Exhaled acetone multi-sensor response adaptive detection method and device

By defining an effective detection window during exhalation, suppressing common-mode responses during exhalation, and eliminating transient anomalies, the problem of aliasing in multi-sensor responses was solved, thereby improving the accuracy and stability of acetone detection in exhalation.

CN123058705APending Publication Date: 2026-11-10ZHONGSHAN HOSPITAL FUDAN UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611410101.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-09-11
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

Existing breath acetone detection devices are prone to aliasing of multi-sensor responses under changes in exhalation state and transient disturbances, affecting detection accuracy and stability.

Method used

By acquiring gas sensor response, flow rate, and humidity data during exhalation, an effective detection window is determined, common-mode response during exhalation is suppressed, transient abnormal responses are excluded, and the acetone concentration in exhalation is identified using the differential response characteristics of continuous detection windows.

Benefits of technology

It improves the accuracy and stability of expiratory acetone detection and reduces the impact of changes in expiratory state and transient disturbances.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123058705A_ABST
    Figure CN123058705A_ABST
Patent Text Reader

Abstract

The application discloses a breath acetone multi-sensor response adaptive detection method and device, obtains multiple gas sensor responses, breath flow and humidity in a breath process, determines a response baseline according to a pre-detection response, and determines an effective breath detection window according to breath flow and humidity changes; calculates a relative response amount of each gas sensor relative to the response baseline, determines a breath state common mode response according to synchronous changes of multiple gas sensors and time corresponding relationships between the synchronous changes and breath flow or humidity changes; suppresses the common mode response to form a difference response feature, excludes transient abnormal responses through difference response consistency between continuous detection windows, obtains an effective difference response feature, and determines a breath acetone concentration according to the effective difference response feature. The application can reduce multi-sensor response aliasing caused by breath state changes and transient disturbances, and improve breath acetone detection accuracy and stability.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of gas detection and exhaled breath component analysis technology, specifically to a multi-sensor adaptive detection method and device for exhaled acetone. Background Technology

[0002] Exhaled breath contains various volatile organic compounds, among which acetone, a volatile component produced during human metabolism, can be non-invasively detected using exhaled breath sampling and gas sensing technology. Existing acetone detection devices typically employ a single gas sensor or a sensor array consisting of multiple gas sensors with cross-sensitivity characteristics. The acetone concentration is determined by collecting the response signals of each sensor to exhaled breath and combining this with signal processing or pattern recognition methods. Compared to a single sensor, a multi-sensor array can provide different dimensions of gas response, but its detection results are easily affected by the actual exhalation state.

[0003] During an exhalation, the expiratory flow rate and humidity continuously change during the initiation, stabilization, and end of the exhalation. In addition to responding to acetone, multiple cross-sensitive gas sensors are also affected by changes in humidity and other volatile gases.

[0004] When the aforementioned factors act on multiple sensors simultaneously, they can easily lead to responses with similar directions or synchronized times, causing the effective response induced by acetone to overlap with the shared response induced by the exhalation state. Existing methods that employ temperature and humidity compensation or directly input the sensor array response into a pattern recognition model struggle to further distinguish between the shared components of the exhalation state and the differential components with gas compositional significance in the multi-sensor responses. Furthermore, transient flow rate changes during exhalation can cause localized anomalous responses, further affecting the consistency of results from different tests.

[0005] In view of this, the present invention provides a method and apparatus for adaptive detection of exhaled acetone using multiple sensors, thereby solving the above problems. Summary of the Invention

[0006] The purpose of this invention is to provide a method and apparatus for adaptive detection of exhaled acetone using multiple sensors, which addresses the technical problem of aliasing of multi-sensor responses caused by changes in exhalation state and transient disturbances, thereby affecting the accuracy and stability of exhaled acetone detection.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] In a first aspect, the present invention provides a multi-sensor adaptive detection method for exhaled acetone, comprising the following steps:

[0009] The response of multiple gas sensors with different cross-sensitivity characteristics during a single exhalation, as well as the corresponding exhalation flow rate and humidity, are obtained. The pre-detection response baseline of each gas sensor is determined based on the gas sensor response before the exhaled gas stably enters the detection area.

[0010] Candidate detection windows are determined based on changes in expiratory flow and humidity. The time interval in which the fluctuation of expiratory flow and humidity both meet the preset stability conditions is determined as the effective expiratory detection window. The effective expiratory detection window is divided into multiple consecutive detection windows.

[0011] The relative response quantity is determined by comparing the gas sensor response in each detection window with the corresponding pre-detection response baseline. The common mode response of the expiratory state is determined by the synchronous change of the relative response quantities of multiple gas sensors in the same direction and their time correspondence with the change of expiratory flow or humidity.

[0012] Based on the degree of response of each gas sensor to the common mode response of the exhalation state, the common mode response component is suppressed from the corresponding relative response quantity to form the differential response characteristics of each detection window;

[0013] Compare the differential response features of consecutive detection windows, exclude candidate transient abnormal responses that have isolated changes relative to the preceding and following detection windows, and determine the effective differential response features based on the differential responses of consecutive detection windows after exclusion.

[0014] Based on the pre-established correspondence between effective differential response characteristics and acetone concentration, the acetone concentration in exhaled breath is determined and the detection results are output.

[0015] Furthermore, response data from each gas sensor at multiple consecutive sampling times before the exhaled gas stably enters the detection area are acquired. When the variation range of the response data is within a preset stable range, the corresponding pre-detection response baseline is determined based on the statistical value of the response data.

[0016] Furthermore, when the expiratory flow rate reaches the preset flow rate condition and the humidity changes accordingly as the exhaled air enters the detection area, a candidate detection window is determined; when the fluctuation of expiratory flow rate and humidity within the candidate detection window meets the preset stability condition, the candidate detection window is determined as a valid expiratory detection window, and the time intervals corresponding to expiratory interruption, abnormal decrease in expiratory flow rate, or abnormal change in humidity are excluded from the valid expiratory detection window.

[0017] Furthermore, the first The gas sensor in the first The relative response within each detection window can be expressed as:

[0018] ;

[0019] in, Let be the relative response of the i-th gas sensor in the k-th detection window. Indicates the first The gas sensor in the first Response statistics in each detection window Indicates the first Pre-detection response baseline of a gas sensor.

[0020] Furthermore, the direction and time of change of the relative response quantities of multiple gas sensors are compared, and the response changes with the same direction and corresponding time are determined as candidate common responses; when the candidate common responses have a time correspondence with the changes in expiratory flow rate or humidity, the common mode response of the expiratory state of the corresponding detection window is determined based on the candidate common responses.

[0021] Furthermore, the first common-mode response coefficient obtained from calibration is determined. The gas sensor in the first Differential response values ​​in each detection window:

[0022] ;

[0023] in, For the first The gas sensor in the first Differential response quantity in each detection window To correspond to the relative response quantity, For the first The common-mode response of the expiratory state corresponding to each detection window For the first The common-mode response coefficients of each gas sensor; the differential response values ​​of multiple gas sensors within the same detection window are combined to form the differential response characteristics.

[0024] Furthermore, the difference response quantities of the current detection window and the detection windows before and after it are compared; when the difference response quantity of the current detection window changes more than the preset conditions relative to the detection windows before and after it, but the difference response quantities of the detection windows before and after it do not change, the difference response of the current detection window is identified as a candidate transient abnormal response and excluded; the effective difference response characteristics are determined based on the difference responses of the continuous detection windows after the exclusion.

[0025] Furthermore, a feature mapping model is established based on the effective difference response characteristics corresponding to different known acetone concentrations. The effective difference response characteristics corresponding to the current exhalation process are input into the feature mapping model to determine the exhaled acetone concentration. When the number of effective exhalation detection windows does not meet the preset conditions or the difference response between consecutive detection windows does not meet the preset consistency conditions, the current concentration calculation value is not used as a valid detection result, and a re-detection prompt is output.

[0026] Secondly, the present invention provides a multi-sensor response adaptive detection device for exhaled acetone, including an exhaled breath acquisition module, a sensor detection module, a signal acquisition module, a processing module, and a result output module;

[0027] The exhalation collection module is used to deliver exhaled gas to the sensor detection module;

[0028] The sensor detection module includes multiple gas sensors, humidity sensors, and flow sensors with different cross-sensitivity characteristics, used to acquire gas sensor responses, humidity, and exhalation flow.

[0029] The signal acquisition module is used to acquire the gas sensor response, humidity, and exhalation flow rate according to a uniform sampling sequence.

[0030] The processing module is used to determine the pre-detection response baseline and effective exhalation detection window, determine the common mode response of the exhalation state based on the relative responses of multiple gas sensors and their time correspondence with changes in exhalation flow or humidity, suppress the formation of differential response features by the common mode response of the exhalation state, exclude candidate transient abnormal responses based on the differential response features of the continuous detection window, and determine the exhalation acetone concentration based on the obtained effective differential response features.

[0031] The result output module is used to output the acetone concentration in exhaled breath and the validity of the detection.

[0032] Furthermore, the exhalation acquisition module includes a mouthpiece, a one-way valve, and an air delivery tube. The one-way valve is used to limit the flow direction of the exhaled gas, and the air delivery tube is used to deliver the exhaled gas to the sensor detection module. The signal acquisition module includes an analog signal conditioning circuit and an analog-to-digital conversion circuit. The processing module includes a microprocessor and a storage unit. The storage unit is used to store the pre-detection response baseline, common-mode response coefficients, differential response features, and feature mapping models. The microprocessor is used to call the data in the storage unit to perform common-mode response identification, differential response extraction, cross-window consistency verification, and determination of exhaled acetone concentration.

[0033] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0034] This invention determines the effective exhalation detection window based on changes in exhalation flow and humidity, enabling comparison of the responses of multiple gas sensors under relatively stable exhalation conditions. This reduces interference from the start, end, and abnormal fluctuation phases of exhalation on the detection data. Furthermore, it determines the common-mode response of the exhalation state based on the synchronous changes in the relative responses of multiple gas sensors and their temporal correspondence with changes in exhalation flow or humidity. This distinguishes the sensor responses caused by changes in exhalation state from the differential responses between different sensor channels. By suppressing the common-mode response, differential response characteristics are obtained, thereby reducing the impact of common changes in exhalation state on acetone response identification. Further, the consistency of differential response characteristics between continuous detection windows is used to identify and exclude isolated transient abnormal responses, resulting in effective differential response characteristics for concentration determination with good continuity and consistency. Therefore, it reduces the aliasing of multi-sensor responses caused by changes in exhalation state and transient disturbances, improving the accuracy and stability of exhaled acetone detection results. Attached Figure Description

[0035] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the embodiments of the present invention will be briefly described below.

[0036] Figure 1 This is a schematic diagram of the overall structure of the breath acetone multi-sensor response adaptive detection device of the present invention;

[0037] Figure 2 This is a schematic flowchart of the breath acetone multi-sensor response adaptive detection method of the present invention;

[0038] Figure 3 This is a schematic diagram illustrating the signal processing, common-mode response identification, differential response extraction, and acetone concentration estimation processes of the multi-sensor response of the present invention.

[0039] Figure 4 This is a schematic diagram illustrating the temporal relationship between expiratory flow rate, humidity, and the responses of multiple gas sensors and the effective expiratory detection window during the exhalation process of this invention. Detailed Implementation

[0040] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. The described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit the scope of protection of the present invention.

[0041] This invention addresses the problem that during acetone detection in exhaled breath, changes in expiratory flow rate, humidity, and non-target volatile gases can simultaneously affect multiple gas sensors with cross-sensitivity characteristics, causing synchronous responses from different gas sensors related to the exhalation state. This results in an overlap between acetone responses and exhalation state disturbances. The invention synchronously collects expiratory flow rate, humidity, and the responses of multiple gas sensors during a single exhalation. An effective exhalation detection window is determined based on changes in expiratory flow rate and humidity. Within this effective detection window, the relative response of each gas sensor relative to the baseline before detection is determined. A common-mode response of the exhalation state is determined based on the synchronous changes among multiple gas sensors. By suppressing the common-mode response to form differential response characteristics, the consistency of differential response characteristics between consecutive detection windows is used for verification. The exhaled acetone concentration is determined based on the verified effective differential response characteristics.

[0042] Example 1

[0043] like Figure 2 As shown, this embodiment provides a multi-sensor adaptive detection method for exhaled acetone, including the following steps:

[0044] S101. Obtain breath test data

[0045] The subject exhales continuously through the mouthpiece, and the exhaled air enters the sensor detection area through a one-way valve and an air guide tube.

[0046] During a single exhalation, response data from multiple gas sensors are continuously acquired according to a uniform sampling sequence, and exhalation flow, temperature, and humidity are acquired simultaneously, so that the responses of each gas sensor correspond to the exhalation flow, temperature, and humidity according to the sampling time.

[0047] Multiple gas sensors have different cross-sensitivity characteristics, and each produces a different degree of response to acetone and other volatile gases in exhaled breath, thus forming multiple sensing response channels.

[0048] Among them, expiratory flow rate and humidity are used to characterize the expiratory state corresponding to the current sensor response, and temperature is collected synchronously as an auxiliary environmental parameter of the detection area; the subsequent determination of the effective expiratory detection window and the identification of the common mode response of the expiratory state are mainly based on expiratory flow rate and humidity.

[0049] S102. Determine the response baseline before detection.

[0050] Before the current exhaled gas stably enters the sensor detection area, the response data of each gas sensor is acquired at multiple consecutive sampling times.

[0051] When the variation range of the continuously sampled data is within a preset stable range, the pre-detection response baseline of each gas sensor is determined based on the statistical value of the corresponding sampled data.

[0052] For the A gas sensor, whose pre-detection response baseline is represented as... .

[0053] The pre-detection response baseline is used to characterize the initial response state of the corresponding gas sensor at the start of the current detection, so that subsequent sensing responses are based on the current actual initial state, reducing the impact of changes in the sensor's initial state at different detection times on the detection results.

[0054] S103. Determine the effective expiratory detection window based on changes in expiratory flow rate and humidity.

[0055] like Figure 4 As shown, a single exhalation process consists of the baseline phase, the expiratory rise phase, the effective exhalation detection phase, and the expiratory fall phase.

[0056] During the baseline phase, the expiratory flow rate is at its initial state, and the responses of each gas sensor remain near the pre-detection response baseline.

[0057] During the rising phase of exhalation, as exhaled air enters the detection area, the expiratory flow rate gradually increases, and humidity changes synchronously, causing the gas sensors to produce corresponding response changes. Because expiratory flow rate and humidity are still significantly changing during this phase, the corresponding sensor responses are not directly used as characteristics for stable acetone detection.

[0058] When the expiratory flow rate reaches the preset flow rate condition and enters a relatively stable state, and the humidity reaches a stable change state corresponding to the exhaled gas, the corresponding time interval is determined as the candidate detection window.

[0059] Further analysis is conducted on the fluctuations in expiratory flow rate and humidity within the candidate detection window to determine whether it meets the stability criteria. When both meet the corresponding stability criteria, the candidate detection window is determined as a valid expiratory detection window.

[0060] The time intervals corresponding to sudden interruption of exhalation, abnormal decrease in exhalation flow, or abnormal change in humidity are not considered as valid exhalation detection windows.

[0061] Therefore, subsequent multi-sensor response analysis is mainly based on Figure 4 The effective exhalation detection phase is conducted when both exhalation flow and humidity are in a relatively stable state. The effective exhalation detection window is divided into multiple time-continuous detection windows according to a preset time length, so that each detection window generates corresponding multi-sensor response data, which is used for subsequent common-mode response identification, differential response extraction, and cross-window consistency verification.

[0062] S104. Determine the relative response of each gas sensor.

[0063] Within the effective exhalation detection window, the relative response quantity is determined based on the change between the current response of each gas sensor and the corresponding pre-detection response baseline.

[0064] No. The gas sensor in the first The relative response within each detection window can be expressed as:

[0065] ;

[0066] in, Indicates the first The gas sensor in the first Response statistics in each detection window Indicates the first Pre-detection response baseline of a gas sensor.

[0067] When different gas sensors have different response scales, the relative response quantities are normalized to make the changes between different sensing response channels comparable.

[0068] The relative response values ​​of multiple gas sensors are combined to form the multi-sensor response vector corresponding to the current detection window: .

[0069] S105. Determine the common-mode response of expiratory state based on synchronous changes from multiple sensors.

[0070] Furthermore, such as Figure 3 As shown, synchronous change analysis was performed on multiple sensor responses after signal acquisition, filtering and noise reduction, data alignment, baseline determination before detection, and identification of the effective expiratory window.

[0071] By comparing the direction and time of change of the relative response quantities of multiple gas sensors within the current detection window, the response changes with the same direction and corresponding time are identified as candidate common responses.

[0072] The candidate common response is further compared with the time-corresponding changes in expiratory flow rate and humidity; when the candidate common response has a time-corresponding relationship with the state changes of expiratory flow rate or humidity, the common mode response of the current detection window is determined based on the response changes shared by multiple gas sensors.

[0073] The first The common-mode response of the expiratory state corresponding to each detection window is represented as follows: The common-mode response of the exhalation state is used to characterize the common response components generated by the combined effect of the current exhalation state change on multiple gas sensors, and is not directly used as a characteristic quantity for determining the acetone concentration.

[0074] Therefore, only response changes that simultaneously exhibit common changes among multiple sensing channels and have a time-corresponding relationship with changes in expiratory flow or humidity can be used to determine the common mode response in the expiratory state, thus avoiding the direct determination of the common mode component based solely on the response changes of a single gas sensor.

[0075] S106. Suppress common-mode response and form differential response characteristics.

[0076] Based on the determined common-mode response of the exhalation state, common-mode suppression is performed on the relative response of each gas sensor.

[0077] No. The gas sensor in the first The difference response in each detection window can be expressed as:

[0078] ;

[0079] in, Let be the common-mode response coefficient corresponding to the i-th gas sensor, used to characterize the degree of response of the gas sensor to the common-mode response in the exhalation state.

[0080] The common-mode response coefficient can be determined based on the response data of each gas sensor under different exhalation states obtained during the equipment calibration phase.

[0081] The differential response values ​​of multiple gas sensors after common-mode suppression are combined to form:

[0082] ;

[0083] in, For the first The difference response features corresponding to each detection window.

[0084] like Figure 3 As shown, through the above processing, the original response of the gas sensor first undergoes common-mode response identification and common-mode suppression in the exhalation state, and then forms a differential response feature for subsequent concentration estimation, instead of directly using the original sensor response as the basis for calculating acetone concentration.

[0085] S107. Perform cross-window consistency verification.

[0086] Will Figure 4 The effective exhalation detection phase shown is further divided into multiple time-continuous detection windows, and the differential response characteristics corresponding to each detection window are determined according to the above processing procedure.

[0087] For the preceding detection window, the kth detection window, and the following detection window that are sequentially connected in time, compare the difference in response quantities corresponding to the same sensing response channel.

[0088] When the differential response of a gas sensor in the k-th detection window changes beyond the preset condition relative to both the previous and subsequent detection windows, but the differential response of the gas sensor in the previous and subsequent detection windows does not change, it indicates that the change is mainly an isolated change in the current detection window, and it is identified as a candidate transient abnormal response.

[0089] For the candidate transient abnormal response, the corresponding differential response quantity or the corresponding detection window can be excluded so that the candidate transient abnormal response does not participate in the determination of the current valid differential response features.

[0090] After excluding candidate transient abnormal responses, the direction and magnitude of change of the remaining differential responses in multiple consecutive detection windows and the correspondence between the sensor response channels are compared. When the differential responses between consecutive detection windows meet the preset consistency condition, the effective differential response characteristics of the current exhalation process are determined based on the corresponding differential responses.

[0091] Therefore, occasional changes in response within a single detection window do not directly determine the final acetone concentration, but require consistency verification across consecutive detection windows.

[0092] S108. Determine the acetone concentration in exhaled breath and output the test results.

[0093] The acetone concentration corresponding to the current exhalation process is determined based on the effective difference response characteristics after cross-window consistency verification.

[0094] Specifically, a feature mapping model is established in advance using the multi-sensor differential response features obtained under different acetone concentration conditions, and the current effective differential response features are input into the feature mapping model. The current exhaled acetone concentration is obtained according to the correspondence between the differential response features and the acetone concentration.

[0095] When establishing the feature mapping model, the calibration data corresponding to different known acetone concentrations are processed in the following order: determining the response baseline before detection, screening the effective detection window, determining the relative response amount, identifying the common mode response in the expiratory state, suppressing the common mode response, and verifying the consistency across the window to obtain effective differential response features. The mapping relationship between the effective differential response features and the corresponding known acetone concentrations is then established.

[0096] The input to the feature mapping model is the effective differential response features after effective expiratory window limitation, pre-detection baseline correction, common-mode response suppression, and cross-window consistency verification, rather than the unprocessed raw gas sensor response.

[0097] When the number of effective exhalation detection windows meets the preset conditions, and the difference response between consecutive detection windows meets the preset consistency conditions, the exhalation acetone concentration obtained based on the effective difference response characteristics will be taken as the effective detection result, and the exhalation acetone concentration and detection validity status will be output.

[0098] If the number of valid exhalation detection windows does not meet the preset conditions, there is a significant interruption during exhalation, or the difference response between consecutive detection windows does not meet the preset consistency conditions, the current concentration calculation value will not be used as a valid detection result, and a prompt to re-detect will be output.

[0099] Therefore, this embodiment acquires exhaled breath detection data and determines the pre-detection response baseline of each gas sensor. It then filters effective exhaled breath detection windows based on changes in exhaled flow and humidity, and calculates the relative response of multiple gas sensors relative to the pre-detection response baseline within these effective windows. By analyzing the synchronous changes among multiple gas sensors and their temporal correspondence with changes in exhaled flow or humidity, it identifies common-mode responses in the exhaled state and suppresses these common-mode responses to obtain differential response characteristics that characterize the differences in responses from different sensor channels. Furthermore, it eliminates transient abnormal responses by comparing the consistency of differential response characteristics between consecutive detection windows, and determines the exhaled acetone concentration based on the verified effective differential response characteristics, thereby reducing the impact of exhaled state changes and transient disturbances on the exhaled acetone detection results.

[0100] Example 2

[0101] like Figure 1 As shown, this embodiment provides a breath acetone multi-sensor response adaptive detection device, including a breath acquisition module, a sensor detection module, a signal acquisition module, a processing module, and a result output module.

[0102] The exhalation sampling module includes a mouthpiece, a one-way valve, and an air delivery tube. The mouthpiece is used to receive the exhaled gas from the subject being tested; the one-way valve is used to limit the flow direction of the exhaled gas, reducing the backflow of outside air into the exhalation sampling path; and the air delivery tube is used to deliver the exhaled gas to the sensor detection module.

[0103] The sensor detection module includes multiple gas sensors, a temperature sensor, and a humidity sensor. The multiple gas sensors have different cross-sensitivity characteristics and each forms an independent sensing response channel, used to generate corresponding gas response signals after exhaled air enters the detection area.

[0104] Temperature and humidity sensors are used to acquire the temperature and humidity within the detection area corresponding to the current exhalation process, respectively. The exhalation acquisition pathway also includes a flow sensor to acquire the exhalation flow rate, enabling the processing module to determine [the appropriate parameters] based on changes in exhalation flow rate and humidity. Figure 4The effective exhalation detection window is shown.

[0105] Among them, the humidity sensor and the flow sensor are used to obtain the humidity and expiratory flow rate, respectively, which are used to determine the effective expiratory detection window and identify the common mode response of the expiratory state; the temperature sensor is used to obtain the auxiliary environmental parameters of the detection area.

[0106] The number of gas sensors is not limited to Figure 1 The three shown can be used to set up three or more gas sensors with different cross-sensitivity characteristics according to actual detection needs.

[0107] The signal acquisition module includes an analog signal conditioning circuit and an analog-to-digital conversion circuit. The analog signal conditioning circuit is used to condition the analog signals output by each sensor, and the analog-to-digital conversion circuit is used to convert the conditioned analog signals into digital signals and send the responses of multiple gas sensors, exhalation flow, temperature and humidity data to the processing module according to a unified sampling sequence.

[0108] The processing module includes a microprocessor and may further include a storage unit connected to the microprocessor. The storage unit is used to store the pre-detection response baseline, common-mode response coefficients, differential response features corresponding to the continuous detection window, and feature mapping model.

[0109] Specifically, the processing module determines the pre-detection response baseline of each gas sensor based on the collected data, and determines the effective exhalation detection window based on changes in exhalation flow and humidity. Within the effective exhalation detection window, it determines the relative response of each gas sensor relative to the pre-detection response baseline. Based on the synchronous changes in the relative response of multiple gas sensors with consistent direction and their time correspondence with changes in exhalation flow or humidity, it determines the common-mode response of the exhalation state. Based on the common-mode response coefficients corresponding to each gas sensor, it suppresses the common-mode response of the exhalation state to form differential response features. It compares the differential response features of continuous detection windows and excludes candidate transient abnormal responses, obtaining effective differential response features based on the excluded differential responses. The effective differential response features are input into a feature mapping model to determine the exhaled acetone concentration.

[0110] The result output module includes a display unit, a wireless communication unit, and a prompting unit. The display unit displays the acetone concentration in exhaled breath and the validity of the test; the wireless communication unit sends the test data to an external terminal; and the prompting unit outputs a prompt to retest when the test conditions do not meet preset requirements.

[0111] During continuous use, the pre-detection response baseline and common-mode response parameters of each gas sensor can be recorded when stable detection conditions are met. When multiple consecutive valid detections indicate that the pre-detection response baseline of a certain gas sensor is slowly changing, the corresponding baseline parameters used for subsequent detections are updated based on the data that meets stable detection conditions. Detection data with interrupted exhalation, insufficient effective exhalation detection window, or failed cross-window consistency verification are not used for updating the baseline parameters.

[0112] Thus, multi-sensor exhalation detection data with a unified time correspondence is obtained through the exhalation acquisition module, sensor detection module, and signal acquisition module. The processing module then performs the effective exhalation window determination, common mode response identification, differential response extraction, and cross-window consistency verification as described in Example 1. Finally, the exhalation acetone detection result is output through the result output module.

[0113] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A multi-sensor adaptive detection method for acetone in exhaled breath, characterized in that, Includes the following steps: The response of multiple gas sensors with different cross-sensitivity characteristics during a single exhalation, as well as the corresponding exhalation flow rate and humidity, are obtained. The pre-detection response baseline of each gas sensor is determined based on the gas sensor response before the exhaled gas stably enters the detection area. Candidate detection windows are determined based on changes in expiratory flow and humidity. The time interval in which the fluctuation of expiratory flow and humidity both meet the preset stability conditions is determined as the effective expiratory detection window. The effective expiratory detection window is divided into multiple consecutive detection windows. The relative response quantity is determined by comparing the gas sensor response in each detection window with the corresponding pre-detection response baseline. The common mode response of the expiratory state is determined by the synchronous change of the relative response quantities of multiple gas sensors in the same direction and their time correspondence with the change of expiratory flow or humidity. Based on the degree of response of each gas sensor to the common mode response of the exhalation state, the common mode response component is suppressed from the corresponding relative response quantity to form the differential response characteristics of each detection window; Compare the differential response features of consecutive detection windows, exclude candidate transient abnormal responses that have isolated changes relative to the preceding and following detection windows, and determine the effective differential response features based on the differential responses of consecutive detection windows after exclusion. Based on the pre-established correspondence between effective differential response characteristics and acetone concentration, the acetone concentration in exhaled breath is determined and the detection results are output.

2. The multi-sensor adaptive detection method for exhaled acetone according to claim 1, characterized in that, The system acquires response data from multiple consecutive sampling times of each gas sensor before the exhaled gas stably enters the detection area. When the variation range of the response data is within a preset stable range, the corresponding pre-detection response baseline is determined based on the statistical value of the response data.

3. The multi-sensor adaptive detection method for exhaled acetone according to claim 1, characterized in that, When the expiratory flow rate reaches the preset flow rate condition and the humidity changes accordingly as the exhaled air enters the detection area, a candidate detection window is determined. When the fluctuation of expiratory flow rate and humidity within the candidate detection window meets the preset stability condition, the candidate detection window is determined as a valid expiratory detection window, and the time intervals corresponding to expiratory interruption, abnormal decrease in expiratory flow rate, or abnormal change in humidity are excluded from the valid expiratory detection window.

4. The multi-sensor adaptive detection method for exhaled acetone according to claim 1, characterized in that, No. The gas sensor in the first The relative response within each detection window can be expressed as: ; in, Let be the relative response of the i-th gas sensor in the k-th detection window. Indicates the first The gas sensor in the first Response statistics in each detection window Indicates the first Pre-detection response baseline of a gas sensor.

5. The multi-sensor adaptive detection method for exhaled acetone according to claim 4, characterized in that, By comparing the direction and time of change of relative response quantities of multiple gas sensors, response changes with consistent direction and corresponding time are identified as candidate common responses. When the candidate common responses have a time correspondence with changes in expiratory flow rate or humidity, the common mode response of the expiratory state in the corresponding detection window is determined based on the candidate common responses.

6. The multi-sensor adaptive detection method for exhaled acetone according to claim 5, characterized in that, The first [condition] is determined based on the common-mode response coefficient obtained from calibration. The gas sensor in the first Differential response values ​​in each detection window: ; in, For the first The gas sensor in the first Differential response quantity in each detection window To correspond to the relative response quantity, For the first The common-mode response of the expiratory state corresponding to each detection window For the first The common-mode response coefficients of each gas sensor; the differential response values ​​of multiple gas sensors within the same detection window are combined to form the differential response characteristics.

7. The multi-sensor adaptive detection method for exhaled acetone according to claim 6, characterized in that, Compare the difference response values ​​of the current detection window with those of the detection windows before and after it; when the difference response value of the current detection window changes more than a preset condition relative to both the detection windows before and after it, but the difference response values ​​of the detection windows before and after it do not change, the difference response value of the current detection window is identified as a candidate transient abnormal response and excluded; the effective difference response characteristics are determined based on the difference responses of the continuous detection windows after the exclusion.

8. The multi-sensor adaptive detection method for exhaled acetone according to claim 7, characterized in that, A feature mapping model is established based on the effective difference response characteristics corresponding to different known acetone concentrations. The effective difference response characteristics corresponding to the current exhalation process are input into the feature mapping model to determine the exhaled acetone concentration. When the number of effective exhalation detection windows does not meet the preset conditions or the difference response between consecutive detection windows does not meet the preset consistency conditions, the current concentration calculation value is not used as a valid detection result, and a re-detection prompt is output.

9. A multi-sensor adaptive detection device for exhaled acetone, characterized in that, It includes a breath sampling module, a sensor detection module, a signal acquisition module, a processing module, and a result output module; The exhalation collection module is used to deliver exhaled gas to the sensor detection module; The sensor detection module includes multiple gas sensors, humidity sensors, and flow sensors with different cross-sensitivity characteristics, used to acquire gas sensor responses, humidity, and expiratory flow rate. The signal acquisition module is used to acquire the gas sensor response, humidity, and exhalation flow rate according to a uniform sampling sequence. The processing module is used to determine the pre-detection response baseline and effective exhalation detection window, determine the common mode response of the exhalation state based on the relative responses of multiple gas sensors and their time correspondence with changes in exhalation flow or humidity, suppress the formation of differential response features by the common mode response of the exhalation state, exclude candidate transient abnormal responses based on the differential response features of the continuous detection window, and determine the exhalation acetone concentration based on the obtained effective differential response features. The result output module is used to output the acetone concentration in exhaled breath and the validity of the detection.

10. The breath acetone multi-sensor response adaptive detection device according to claim 9, characterized in that, The exhalation acquisition module includes a mouthpiece, a one-way valve, and an air delivery tube. The one-way valve is used to limit the flow direction of exhaled gas, and the air delivery tube is used to deliver exhaled gas to the sensor detection module. The signal acquisition module includes an analog signal conditioning circuit and an analog-to-digital conversion circuit. The processing module includes a microprocessor and a storage unit. The storage unit is used to store the pre-detection response baseline, common-mode response coefficients, differential response characteristics, and feature mapping models. The microprocessor is used to call the data in the storage unit to perform common-mode response identification, differential response extraction, cross-window consistency verification, and determination of exhaled acetone concentration.