Medication monitoring device, method and system for capsule inhalers

Abstract

This application provides a medication monitoring device, method, and system for a capsule inhaler, relating to the field of medical devices. The medication monitoring device includes a housing, a photoelectric sensor, an acoustic sensor, and a processor. The photoelectric sensor is disposed in the housing and is used to detect the light signal caused by the periodic change in the light path due to the rotation of the capsule within the medication chamber. The acoustic sensor is disposed inside the housing near the medication chamber and is used to detect the sound signal of the capsule striking the wall of the medication chamber when it rotates. The processor is connected to both the photoelectric sensor and the acoustic sensor, and is used to determine the rotation state of the capsule based on first information and second information. The first information is determined based on the light signal, and the second information is determined based on the sound signal. By using the acoustic sensor and the photoelectric sensor for collaborative monitoring, the technical problem of accurately determining the capsule rotation state in existing technologies is solved.

Description

Technical Field

[0001] This application relates to the field of medical device technology, and in particular to a medication monitoring device, method and system for a capsule inhaler. Background Technology

[0002] Inhalation is a direct method of drug administration. Figure 1 This is a schematic diagram of the structure of a capsule-type inhaler in the prior art. See also... Figure 1 As shown, the capsule inhaler includes a connected base and an inhalation unit. A puncture point 104 is provided on the base, and a medication chamber is disposed within the base. The medication chamber includes a connected receiving slot 101 and a spiral air channel 102, with the spiral air channel 102 located above the receiving slot 101. When using the device, the user presses the puncture point 104, which punctures the capsule in the receiving slot. The user then inhales through the inhalation unit. External gas enters the medication chamber through the air inlet 103 under negative pressure, forming a spiral airflow along the inner wall of the spiral air channel 102. The capsule, propelled by the spiral airflow, floats from the receiving slot 101 into the spiral air channel 102, where it rotates and ejects the medication. The medication is then released and inhaled by the user.

[0003] In existing technologies, photoelectric sensors have been proposed to monitor capsule rotation. For example, the optical path of the photoelectric sensor is set to pass through the air inlet and the medication compartment of the inhaler, and the capsule rotation is determined by detecting changes in the periodic blocking of the optical path by the capsule. However, the inventors found in practice that due to uneven drug distribution within the capsule (drug often accumulates at one end of the capsule, causing a shift in the center of gravity), the capsule exhibits irregular movements such as swaying and deflection when rotating. This results in irregular signal waveforms and inconsistent peak values ​​generated by the photoelectric sensor, making it difficult to accurately determine the capsule rotation status and affecting the reliability of medication monitoring. Summary of the Invention

[0004] This application provides a medication monitoring device, method, and system for a capsule inhaler to solve the technical problem of difficulty in accurately determining the capsule rotation state in the prior art.

[0005] The first aspect of this application discloses a medication monitoring device for a capsule inhaler, including a housing, a photoelectric sensor, an acoustic sensor, and a processor. The housing is used to house and fix the inhaler. The photoelectric sensor is disposed in the housing, and the optical path of the photoelectric sensor is configured to pass through the air inlet and the medication compartment of the inhaler. The photoelectric sensor is used to detect the light signal of the periodic change of the optical path caused by the rotation of the capsule in the medication compartment. The acoustic sensor is disposed in the housing near the medication compartment and is used to detect the sound signal of the capsule hitting the wall of the medication compartment when it rotates. The processor is connected to the photoelectric sensor and the acoustic sensor respectively, and the processor is used to determine the rotation state of the capsule based on first information and second information. The first information is determined based on the optical signal, and the second information is determined based on the sound signal.

[0006] In one possible implementation of the first aspect, the processor includes a master clock module connected to both a photoelectric sensor and an acoustic sensor, the master clock module being used to simultaneously trigger data acquisition from both the photoelectric sensor and the acoustic sensor.

[0007] The second aspect of this application discloses a medication monitoring method for a capsule inhaler, applicable to any medication monitoring device provided in this application. The method includes: obtaining first information, the first information being determined based on a light signal detected by a photoelectric sensor, the first information including a first rotation speed and / or a first rotation determination result; obtaining second information, the second information being determined based on a sound signal detected by an acoustic sensor, the second information including a second rotation speed and / or a second rotation determination result; and determining the rotation state of the capsule based on the first information and / or the second information, the rotation state including whether the capsule is rotating and / or a target rotation speed.

[0008] In one possible implementation of the second aspect, determining the rotational state of the capsule based on the first information and / or the second information includes: within each sliding time window, calculating in real time the confidence scores of the photoelectric sensor and the acoustic sensor based on the optical signal and the sound signal, wherein the confidence scores are calculated based on a weighted average of one or more of the following evaluation factors: signal amplitude, signal-to-noise ratio, signal stability, historical reliability, and current environmental parameters; wherein the signal amplitude reflects the sensor output intensity, the signal-to-noise ratio reflects the signal purity, the signal stability is characterized by the coefficient of variation, the historical reliability is based on the judgment accuracy rate in historical medication, and the current environmental parameters include the ambient noise level and / or ambient light intensity; and determining the rotational state of the capsule based on the confidence scores of the photoelectric sensor and the acoustic sensor, according to the first information and / or the second information.

[0009] In one possible implementation of the second aspect, the rotational state of the capsule is determined based on the confidence scores of the photoelectric sensor and the acoustic sensor, according to first information and / or second information, including: if the difference between C_photo and C_audio is less than a first difference threshold, and the first rotational speed in the first information is inconsistent with the second rotational speed in the second information, then the first rotational speed and the second rotational speed are weighted and fused to obtain the target rotational speed of the capsule; the weighted fusion of the first rotational speed and the second rotational speed to obtain the target rotational speed of the capsule includes: determining a first weighted rotational speed based on the product of C_photo and the first rotational speed, determining a second weighted rotational speed based on the product of C_audio and the second rotational speed, and determining the target rotational speed based on the quotient of the sum of the first weighted rotational speed and the second weighted rotational speed and the sum of C_photo and C_audio.

[0010] In one possible implementation of the second aspect, the rotation state of the capsule is determined based on the confidence scores of the photoelectric sensor and the acoustic sensor, according to the first information and / or the second information, including: if the difference between C_photo and C_audio is less than a second difference threshold, and the first rotation judgment result in the first information is inconsistent with the second rotation judgment result in the second information, then the confidence levels of the photoelectric sensor and the acoustic sensor for rotation and non-rotation are determined respectively according to the first rotation judgment result and C_photo in the first information, and the second rotation judgment result and C_audio in the second information; and the capsule rotation is determined by confidence fusion.

[0011] In one possible implementation of the second aspect, based on the first rotation judgment result and C_photo in the first information, and the second rotation judgment result and C_audio in the second information, the confidence levels of the photoelectric sensor and the acoustic sensor for rotation and non-rotation are determined respectively, including: if the first rotation judgment result is rotation, then the confidence level of the photoelectric sensor for rotation is determined based on C_photo, and the confidence level of the photoelectric sensor for non-rotation is determined based on the confidence level of the photoelectric sensor for rotation; if the first rotation judgment result is non-rotation, then the confidence level of the photoelectric sensor for non-rotation is determined based on C_photo, and the confidence level of the photoelectric sensor for non-rotation is determined based on the confidence level of the photoelectric sensor for non-rotation. The photoelectric sensor's confidence level for rotation is set; the sum of the photoelectric sensor's confidence level for rotation and its confidence level for no rotation equals a set value; if the second rotation judgment result is rotation, then the acoustic sensor's confidence level for rotation is determined based on C_audio, and the acoustic sensor's confidence level for no rotation is determined based on the acoustic sensor's confidence level for rotation; if the second rotation judgment result is no rotation, then the acoustic sensor's confidence level for no rotation is determined based on C_audio, and the acoustic sensor's confidence level for rotation is determined based on the acoustic sensor's confidence level for no rotation; the sum of the acoustic sensor's confidence level for rotation and its confidence level for no rotation equals a set value.

[0012] In one possible implementation of the second aspect, determining whether the capsule is rotating through trust fusion includes: determining a rotation trust product based on the product of the trust in rotation by the photoelectric sensor and the trust in rotation by the acoustic sensor; determining a non-rotation trust product based on the product of the trust in non-rotation by the photoelectric sensor and the trust in non-rotation by the acoustic sensor; comparing the rotation trust product and the non-rotation trust product, and determining whether the capsule is rotating based on the comparison result.

[0013] In one possible implementation of the second aspect, before obtaining the second information, the method further includes: before inhalation begins, acquiring ambient noise and determining the mean and standard deviation of the ambient noise; during inhalation, calculating a dynamic threshold in real time based on the mean and standard deviation of the ambient noise and the peak value of the real-time acquired sound signal; filtering the sound signal using the dynamic threshold to filter out noise components below the dynamic threshold; and determining the second information based on the filtered sound signal.

[0014] A third aspect of this application also discloses a medication monitoring system for a capsule inhaler, including a medication monitoring device for a capsule inhaler and a capsule inhaler as disclosed in the embodiments of this application.

[0015] The technical solution of this application combines the synergistic effect of photoelectric sensors and acoustic sensors to achieve multi-dimensional monitoring of the capsule's rotation state, effectively solving the problem of insufficient monitoring reliability caused by irregular capsule movement in the prior art. Specifically, the photoelectric sensor captures the periodic changes in the optical path caused by the capsule's rotation, while the acoustic sensor collects the sound signals generated when the capsule strikes the drug chamber wall during rotation. The processor fuses these two pieces of information to make a more comprehensive and accurate determination of the capsule's rotation state. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of the structure of a capsule-type inhaler in the prior art; Figure 2 This is a schematic diagram of the structure of a medication monitoring device for a capsule inhaler and the assembled capsule inhaler, provided in an embodiment of this application. Figure 3 This is a schematic diagram of the medication monitoring device for a capsule inhaler and the assembled capsule inhaler, provided in an embodiment of this application, from another perspective. Figure 4 This is a schematic flowchart of a medication monitoring method for a capsule inhaler provided in an embodiment of this application; Figure 5 This is a schematic flowchart of another medication monitoring method for a capsule inhaler provided in the embodiments of this application; Figure 6 This is a schematic flowchart of another medication monitoring method for a capsule inhaler provided in the embodiments of this application.

[0018] In the figure, 100 is the inhaler; 101 is the receiving groove; 102 is the spiral air groove; 103 is the air inlet; 104 is the puncture site; 105 is the inhalation port, transmitter 10, receiver 11, and acoustic sensor 12. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be described in detail below. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments of this application, all other implementation methods obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0020] In the description of the embodiments of this application, it should be understood that the terms "center," "length," "width," "height," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," and "side," etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the accompanying drawings and are only for the convenience of describing the embodiments of this application and simplifying the description, and do not indicate or imply that the device or component referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation on the embodiments of this application. In the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0021] See Figure 1 As shown, in the prior art, a capsule inhaler includes a connected base and an inhalation unit. The base is provided with a puncture part 104 and a drug chamber is provided inside the base. The drug chamber includes a connected receiving groove 101 and a spiral air groove 102. The spiral air groove 102 is located above the receiving groove 101. When the user uses it, he / she presses the puncture part 104, which punctures the capsule in the receiving groove. The user inhales with the inhalation unit in his / her mouth. Under the action of negative pressure, external gas enters the drug chamber from the air inlet 103 and forms a spiral airflow along the inner wall of the spiral air groove 102. The capsule is affected by the spiral airflow and floats from the receiving groove 101 into the spiral air groove 102. While rotating in the spiral air groove 102, it throws out the powdered drug. The powdered drug is released and inhaled into the user's mouth and lungs.

[0022] Current methods for monitoring capsule rotation status are limited, typically relying solely on photoelectric sensors. When the capsule undergoes irregular movement within the spiral air channel (such as swaying, jamming, or abnormal collisions with the channel wall), the photoelectric sensor is susceptible to interference, leading to inaccurate monitoring results. For example, when the capsule deviates from the center of the optical path due to unstable airflow or uneven surface reflection, fluctuations in the optical signal amplitude and a decrease in the signal-to-noise ratio can occur, resulting in misjudgments of whether the capsule is rotating or misestimation of its rotation speed. Ultimately, this affects the reliability of the capsule rotation status assessment and fails to provide accurate data support for medication efficacy evaluation and patient compliance management.

[0023] To address the aforementioned issues, this application provides a medication monitoring device for a capsule inhaler. The device integrates a photoelectric sensor and an acoustic sensor, forming a dual-sensor monitoring mechanism. The photoelectric sensor is configured to align with a specific area of ​​the spiral air groove in the capsule inhaler, emitting a detection beam and receiving light signals reflected or blocked by the capsule. By analyzing the periodic variation characteristics of this light signal (such as pulse frequency and duty cycle), a sensor characterizing the capsule's rotation state is generated. Simultaneously, the acoustic sensor is positioned on the outer wall of the inhaler's medication chamber or at a location with good vibration transmission to the chamber. It collects sound signals generated when the capsule rotates within the spiral air groove due to impacts with the groove wall and airflow disturbances. By performing spectral analysis and feature extraction (such as energy values ​​in specific frequency bands and the interval of impact pulses) on these sound signals, a sensor characterizing the capsule's rotation state is generated. The medication monitoring device also includes a processor electrically connected to both the photoelectric sensor and the acoustic sensor. The processor performs fusion decision-making based on information from the two sensors to improve the accuracy and reliability of the capsule's rotation state determination.

[0024] However, in the process of realizing this invention, the inventors discovered the following technical difficulties in the fusion decision-making process of photoelectric sensors and acoustic sensors.

[0025] The incompatibility of heterogeneous data types. The target rotational speed is a numerical regression result; while the rotation judgment result is a classification judgment (e.g., rotation or no rotation), which is a Boolean classification result. The two types of data differ in physical meaning, dimensions, and noise distribution, making direct fusion difficult.

[0026] Arbitration of Boolean-based conflicting judgments is difficult. When two sensors give conflicting conclusions about whether the capsule is rotating (e.g., photoelectric detection determines rotation due to periodic obstruction, while acoustic detection determines no rotation due to lack of regular impact caused by ambient noise), there is a lack of a unified arbitration basis. Simply using "AND" or "OR" logic introduces a high risk of false alarms or false negatives.

[0027] The challenge of fusing numerical estimation biases. When two sensors both determine rotation but measure different rotational speeds, the fixed weighting strategy cannot adapt to real-time signal quality fluctuations due to the dynamic changes in confidence levels of each sensor under different operating conditions (e.g., weak impact signals in acoustic sensors at low speeds, and inaccurate counting in photoelectric sensors due to motion blur at high speeds).

[0028] The misalignment is caused by timescale asynchrony. The photoelectric sensor and the acoustic sensor have different sensitivities at the start and end of the rotation, resulting in lag or acceleration of the fusion results near the start and end points of the rotation.

[0029] Adaptive adjustment in dynamic environments. Ambient lighting, background noise, capsule powder distribution, and inhalation airflow velocity all change in real time during medication administration. The fusion strategy must have adaptive capabilities; otherwise, fixed parameters will lead to a sharp decline in fusion performance.

[0030] The technical solution of this application overcomes the above-mentioned technical difficulties and designs a layered heterogeneous fusion architecture: a confidence allocation mechanism is introduced to address Boolean type judgment conflicts. A confidence-weighted fusion method is used to address numerical estimation biases. This solves the problem of incompatibility between heterogeneous data types.

[0031] By establishing a trust level allocation mechanism for each sensor, multiplying the trust levels of two sensors for rotation to obtain the rotation trust level product, and multiplying the trust levels of two sensors for non-rotation to obtain the non-rotation trust level product, and comparing the size of the products to determine the final judgment result, the technical problems of lacking a unified arbitration basis when two sensors give opposite conclusions on "whether the capsule is rotating" and the simple "AND / OR" logic easily leading to high false alarms or missed alarms are solved.

[0032] By dynamically calculating the confidence scores of each sensor in real time based on multiple evaluation factors such as signal amplitude, signal-to-noise ratio, signal stability, historical reliability, and current environmental parameters, and employing weighted fusion when both sensors have high confidence scores but the measured rotational speeds are inconsistent, the high-confidence sensor contributes more to the rotational speed, while the contribution of the low-confidence sensor is suppressed. This solves the technical problem of dynamic changes in sensor confidence under different operating conditions and the inability of fixed weighting strategies to adapt to signal quality fluctuations that lead to rotational speed fusion deviations.

[0033] By setting a master clock module in the processor and connecting it to both the photoelectric sensor and the acoustic sensor respectively, the data acquisition of both is triggered simultaneously to achieve hardware-level synchronization. Each frame of data is assigned a microsecond-level hardware timestamp, and a linear interpolation algorithm is used for software alignment of the sensor with a low sampling rate. This solves the misalignment problem caused by the time scale asynchrony between the photoelectric sensor and the acoustic sensor.

[0034] By incorporating current environmental parameters (ambient noise level and / or ambient light intensity) into the evaluation factors of the confidence score, the sensor confidence level decreases and the fusion weight automatically decreases as the environment deteriorates. In the preprocessing of the sound signal, the filtering threshold is dynamically calculated based on the mean and standard deviation of ambient noise and the real-time signal-to-noise ratio to achieve adaptive noise suppression. This solves the technical problem that the fusion performance of fixed parameters drops sharply due to real-time changes in ambient light, background noise, powder distribution, and inhalation airflow velocity.

[0035] This application provides a medication monitoring device for a capsule inhaler, including a housing, a photoelectric sensor, an acoustic sensor, and a processor.

[0036] The housing is used to house and fix the inhaler; a photoelectric sensor is disposed in the housing, and the optical path of the photoelectric sensor is configured to pass through the air inlet and the capsule of the inhaler. The photoelectric sensor is used to detect the light signal of the periodic change of the optical path caused by the rotation of the capsule in the capsule; an acoustic sensor is disposed in the housing near the capsule, and the acoustic sensor is used to detect the sound signal of the capsule hitting the wall of the capsule when it rotates; a processor is connected to the photoelectric sensor and the acoustic sensor respectively, and the processor is used to determine the rotation state of the capsule based on first information and second information; the first information is determined based on the light signal, and the second information is determined based on the sound signal.

[0037] The first information includes a set of data characterizing capsule rotation information determined based on light signals detected by a photoelectric sensor. The first information includes at least a first rotational speed and / or a first rotation determination result. The first rotational speed includes the angular velocity or linear velocity of the capsule rotation determined based on the light signals. The first rotation determination result includes a Boolean value indicating whether the capsule is rotating, determined based on the light signals. For example, if the periodic change characteristics of the detected light signals meet a preset condition, it is determined to be rotating, and "yes" is output; otherwise, it is determined not to be rotating, and "no" is output.

[0038] The second information includes a set of data characterizing capsule rotation information determined based on sound signals detected by acoustic sensors. The second information includes at least a second rotational speed and / or a second rotation determination result. The second rotational speed includes the angular velocity or linear velocity of the capsule rotation determined based on the interval time of impact pulses or the periodic variation of energy in a specific frequency band in the sound signal. The second rotation determination result includes a Boolean value indicating whether the capsule is rotating, based on the presence of a regular impact signal in the sound signal that matches the characteristics of capsule rotation. For example, if the detected sound signal, after filtering and feature extraction, is determined to be rotating based on the features, output "Yes"; otherwise, it is determined not to be rotating, and output "No".

[0039] The determination of the first and second information can be achieved by having the sensor directly output the original light signal and the original sound signal, and the processor obtaining the rotation speed and rotation judgment result through analog-to-digital conversion and algorithm calculation.

[0040] Another way to determine the first and second information is to have the sensor integrate a signal processing circuit to directly output the calculated rotational speed value and the judgment result, while the processor only reads the calculated rotational speed value and the judgment result.

[0041] For example, see Figures 1 to 3 This embodiment provides a medication monitoring device for a capsule inhaler, including a housing 1, a photoelectric sensor, an acoustic sensor 12, and a processor (not shown in the figure). The photoelectric sensor may include a transmitter 10 and a receiver 11, and the positions of the transmitter 10 and the receiver 11 may be interchanged; this embodiment does not limit this.

[0042] The housing 1 is used to house and secure the inhaler 100. The housing 1 can be a split structure, for example, comprising a base 7 and a cover 5. One side of the cover 5 is hinged to the base 7 via a pivot, and the other side is snap-fitted into the base 7. The cover 5 has a limiting cavity whose shape matches the outer contour of the inhaler 100. When the inhaler 100 is inserted into the limiting cavity and the cover 5 is engaged, axial fixation is achieved. The patient can administer medication through the inhalation port 105 of the inhaler 100.

[0043] A photoelectric sensor is mounted on the housing 1. Its optical path is configured to pass through the air inlet 103 and the medication compartment (including the receiving slot 101 and the spiral air slot 102) of the inhaler 100. The photoelectric sensor can be a photoelectric through-beam sensor or a photoelectric reflective sensor.

[0044] The acoustic sensor 12 is located inside the housing 1 near the medicine compartment. It can be a capacitive microphone or a piezoelectric sensor. A capacitive microphone is small, has low power consumption, and can achieve sampling frequencies above 20kHz. A piezoelectric sensor is directly attached to the outer wall of the medicine compartment and is sensitive to impact vibrations.

[0045] The processor is housed within the casing 1 and is electrically connected to both the photoelectric sensor and the acoustic sensor. The processor can be a microcontroller (MCU), digital signal processor (DSP), field-programmable gate array (FPGA), or application-specific integrated circuit (ASIC). The processor is used to acquire first information and second information; it is also used to determine the capsule's rotational state based on the first and second information.

[0046] To achieve synchronous triggering, the processor in this embodiment may further include a master clock module, which is connected to both the photoelectric sensor and the acoustic sensor. The master clock module is used to simultaneously trigger data acquisition from both the photoelectric sensor and the acoustic sensor. The master clock module can generate a synchronization pulse at a fixed frequency (e.g., 10kHz), which serves as the sampling trigger signal for both the photoelectric sensor and the acoustic sensor, thereby ensuring strict alignment of the initial acquisition time.

[0047] This application embodiment adds an acoustic sensor and a photoelectric sensor to the medication monitoring device to form a dual-sensor monitoring system. By collecting the sound signal generated when the capsule rotates and hits the wall of the medicine chamber, it supplements the monitoring blind spot of a single photoelectric sensor when the capsule deviates from the center of the light path, the surface reflection is uneven, or the high-speed motion is blurred, thus reducing misjudgments caused by light signal interference. At the same time, by utilizing the direct response characteristics of sound signals to mechanical vibration, it enhances the detection capability of abnormal motion states such as capsule jamming and irregular swinging, and improves the adaptability of the equipment to complex working conditions.

[0048] Figure 4 This is a schematic flowchart illustrating a medication monitoring method for a capsule inhaler provided in an embodiment of this application. See also... Figure 4 This application provides a medication monitoring method for a capsule inhaler, applicable to any of the aforementioned device embodiments. The method includes the following steps.

[0049] S110, first information obtained.

[0050] The first information is the same as described above and will not be repeated here. Optionally, the processor or photoelectric sensor processes the optical signal as follows: detects the peak value of the optical signal and records the time interval between adjacent peak values; calculates the first rotation speed based on the number of peak values ​​per unit time, or calculates the first rotation speed based on the average interval; and determines whether periodic occlusion is detected. If peak values ​​with a regular interval of at least the first number appear within the time window, the first rotation determination result is "rotation"; otherwise, it is "no rotation". The first number can be set according to actual needs, and can be 1 time or multiple times.

[0051] To reduce the impact of noise, the original optical signal can be denoised before calculating the first rotation speed or the first rotation determination result based on the optical signal. For example, before acquiring the original optical signal, the ambient light intensity can be acquired in real time as a dynamic baseline, and the original optical signal can be differentially analyzed with the ambient light baseline to directly eliminate interference from slow changes in illumination.

[0052] S120, obtain the second information.

[0053] The second information is the same as described above and will not be repeated here. Optionally, the processor or acoustic sensor processes the sound signal as follows: detects signal peaks exceeding a dynamic threshold, with each peak considered an impact; calculates the second rotation speed based on the number of impacts per unit time; determines whether regular impacts are detected: if at least a second number of impacts occur within the time window and the intervals are relatively stable, the second rotation result is "rotation"; otherwise, it is "no rotation". The second number of impacts can be set according to actual needs, and can be once or multiple times.

[0054] To reduce the impact of noise, the original sound signal may be denoised before calculating the second rotation speed or the second rotation determination result based on the sound signal. For example, before inhalation begins, ambient noise is collected, and the mean and standard deviation of the ambient noise are determined; during inhalation, a dynamic threshold is calculated in real time based on the mean and standard deviation of the ambient noise and the peak value of the real-time collected sound signal; the sound signal is filtered using the dynamic threshold to remove noise components below the dynamic threshold; and the second information is determined based on the filtered sound signal. The sensor for collecting ambient noise can be the same as or different from the sensor for collecting sound signals. Optionally, the sensor for collecting ambient noise is placed in a location within the medication monitoring device that facilitates the collection of ambient noise, while the sensor for collecting sound signals can be placed in a location within the medication monitoring device that facilitates the collection of the sound of the capsule rotating and impacting the drug compartment wall.

[0055] Dynamic thresholding is a filtering threshold that is adaptively calculated in the preprocessing of audio signals based on the statistical characteristics of ambient noise and the real-time signal-to-noise ratio, and is used to distinguish effective impact signals from background noise.

[0056] S130, determine the capsule rotation state based on the first information and the second information.

[0057] Rotation status can include whether the capsule is rotating and / or the target rotation speed. The target rotation speed includes the final angular velocity or linear velocity of the capsule after fusion, which is used to determine whether the user's medication is up to standard. Generally, the capsule rotation speed that meets the medication requirements needs to reach a preset minimum rotation speed threshold.

[0058] Optionally, determining the capsule rotation state based on the first information and the second information may include: when the first information and the second information are consistent, determining the final result based on the first information or the second information; when there is a conflict between the two, comparing the regularity of the light signal and the sound signal, and determining the final result based on the first information or the second information with the higher regularity.

[0059] The execution order of S110 and S120 is not limited in this embodiment. Optionally, S120 may be executed before or simultaneously with S110.

[0060] This application embodiment determines the capsule rotation state by combining the first information and the second information. It can comprehensively utilize the precise detection capability of optical signals for rotation cycle and the direct sensing characteristics of sound signals for mechanical motion, effectively making up for the monitoring limitations of a single sensing method under complex working conditions.

[0061] Figure 5 This is a schematic flowchart illustrating another medication monitoring method for a capsule inhaler provided in this application. See also... Figure 5Based on the above-described medication monitoring method, this application further refines step S130, which includes the following steps.

[0062] S131, within each sliding time window, calculate the confidence scores of the photoelectric sensor and the acoustic sensor in real time based on the optical signal and the sound signal. The confidence scores are calculated based on a weighted average of one or more of the following evaluation factors: signal amplitude, signal-to-noise ratio, signal stability, historical reliability, and current environmental parameters.

[0063] Signal amplitude reflects the sensor output intensity, signal-to-noise ratio reflects signal purity, signal stability is characterized by the coefficient of variation, historical reliability is based on the accuracy of judgments in historical medication use, and current environmental parameters include environmental noise level and / or ambient light intensity.

[0064] S132, based on the confidence scores of the photoelectric sensor and the acoustic sensor, determine the rotation state of the capsule according to the first information and / or the second information.

[0065] The confidence score is a quantitative indicator that represents the reliability of the sensor's current output signal, ranging from 0 to 1. A higher value indicates a more reliable signal.

[0066] For example, determining the capsule's rotation state based on the confidence scores of the photoelectric sensor and the acoustic sensor, according to first information and / or second information, may include: comparing whether the first information and the second information are consistent. If they are consistent, the capsule's rotation state is determined based on the consistency result. If they are inconsistent, the difference between the confidence scores of the photoelectric sensor and the acoustic sensor is further calculated, and the calculated confidence difference is compared with a preset difference threshold. If the confidence difference is greater than the difference threshold, it indicates that the judgment reliability of one sensor (photoelectric sensor or acoustic sensor) is significantly higher than that of the other. In this case, the information corresponding to the sensor with higher confidence is used as the final judgment basis to quickly converge the decision.

[0067] This application embodiment calculates the confidence scores of the photoelectric sensor and the acoustic sensor in real time based on the light signal and the sound signal within each sliding time window, and dynamically adjusts the decision strategy based on the confidence scores. When the two sensors give inconsistent results, the output of the sensor with higher reliability is automatically given priority, which greatly reduces the probability of misjudgment caused by changes in operating conditions and improves the robustness of rotation state judgment.

[0068] Figure 6 This is a schematic flowchart illustrating another medication monitoring method for a capsule inhaler provided in this application. See also... Figure 6 Based on the above-described medication monitoring method, this application further refines step S132, which includes the following steps.

[0069] S1321, if the difference between C_photo and C_audio is less than the first difference threshold, and the first rotation speed in the first information is inconsistent with the second rotation speed in the second information, then the first rotation speed and the second rotation speed are weighted and fused to obtain the target rotation speed of the capsule.

[0070] C_photo includes the confidence score of the photoelectric sensor. C_audio includes the confidence score of the acoustic sensor. The first difference threshold is a preset positive number used to determine whether the difference between the confidence scores of the two sensors is significant. When |C_photo - C_audio| < the first difference threshold, the confidence scores of the two sensors are considered to be similar.

[0071] For example, weighted fusion of the first speed and the second speed to obtain the target speed of the capsule may include: determining the first weighted speed based on the product of C_photo and the first speed, determining the second weighted speed based on the product of C_audio and the second speed, and determining the target speed based on the quotient of the sum of the first weighted speed and the second weighted speed and the sum of C_photo and C_audio.

[0072] Optionally, determining the first weighted speed based on the product of C_photo and the first speed can be achieved by multiplying C_photo by the first speed, or by multiplying C_photo by the first speed and then by a set weighting coefficient. Correspondingly, the calculation of the second weighted speed and the target speed can or can not incorporate weighting coefficients, depending on actual needs.

[0073] S1322, if the difference between C_photo and C_audio is less than the second difference threshold, and the first rotation judgment result in the first information is inconsistent with the second rotation judgment result in the second information, then based on the first rotation judgment result and C_photo in the first information, and the second rotation judgment result and C_audio in the second information, determine the confidence level of the photoelectric sensor and the acoustic sensor for rotation and non-rotation respectively; determine whether the capsule rotates by confidence level fusion.

[0074] The second difference threshold is similar to the first difference threshold and is used to determine the confidence difference when fusing rotation judgment results. The first difference threshold and the second difference threshold can be the same or different. In Boolean decision-making, the confidence level represents the degree of support each sensor has for the "rotate" or "don't rotate" proposition, with a value ranging from 0 to 1, and the sum of the confidence levels for two propositions for the same sensor is 1.

[0075] For example, determining the confidence levels of the photoelectric sensor and the acoustic sensor for rotation and non-rotation based on the first rotation judgment result and C_photo in the first information, and the second rotation judgment result and C_audio in the second information, respectively, may include: if the first rotation judgment result is rotation, then determining the confidence level of the photoelectric sensor for rotation based on C_photo, and determining the confidence level of the photoelectric sensor for non-rotation based on the confidence level of the photoelectric sensor for rotation; if the first rotation judgment result is non-rotation, then determining the confidence level of the photoelectric sensor for non-rotation based on C_photo, and determining the confidence level of the photoelectric sensor for non-rotation based on the confidence level of the photoelectric sensor for non-rotation. The confidence level of the device for rotation; the sum of the confidence level of the photoelectric sensor for rotation and the confidence level of the photoelectric sensor for non-rotation equals a set value; if the second rotation judgment result is rotation, then the confidence level of the acoustic sensor for rotation is determined according to C_audio, and the confidence level of the acoustic sensor for non-rotation is determined according to the confidence level of the acoustic sensor for rotation; if the second rotation judgment result is non-rotation, then the confidence level of the acoustic sensor for non-rotation is determined according to C_audio, and the confidence level of the acoustic sensor for rotation is determined according to the confidence level of the acoustic sensor for non-rotation; the sum of the confidence level of the acoustic sensor for rotation and the confidence level of the acoustic sensor for non-rotation equals a set value. The normalization constant in the confidence level allocation is typically set to 1.

[0076] Optionally, if the first rotation determination result is rotation, the confidence level of the photoelectric sensor for rotation is determined according to C_photo. This can include the confidence level of the photoelectric sensor for rotation being equal to C_photo, or the confidence level of the photoelectric sensor for rotation being equal to the product of C_photo and a set coefficient, or the confidence level of the photoelectric sensor for rotation being equal to the difference between C_photo and a set value. This embodiment of the application does not limit this. Correspondingly, the confidence level of the acoustic sensor for different states can also be allocated in a similar manner. The specific calculation rules can be adjusted according to the actual test results, and this embodiment of the application does not limit this.

[0077] Determining whether a capsule is rotating by trust fusion can include: determining a rotation trust product based on the product of the trust in rotation from the photoelectric sensor and the trust in rotation from the acoustic sensor; determining a non-rotation trust product based on the product of the trust in non-rotation from the photoelectric sensor and the trust in non-rotation from the acoustic sensor; comparing the rotation trust product with the non-rotation trust product, and determining whether the capsule is rotating based on the comparison result.

[0078] For example, determining whether the capsule is rotating based on the comparison result may include determining that the capsule is rotating if the product of the rotation confidence level is greater than the product of the non-rotation confidence level; otherwise, determining that the capsule is not rotating.

[0079] Optionally, the execution order of S1321 and S1322 is not limited in the embodiments of this application. S1322 can be executed before S1321 or at the same time as S1321.

[0080] The technical solution of this application embodiment, when the difference between C_photo and C_audio is less than a second difference threshold and the first rotation judgment result is inconsistent with the second rotation judgment result, uses the confidence scores of the two sensors to assign the corresponding state confidence level and then completes the fusion decision. This fully preserves the effective judgment information of the two sensors, avoids information waste caused by forced binary selection, and further improves the accuracy of rotation state judgment. When the confidence scores of the two sensors are similar and the rotation speed results are inconsistent, the target rotation speed is obtained by fusing through confidence weighting, which can better fit the actual rotation of the capsule and provide a more accurate data foundation for subsequent drug effect judgment.

[0081] For example, an optional medication monitoring method can be described as follows: Before the user inhales, the medication monitoring device first collects ambient light through a photoelectric sensor to determine the baseline intensity of ambient light, and at the same time collects ambient noise to determine the mean and standard deviation of ambient noise, thus completing the environmental baseline initialization.

[0082] After the user begins inhalation, the medication monitoring device synchronously collects light and sound signals in real time: the light signal undergoes ambient light baseline differential processing (e.g., subtracting the ambient light baseline intensity from the original light signal), and first information (including a first rotation judgment result and a first rotation speed) is determined based on the differential light signal. The first rotation judgment result is generated by detecting the periodic occlusion characteristics of the light signal, and the first rotation speed is calculated based on the number of peaks or the average interval per unit time. The sound signal is filtered using a dynamic threshold determined based on the mean and standard deviation of ambient noise, and second information (including a second rotation judgment result and a second rotation speed) is determined based on the filtered sound signal. The second rotation judgment result is generated by detecting the regular impact characteristics of the filtered sound signal, and the second rotation speed is calculated based on the number of impacts per unit time.

[0083] During the user's inhalation, C_photo and C_audio are calculated in real time within each sliding time window (the window length and sliding step size can be preset; the window can be a moment or a time period), and the first information and the second information are compared: if the rotation judgment results are consistent, the consistent result is directly output as the conclusion of whether the capsule is rotating; if they are inconsistent, |C_photo-C_audio| is calculated. When the difference is greater than or equal to the set difference threshold (such as the first difference threshold or the second difference threshold), the calculation result of the sensor with higher confidence is adopted; otherwise, based on the above fusion strategy, the calculation results of the two sensors are fused to determine the rotation state of the capsule.

[0084] After the user finishes inhaling, the inhalation duration is calculated based on the start and end times of rotation. The average rotation speed of the entire inhalation process is determined based on the rotation speed at different times. Finally, the medication is evaluated based on the results of whether the capsule rotates, the target rotation speed, the inhalation duration, and the overall average rotation speed to determine whether the medication has met the target and complete the medication monitoring.

[0085] This application also provides a medication monitoring system for a capsule inhaler, including any of the medication monitoring devices described in this application and a capsule inhaler. The capsule inhaler can be a Breezhaler-type inhaler or other capsule-type powder inhaler.

[0086] The specific features, structures, or characteristics described in this specification may be combined in any suitable manner in one or more embodiments or examples.

[0087] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0088] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A medication monitoring device for a capsule inhaler, characterized in that, Includes housing, photoelectric sensor, acoustic sensor, and processor; The housing is used to accommodate and secure the inhaler; The photoelectric sensor is disposed in the housing, and the optical path of the photoelectric sensor is configured to pass through the air inlet and the medication compartment of the inhaler. The photoelectric sensor is used to detect the light signal of the periodic change of the optical path caused by the rotation of the capsule in the medication compartment. The acoustic sensor is located inside the housing near the medicine compartment, and is used to detect the sound signal of the capsule hitting the wall of the medicine compartment when it rotates; The processor is connected to the photoelectric sensor and the acoustic sensor respectively, and the processor is used to determine the rotation state of the capsule based on the first information and the second information; The first information is determined based on the optical signal, and the second information is determined based on the sound signal; Determining the capsule's rotation state based on the first and second information includes: within each sliding time window, calculating the confidence scores of the photoelectric sensor and the acoustic sensor in real time based on the light signal and the sound signal. The confidence scores are calculated based on a weighted average of one or more of the following evaluation factors: signal amplitude, signal-to-noise ratio, signal stability, historical reliability, and current environmental parameters. The signal amplitude reflects the sensor's output intensity, the signal-to-noise ratio reflects signal purity, the signal stability is characterized by the coefficient of variation, the historical reliability is based on the accuracy rate of judgments in historical medication use, and the current environmental parameters include... The ambient noise level and / or ambient light intensity; if the difference between the confidence score C_photo of the photoelectric sensor and the confidence score C_audio of the acoustic sensor is less than a first difference threshold, and the first rotational speed in the first information is inconsistent with the second rotational speed in the second information, then a first weighted rotational speed is determined based on the product of C_photo and the first rotational speed, a second weighted rotational speed is determined based on the product of C_audio and the second rotational speed, and the target rotational speed of the capsule is determined based on the quotient of the sum of the first weighted rotational speed and the second weighted rotational speed and the sum of C_photo and C_audio.

2. The medication monitoring device according to claim 1, characterized in that, The processor includes a master clock module, which is connected to the photoelectric sensor and the acoustic sensor respectively. The master clock module is used to simultaneously trigger data acquisition from the photoelectric sensor and the acoustic sensor.

3. A method for monitoring medication use in a capsule inhaler, applied to the medication monitoring device according to claim 1 or 2, characterized in that, include: Obtain first information, which is determined based on the light signal detected by the photoelectric sensor, and the first information includes a first rotation speed and / or a first rotation determination result; Obtain second information, which is determined based on the sound signal detected by the acoustic sensor, and the second information includes a second rotation speed and / or a second rotation determination result; Based on the first information and / or the second information, determine the rotation state of the capsule, the rotation state including whether the capsule is rotating and / or the target rotation speed; Determining the rotational state of the capsule based on the first information and / or the second information includes: Within each sliding time window, confidence scores for the photoelectric sensor and the acoustic sensor are calculated in real time based on the optical signal and the sound signal. The confidence scores are calculated based on a weighted average of one or more of the following evaluation factors: signal amplitude, signal-to-noise ratio, signal stability, historical reliability, and current environmental parameters. The signal amplitude reflects the sensor output intensity, the signal-to-noise ratio reflects the signal purity, the signal stability is characterized by the coefficient of variation, the historical reliability is based on the accuracy rate of judgments in historical medication use, and the current environmental parameters include the ambient noise level and / or ambient light intensity. Based on the confidence scores of the photoelectric sensor and the acoustic sensor, the rotation state of the capsule is determined according to the first information and / or the second information; The determination of the capsule's rotational state based on the confidence scores of the photoelectric sensor and the acoustic sensor, according to the first information and / or the second information, includes: If the difference between the confidence score C_photo of the photoelectric sensor and the confidence score C_audio of the acoustic sensor is less than a first difference threshold, and the first rotation speed in the first information is inconsistent with the second rotation speed in the second information, then the first rotation speed and the second rotation speed are weighted and fused to obtain the target rotation speed of the capsule; The step of weightedly fusing the first rotational speed and the second rotational speed to obtain the target rotational speed of the capsule includes: determining a first weighted rotational speed based on the product of C_photo and the first rotational speed; determining a second weighted rotational speed based on the product of C_audio and the second rotational speed; and determining the target rotational speed by taking the quotient of the sum of the first weighted rotational speed and the second weighted rotational speed and the sum of C_photo and C_audio.

4. The method according to claim 3, characterized in that, The determination of the capsule's rotational state based on the confidence scores of the photoelectric sensor and the acoustic sensor, according to the first information and / or the second information, includes: If the difference between C_photo and C_audio is less than the second difference threshold, and the first rotation judgment result in the first information is inconsistent with the second rotation judgment result in the second information, then the confidence levels of the photoelectric sensor and the acoustic sensor for rotation and non-rotation are determined according to the first rotation judgment result and C_photo in the first information, and the second rotation judgment result and C_audio in the second information, respectively. Whether the capsule rotates is determined by trust level fusion.

5. The method according to claim 4, characterized in that, The step of determining the confidence levels of the photoelectric sensor and the acoustic sensor for rotation and non-rotation based on the first rotation judgment result and C_photo in the first information, and the second rotation judgment result and C_audio in the second information, respectively, includes: If the first rotation determination result is rotation, then the confidence level of the photoelectric sensor for rotation is determined according to C_photo, and the confidence level of the photoelectric sensor for no rotation is determined according to the confidence level of the photoelectric sensor for rotation; if the first rotation determination result is no rotation, then the confidence level of the photoelectric sensor for no rotation is determined according to C_photo, and the confidence level of the photoelectric sensor for rotation is determined according to the confidence level of the photoelectric sensor for no rotation; the sum of the confidence level of the photoelectric sensor for rotation and the confidence level of the photoelectric sensor for no rotation is equal to a set value; If the second rotation determination result is rotation, then the acoustic sensor's confidence level for rotation is determined based on C_audio, and the acoustic sensor's confidence level for non-rotation is determined based on the acoustic sensor's confidence level for rotation; if the second rotation determination result is non-rotation, then the acoustic sensor's confidence level for non-rotation is determined based on C_audio, and the acoustic sensor's confidence level for rotation is determined based on the acoustic sensor's confidence level for non-rotation; the sum of the acoustic sensor's confidence level for rotation and the acoustic sensor's confidence level for non-rotation is equal to the set value.

6. The method according to claim 4, characterized in that, The step of determining whether the capsule is rotated through trust fusion includes: The rotation confidence product is determined by multiplying the confidence level of the photoelectric sensor for rotation by the confidence level of the acoustic sensor for rotation. The product of the confidence level of the photoelectric sensor and the confidence level of the acoustic sensor regarding the non-rotation is determined. The product of the rotation confidence level and the product of the non-rotation confidence level are compared, and the capsule is determined to be rotated based on the comparison result.

7. The method according to claim 3, characterized in that, Before obtaining the second information, the method further includes: Before inhalation begins, ambient noise is collected to determine the mean and standard deviation of the ambient noise. During inhalation, a dynamic threshold is calculated in real time based on the mean and standard deviation of the ambient noise and the peak value of the real-time collected sound signal. The sound signal is filtered using the dynamic threshold to remove noise components below the dynamic threshold. The second information is determined based on the filtered sound signal.

8. A medication monitoring system for a capsule inhaler, characterized in that, Includes the medication monitoring device for the capsule inhaler as described in claim 1 and the capsule inhaler.

Images