23 and / or the condition module 15.

[0326] The safety module of estimation 23 is adapted to determine whether the estimated size of the unplanned meal exceeds a meal estimation threshold value and to set the estimated size to a value equal to the meal estimation threshold value if the estimated size of the unplanned meal exceeds the meal estimation threshold value.

[0327] It can be considered as an input supervisor since the safety module of estimation 23 supervises the algorithmic input of compensation.

[0328] Thus, by controlling the maximum input value of the compensation module 103, the amount of insulin estimated by the compensation module 103 is also controlled. This is a first safety measure.

[0329] On the other hand, the condition module 15 can be considered as an output supervisor since it is suitable for checking whether at least one condition is met before applying compensation and, above all, the calculation step and / or before recommending to the user the estimate provided by the compensation module 103.

[0330] This condition can also be considered a safety requirement. To ensure user safety, it may be relevant to monitor the user's active insulin (OIB). The minimum condition may also be related to the total amount of insulin calculated over a predetermined past period, by example the last hour. This is why the minimum condition can be: an IOB(t) variable, representing the variation over time of the amount of active insulin of the user, is less than a predetermined value and / or a total amount of insulin calculated during a predetermined past period is less than a predetermined value.

[0331] If the minimum condition is not met, the calculation step is not applied or the output of the calculation step is set to zero or is null. The output can also be: "no recommendation".

[0332] Frequency

[0333] The compensation is applied by the system at the end of each predetermined period, for example, every five minutes. In this embodiment, the estimated size of the unplanned meal corresponds to a blood glucose deviation over five minutes. This is generally a small deviation, resulting in a small estimated size and therefore a small insulin bolus compared to a deviation over a longer period, for example, 15 minutes, as is usually the case.

[0334] For this reason, the estimation step and the calculation step are repeated by the estimation module 102 and the compensation module 103 a determined number of times, for example four times in succession, with a determined frequency, for example every five minutes.

[0335] The predetermined number of times can be between 2 and 10, preferably between 3 and 5, and ideally equal to 4, as described above.

[0336] The predetermined frequency is between 1 and 15 minutes, preferably between 1 and 10 minutes, and ideally equal to 5 minutes, as described above.

[0337] This short period allows for fine management of blood glucose from the start of an unexpected meal.

[0338] For example, an aperitif may last 20 minutes. Every five minutes, compensation is applied and an estimated size of the corresponding ingested meal is determined, then used to manage the imminent rise in blood glucose due to the five minutes of aperitif.

[0339] The system allows the imminent rise to be managed from the beginning, here the first 5 minutes, and therefore promotes smooth blood glucose management by administering insulin if necessary based on the estimate of the unforeseen meal ingested.

Claims

Demands

1. A method for estimating the size of an unplanned meal ingested by a user, the method being applied by a system, this system comprising: - a storage module (301), this storage module recording a time series of the user's blood glucose values ​​in a blood glucose storage, - a forecasting module (302), this forecasting module (302) determining an expected change in blood glucose at a given time, - an estimation module (11), this estimation module (11) estimating the size of the unplanned meal ingested based on: - a difference in change between - a change in blood glucose measured between a blood glucose value at a given time and a blood glucose value at a time in the time series preceding the given time, the blood glucose values ​​being collected by a blood glucose acquisition system or recorded in the blood glucose storage;and - the expected change in blood glucose between a blood glucose value at a given time and a blood glucose value at a point in the time series preceding that time; - at least one user-related parameter that is customized according to the user.

2. A method for estimating the size of an unplanned meal ingested according to the preceding claim, wherein the prediction module (302) determines the expected variation in blood glucose at a given time based on at least: - the user's insulin sensitivity factor, and / or - an instantaneous consumption of insulin by the user.

3. A method for estimating the size of an unplanned meal ingested in accordance with any one of the preceding claims 1 to 2, wherein the user-linked and user-customized parameter is based on the user's weight.

4. A method for estimating the size of an unplanned meal ingested in accordance with any one of the preceding claims 1 to 3, wherein the system also includes a first estimation safety module (13a), this first estimation safety module (13a) determining: - whether the measured blood glucose variation determined by the estimation module (11) exceeds a threshold value of difference of variation and setting the measured blood glucose variation to a value equal to the threshold value of difference of variation if the measured blood glucose variation determined by the estimation module (11) exceeds the threshold value of difference of variation.

5. A method for estimating the size of an unplanned meal ingested in accordance with any one of the preceding claims 1 to 4, wherein the system also includes a second estimation safety module (13b), this second estimation safety module (13b) determining: - whether the size of the unplanned meal ingested estimated by the estimation module (11) exceeds a meal estimation threshold value and setting the size to a value equal to the meal estimation threshold value if the size of the unplanned meal ingested estimated by the estimation module (11) exceeds the meal estimation threshold value.

6. A method for estimating the size of an unplanned meal ingested in accordance with the preceding claim, wherein the threshold value for meal estimation is calculated based on a threshold parameter determined for a given time.

7. A method for estimating the size of an unplanned meal ingested in accordance with any one of the preceding claims 1 to 6, wherein the system also includes a correction module (12), this correction module (12) correcting the size of the unplanned meal ingested estimated by the estimation module (11) by applying at least a correction coefficient, this minimum correction coefficient being predetermined according to the user's profile based at least on the user's age, weight, sex and / or insulin requirements.

8. Method for estimating the size of an unexpectedly ingested meal in accordance with any one of the preceding claims 1 to 7, wherein the system also comprises an aggressiveness module (14), this aggressiveness module (14) calculates a personalized aggressiveness estimate of the meal size by applying an aggressiveness factor to the size of the unforeseen ingested meal estimated by the estimation module (11).

9. A system for estimating the size of an unexpected meal ingested by a user, comprising: - a storage module (301) adapted to record a time series of the user's blood glucose values ​​in a blood glucose storage device, - a forecasting module (302) adapted to determine an expected variation in blood glucose at a given time, - an estimation module (11) adapted to estimate the size of an unexpected meal ingested based on: - a difference in variation between - a variation in blood glucose measured between a blood glucose value at a given moment and a blood glucose value at a point in the time series preceding that moment, blood glucose values ​​are collected by a blood glucose acquisition system or recorded in blood glucose storage; And - the expected change in blood glucose between a blood glucose value at a given moment and a blood glucose value at a point in the time series preceding the given moment; - at least one user setting linked to and customized according to the user.

10. A system for estimating the size of an unexpected meal ingested by a user in accordance with the preceding claim, this system also comprising a first estimation safety module (13a) adapted for: - determine if the difference in variation determined by the estimation module (11) exceeds a threshold value for the difference in variation and set the difference in variation to a value equal to the value threshold difference of variation if the difference of variation determined by the estimation module (11) exceeds the threshold value of difference of variation.

11. A system for estimating the size of an unplanned meal ingested by a user in accordance with any one of the preceding claims 9 or 10, this system also comprising a second estimation safety module (13b) adapted to: - determine whether the size of the unplanned meal ingested estimated by the estimation module (11) exceeds a meal estimation threshold value and set the size to a value equal to the meal estimation threshold value if the size of the unplanned meal ingested estimated by the estimation module (11) exceeds the meal estimation threshold value.

12. A system for estimating the size of an unexpected meal ingested by a user in accordance with any one of the preceding claims 9 to 11, wherein the threshold value for meal estimation is calculated based on a threshold parameter determined for a given time.

13. A system for estimating the size of an unplanned meal ingested by a user in accordance with any one of the preceding claims 9 to 12, also comprising a correction module (12) adapted to correct the size of the unplanned meal ingested estimated by the estimation module (11) by applying at least a correction coefficient, this minimum correction coefficient being predetermined according to the user's profile based at least on the user's age, weight, sex and / or insulin requirements.

14. A system for estimating the size of an unplanned meal ingested by a user in accordance with any one of the preceding claims 9 to 13, also comprising an aggressiveness module (14) adapted to calculate a personalized aggressiveness estimate of the meal size by applying an aggressiveness factor to the size of the unplanned meal ingested estimated by the estimation module (11).

15. A computer program for estimating the size of an unexpected meal ingested by a user, this computer program being adapted, when executed on a processor, to request the processor to apply the method of claims 1 to 8.

Images