Respiratory data processing method and apparatus, and device and storage medium

Abstract

A respiratory data processing method and apparatus, and a device and a storage medium. The method comprises: on the basis of three-dimensional motion information of at least one observation site provided on the chest and / or abdomen of a target object, determining a first respiratory curve for reflecting a respiratory process of the target object (S101); determining the current respiratory phase of the target object on the basis of the first respiratory curve (S102); determining a current observation sequence on the basis of a prior probability corresponding to the respiratory phase (S103); and determining a respiratory stage corresponding to the current observation sequence in a hidden Markov model, wherein model parameters of the hidden Markov model are determined on the basis of a Baum-Welch algorithm (S104).

Description

Respiratory data processing method, device, equipment and storage medium

[0001] This application claims priority to the Chinese patent application filed with the China Patent Office on October 30, 2023 with application number 202311423799.0, and claims priority to the Chinese patent application filed with the China Patent Office on October 30, 2023 with application number 202311426661.6, the entire contents of the above applications are incorporated by reference into this application. Technical Field

[0002] The present application relates to the technical field of physiological data processing, for example, to a respiratory data processing method, apparatus, device and storage medium. Background Art

[0003] During percutaneous interventional procedures, a patient's steady breathing facilitates needle insertion, thereby improving the success rate of percutaneous interventions and avoiding medical errors. Surgical navigation systems focus solely on respiratory amplitude to determine puncture timing using respiratory gating, making it difficult to improve the accuracy of puncture strategies.

[0004] For thoracic and abdominal organs, respiratory motion can lead to uncertainty in the location of the target point and the position of the puncture needle relative to the tissue during percutaneous intervention, which poses a challenge to the accuracy of intervention. Among the different ways to control or compensate for respiratory motion, the respiratory gating method has lower requirements on the patient's lung function and cooperation. At the same time, it does not require complex preoperative and intraoperative imaging acquisition or marker implantation, does not introduce additional risks, and can be applied to most interventional scenarios. Among them, under normal breathing conditions, the respiratory muscles are in a relaxed state at the end of exhalation, and the movement caused by breathing is slow, which is an ideal stage for interventional surgery. Therefore, in order to improve the accuracy of interventional surgery, it is necessary to accurately estimate the patient's end-expiration stage.

[0005] Respiratory gating is an amplitude-gated method, meaning that gating is initiated when the amplitude of body surface motion falls within a specified threshold. However, with baseline drift of the respiratory signal or changes in breathing pattern, the respiratory amplitude may not fall within the gating threshold, preventing gating from being activated for extended periods. Alternatively, the respiratory amplitude may reach the gating threshold but not be in the final stage of expiration, indicating that tissue movement caused by respiration is rapid and unsuitable for intervention.

[0006] Summary of the Invention

[0007] The present application provides a respiratory data processing method, apparatus, device and storage medium to solve the problem of being unable to accurately estimate the respiratory stage.

[0008] According to one aspect of the present application, a method for processing respiratory data is provided, comprising:

[0009] determining a first respiratory curve reflecting a respiratory process of the target object based on three-dimensional motion information of at least one observation site set on the chest and / or abdomen of the target object;

[0010] determining a current respiratory phase of the target subject according to the first respiratory curve;

[0011] Determining a current observation sequence according to a priori probability corresponding to the current respiratory phase;

[0012] The breathing phase corresponding to the current observation sequence under a hidden Markov model is determined, wherein the model parameters of the hidden Markov model are determined based on a Baum-Welch algorithm.

[0013] According to another aspect of the present application, a respiratory data processing device is provided, comprising:

[0014] a first respiratory curve determining module configured to determine a first respiratory curve reflecting a respiratory process of the target object based on three-dimensional motion information of at least one observation site set on the chest and / or abdomen of the target object;

[0015] a respiratory phase module, configured to determine a current respiratory phase of the target object based on the first respiratory curve;

[0016] an observation sequence determination module, configured to determine a current observation sequence according to a priori probability corresponding to the current respiratory phase;

[0017] The estimation module is configured to determine a respiratory phase corresponding to a current observation sequence under a hidden Markov model, wherein model parameters of the hidden Markov model are determined based on a Baum-Welch algorithm.

[0018] According to another aspect of the present application, an electronic device is provided, comprising:

[0019] at least one processor; and

[0020] a memory communicatively connected to the at least one processor; wherein,

[0021] The memory stores a computer program that can be executed by the at least one processor. The computer program is executed by the at least one processor so that the at least one processor can perform the respiratory data processing method described in any embodiment of the present application.

[0022] According to another aspect of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the respiratory data processing method described in any embodiment of the present application when executed. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] FIG1 is a flow chart of a respiratory data processing method according to an embodiment of the present application;

[0024] FIG2 is a flowchart of another respiratory data processing method provided according to an embodiment of the present application;

[0025] FIG3 is a flow chart of another respiratory data processing method provided according to an embodiment of the present application;

[0026] FIG4 is a flow chart of another respiratory data processing method provided according to an embodiment of the present application;

[0027] FIG5 is a flow chart of a respiratory stage estimation method according to an embodiment of the present application;

[0028] FIG6A is a schematic diagram of an initial breathing curve provided according to an embodiment of the present application;

[0029] FIG6B is a schematic diagram of a first respiratory curve provided according to an embodiment of the present application;

[0030] FIG7A is a schematic structural diagram of a respiratory data processing device according to an embodiment of the present application;

[0031] FIG7B is a schematic structural diagram of another respiratory data processing device provided according to an embodiment of the present application;

[0032] FIG7C is a schematic structural diagram of another respiratory data processing device provided according to an embodiment of the present application;

[0033] FIG8 is a schematic structural diagram of another respiratory data processing device provided according to an embodiment of the present application;

[0034] FIG9 is a schematic structural diagram of another respiratory data processing device provided according to an embodiment of the present application;

[0035] FIG10 is a schematic diagram of the structure of an electronic device that implements the respiratory data processing method according to an embodiment of the present application. DETAILED DESCRIPTION

[0036] The technical solutions in the embodiments of the present application will be described below in conjunction with the drawings in the embodiments of the present application. The described embodiments are only part of the embodiments of the present application, rather than all the embodiments.

[0037] The terms "first", "second", etc. in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequential order. The numbers used in this way can be interchanged where appropriate, so that the embodiments of the application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units listed, but may include other steps or units that are not listed or that are inherent to these processes, methods, products or devices.

[0038] Figure 1 is a flowchart of a respiratory data processing method provided in an embodiment of the present application. This embodiment is applicable to determining a target subject's current respiratory stage based on their respiratory phase. The method can be performed by a respiratory data processing device, which can be implemented in hardware and / or software and configured in a processor of an electronic device. As shown in Figure 1, the method includes the following steps.

[0039] S101 : Determine a first respiratory curve reflecting a breathing process of a target object based on three-dimensional motion information of at least one observation site located on the chest and / or abdomen of the target object.

[0040] S102: Determine a current respiratory phase of the target object according to the first respiratory curve.

[0041] S103: Determine a current observation sequence according to the prior probability corresponding to the current respiratory phase.

[0042] S104 : Determine a respiratory phase corresponding to the current observation sequence under a hidden Markov model, wherein model parameters of the hidden Markov model are determined based on a Baum-Welch algorithm.

[0043] Figure 2 is a flow chart of another respiratory data processing method provided by an embodiment of the present application. As shown in Figure 2, the method includes the following steps.

[0044] S110 : Determine a current respiratory phase of the target object according to a first respiratory curve, where the first respiratory curve is determined based on three-dimensional motion information of at least one observation site set on the chest and / or abdomen of the target object.

[0045] Among them, the target object is the patient to be punctured.

[0046] In the first respiratory curve, the respiratory phase of a point can be understood as the phase angle of the point, such as a phase angle of 90 degrees or a phase angle of 0 degrees.

[0047] In one embodiment, at least one observation patch is placed on the chest surface of a target subject; three-dimensional position information of a corresponding body surface observation point at the current moment is collected via the at least one observation patch; a first respiratory curve reflecting the patient's respiratory motion is determined based on the at least one three-dimensional position information; and the target subject's current respiratory phase is determined based on the first respiratory curve. The first respiratory curve is drawn based on real-time data, and the target subject's current respiratory phase is determined based on the first respiratory curve, thereby achieving the technical effect of rapidly determining the target subject's current respiratory phase.

[0048] In one embodiment, a first respiratory curve is determined by the following steps: After determining an initial respiratory curve based on the at least one three-dimensional position information, the initial respiratory curve is preprocessed to obtain the first respiratory curve, such as by removing baseline drift, performing data dimensionality reduction, and performing data smoothing. The data dimensionality reduction includes obtaining a projection direction through principal component analysis, where the projection direction is the main motion direction.

[0049] In one embodiment, the first respiratory curve is used to reflect changes in the motion component of the target subject's respiratory motion in a principal motion direction, where the principal motion direction is the direction of the covariance matrix eigenvector corresponding to the first principal component of the respiratory motion signal. Because the motion component of the respiratory motion in the principal motion direction accurately reflects the patient's respiratory process, the target subject's respiratory phase at the current moment can be accurately determined based on the first respiratory curve.

[0050] S120. Determine a current observation sequence according to the prior probability corresponding to the current respiratory phase.

[0051] The embodiment of the present application estimates the respiratory phase of the patient at multiple moments based on a hidden Markov model.

[0052] Among them, the current observation sequence is based on the normal distribution φ(O|μ Oj ,σ Oj ) and the prior probability φ(θ based on the respiratory phase t |μ θi ,σ θi ) Determine the current observation sequence, the observation items in the observation sequence are:

[0053] Among them, μ θi ,σ θi is an empirical value, i is greater than or equal to 1 and less than or equal to the total number of breathing stages, for example, 1≤i≤4. The observation sequence is recorded as B, which can be expressed as, B={μ Oj ,σ Oj}, where j is greater than or equal to 1 and less than or equal to the total number of breathing stages, such as 1≤j≤4.

[0054] Where π is the initial state probability. When the total number of breathing stages is 4, π can be expressed as: π = {π1, π2, π3, π4}, where π j =P{q1=S j},1≤j≤4.

[0055] S130 : Determine a breathing phase corresponding to the current observation sequence under a hidden Markov model, wherein model parameters of the hidden Markov model are determined based on a Baum-Welch algorithm.

[0056] The model parameters of a hidden Markov model can be expressed as λ = {A, B, π}. The Baum-Welch algorithm is an expectation maximization (Em) algorithm used to solve the learning problem of a hidden Markov model. It is an unsupervised learning method. The hidden Markov model can be used to estimate the inspiratory phase, the end-inspiratory phase, the expiratory phase, and the end-expiratory phase.

[0057] The respiratory phase is set as a hidden state. Accordingly, the hidden sequence of the hidden Markov model includes three or four respiratory phase identifiers. The respiratory phase identifiers included in the hidden sequence are related to the division method of the respiratory cycle.

[0058] In one embodiment, a complete respiratory cycle can be divided into four phases, namely, an exhalation phase, a terminal exhalation phase, an inhalation phase, and a terminal inhalation phase; which can be represented as S={S1, S2, S3, S4}.

[0059] In one embodiment, a complete respiratory cycle can be divided into three phases, namely, an exhalation phase, a terminal exhalation phase, and an inhalation phase.

[0060] Assume that the hidden states are converted in the order of S1→S2→S3→S4→S1→…; the state transition probability matrix between hidden states can be defined as:

[0061] Among them, a ij =P{q t+1 =S j |q t =S i},1≤i,j≤4, that is, from state S i Transfer to state S j probability.

[0062] After the model parameters of the hidden Markov model are determined, the breathing phase corresponding to the current observation sequence under the hidden Markov model is determined based on the Viterbi algorithm. The Viterbi algorithm is an algorithm in related art that actually uses dynamic programming to solve the prediction problem of the hidden Markov model.

[0063] If the hidden sequence includes four breathing stage identifiers, the corresponding hidden Markov model can be used to estimate the four breathing stages; if the hidden sequence includes three breathing stage identifiers, the corresponding hidden Markov model can be used to estimate the three breathing stages.

[0064] The breathing process of different patients is different. For example, different patients have different inhalation processes and / or different exhalation processes. Reflected in the respiratory curve, the time span of the partial respiratory curve of the exhalation phase of some patients is relatively large, while the time span of the partial respiratory curve of the exhalation phase of some patients is relatively small. Even if the time span of the partial respiratory curve of the exhalation phase of different patients is the same, some patients may begin to enter the end-expiration phase at phase C, while some patients may begin to enter the end-expiration phase at phase D, and phase D is greater than phase C. Therefore, compared to determining the end-expiration phase based only on a fixed amplitude threshold or phase threshold, the present application determines the model parameters of the hidden Markov model through the Baum-Welch algorithm, and establishes a hidden correspondence between the observation sequence and the respiratory phase for each patient; since the observation sequence is determined based on the prior probability of the respiratory phase, it can be regarded as establishing a correspondence between the respiratory phase and the respiratory phase for each patient, thereby achieving the technical effect of accurately estimating the patient's respiratory phase based on the respiratory phase.

[0065] S140: When the breathing phase is the target phase, perform a setting operation.

[0066] Among them, the target stage can be set based on the actual usage scenario.

[0067] During normal breathing, the respiratory muscles are in a relaxed state at the end of expiration, and breathing-induced movement slows, making it an ideal stage for interventional surgery. Therefore, in this embodiment, the target stage can be selected as the end of expiration. The setting operation includes outputting a prompt message or outputting a control signal to the puncture device, the control signal being used to control the puncture device to perform the set puncture operation. If the estimated respiratory stage is the end of expiration, a prompt message is output or a control signal is output to the puncture device, the control signal being used to control the puncture device to perform the set puncture operation.

[0068] In one embodiment, prompt information is output in a visual interface so that the user can perform a thoracic and abdominal puncture operation according to the prompt information, thereby improving the user's puncture accuracy. The prompt information can be text prompt information, respiratory curve identification information, and respiratory animation identification information. The respiratory curve is a curve corresponding to the estimated respiratory stage for displaying the respiratory process, and the respiratory curve identification information is a set identification set on the end-expiratory stage of the respiratory curve, such as a color identification or a line identification. The respiratory animation is used to display the patient's respiratory process, and the respiratory animation identification information can be set to color identification information during the animation display process. For example, at the end-expiratory stage, the little man used to represent the patient is set to green, and in other respiratory stages, the little man used to represent the patient is set to other colors. This embodiment is suitable for manual puncture scenarios.

[0069] In one embodiment, when the target phase is the end-expiratory phase, a control signal is generated to control the puncture device to perform the puncture operation along a set path. The set path can be determined based on clinical medical images. This embodiment is suitable for automatic puncture scenarios.

[0070] The technical solution of the embodiment of the present application is that since the model parameters of the hidden Markov model are determined based on the Baum-Welch algorithm, the determination of the model parameters can be completed quickly; since the first respiratory curve is used to reflect the change of the motion component of the respiratory motion of the target object in the main motion direction, the respiratory phase of the target object at the current moment can be accurately determined according to the first respiratory curve, and the current observation sequence can be accurately determined according to the prior probability corresponding to the respiratory phase; thereby accurately determining the respiratory stage corresponding to the current observation sequence under the hidden Markov model, achieving the technical effect of accurately estimating the patient's respiratory stage at the current moment based on the respiratory phase and the hidden correspondence between the respiratory phase and the respiratory stage unique to the target object.

[0071] FIG3 is another flow chart of a respiratory data processing method according to an embodiment of the present application, which adds a step of drawing a second respiratory curve based on the previous embodiment. As shown in FIG2 , the method includes the following steps.

[0072] S210 : Determine a current respiratory phase of the target object according to a first respiratory curve, where the first respiratory curve is determined based on three-dimensional motion information of at least one observation site set on the chest and / or abdomen of the target object.

[0073] S220: Determine a current observation sequence according to the prior probability corresponding to the respiratory phase.

[0074] S230 : Determine a breathing phase corresponding to the current observation sequence under a hidden Markov model, wherein model parameters of the hidden Markov model are determined based on a Baum-Welch algorithm.

[0075] S240: When the breathing phase is the target phase, perform a setting operation.

[0076] S2501. Determine the average change rate of the partial respiratory curve of the respiratory stage at the current moment in the first respiratory curve, the change rate at the previous moment, and the difference between the amplitude at the current moment and the amplitude at the previous moment.

[0077] The partial respiratory curve of the respiratory stage at the current moment may include only one plotting point or may include at least two plotting points. Taking the end-expiratory stage as an example, if the current moment is the beginning of the end-expiratory stage, then the partial first respiratory curve of the respiratory stage at the current moment includes only one plotting point. If the current moment is not the beginning of the end-expiratory stage, then the partial first respiratory curve of the respiratory stage at the current moment includes at least two plotting points. In the case where the partial respiratory curve of the respiratory stage at the current moment includes at least two plotting points, the average of the change rates of the at least two plotting points is used as the average of the change rates of the partial respiratory curve of the respiratory stage at the current moment. Wherein, the change rate of at least one plotting point is determined based on the differential result of at least one plotting point.

[0078] S2502: Determine a change rate of the second respiratory curve at the current moment according to the change rate mean, the change rate at the previous moment, and the difference.

[0079] In one embodiment, a weighted average is performed on the mean change rate, the change rate at the previous moment, and the difference to obtain a weighted average result; the weighted average result is used as the change rate of the second respiratory curve at the current moment. This embodiment improves the accuracy of determining the change rate of the second respiratory curve at the current moment by using the weighted average of the mean change rate, the change rate at the previous moment, and the difference as the change rate of the second respiratory curve at the current moment.

[0080] Among them, the weighted average process can be expressed as,

[0081] Among them, μ Oj is the mean of the above change rates, P 1j is the weight of the mean value of the change rate; is the rate of change at the previous moment, P 2j is the weight of the rate of change at the previous moment; is the above difference, A t is the amplitude at the current moment; is the amplitude of the previous moment; P 3j is the weight of the difference.

[0082] S2503: Update the second breathing curve according to the change rate of the second breathing curve at the current moment.

[0083] The second respiratory curve is a respiratory curve presented to the user to assist the user in making medical decisions.

[0084] In one embodiment, the updated second breathing curve is displayed in the visual interactive interface, so that the user can understand the breathing state of the target object through the second breathing curve.

[0085] In one embodiment, when the updated second respiratory curve displayed in the visual interactive interface includes the end-expiration stage, the portion of the second respiratory curve corresponding to the end-expiration stage is provided with marking information, which is color marking information or line marking information.

[0086] The embodiment of the present application determines the change rate of the second respiratory curve at the current moment by taking a weighted average of the change rate mean of the partial respiratory curve in the respiratory stage at the current moment in the first respiratory curve, the change rate at the previous moment, and the difference between the amplitude at the current moment and the amplitude at the previous moment, thereby improving the accuracy of the change rate determination, thereby improving the accuracy of the second respiratory curve updated based on the change rate, and the accuracy of the medical treatment strategy determined by the user based on the second respiratory curve.

[0087] Figure 4 is a flow chart of a respiratory data processing method provided in an embodiment of the present application. This embodiment is applicable to automatically detecting whether a target subject's breathing is regular. The method can be performed by a respiratory data processing device, which can be implemented in hardware and / or software and can be configured in a processor of an electronic device. As shown in Figure 4, the method includes the following steps.

[0088] S310: Determine a first breathing curve for reflecting the breathing process of the target object.

[0089] Among them, the target object is the patient to be punctured.

[0090] In one embodiment, the first respiratory curve is used to reflect changes in the motion component of the target object's respiratory motion in a main motion direction, where the main motion direction is the direction of the covariance matrix eigenvector corresponding to the first principal component of the respiratory motion signal.

[0091] In one embodiment, at least one observation patch is set on the chest surface of the target object; three-dimensional position information of the corresponding body surface observation point at the current moment is collected through the at least one observation patch; and a first respiratory curve reflecting the patient's respiratory movement is determined based on the at least one three-dimensional position information. Exemplarily, after determining the initial respiratory curve based on the at least one three-dimensional position information, the initial respiratory curve is preprocessed to obtain the first respiratory curve, such as removing baseline drift, data dimensionality reduction, and data smoothing. The data dimensionality reduction includes obtaining a projection direction through principal component analysis, and the projection direction is the main motion direction.

[0092] S320: If it is determined that the first respiratory curve has the same first abnormal event N times in a row, output first prompt information indicating abnormal respiratory regularity, where the first abnormal event is abnormal amplitude, abnormal periodicity, or abnormal duration of a set respiratory phase.

[0093] Wherein, N is greater than or equal to 2, the periodic anomaly is determined based on serial correlation, the amplitude anomaly is determined based on the maximum or minimum value of the target breathing phase, the set breathing phase and the target breathing phase are determined based on a hidden Markov model, and the model parameters of the hidden Markov model are determined based on a Baum-Welch algorithm.

[0094] A normal respiratory signal is regular, for example, peaks repeat periodically and multiple respiratory phases have roughly the same duration. A first abnormal event refers to an event related to the regularity of the target subject's breathing, such as abnormal amplitude, abnormal periodicity, or abnormal duration of a set respiratory phase.

[0095] The periodic abnormality means that the respiratory curve corresponding to the respiratory movement of the target object is no longer periodic.

[0096] The abnormal duration of the set breathing phase refers to that the duration of the set breathing phase is too long or too short compared with the previous breathing cycle.

[0097] In one embodiment, if a target object has a maximum amplitude anomaly or a minimum amplitude anomaly, it is considered that the target object has an amplitude anomaly.

[0098] Exemplarily, the following steps are used to determine whether the first respiratory curve has an abnormal amplitude:

[0099] Step a1: performing segmentation processing on the first respiratory curve based on a first segmentation duration to obtain a segmentation result, wherein the first segmentation duration is greater than 0.5 respiratory cycles and less than one respiratory cycle.

[0100] In one embodiment, the duration of the first segment is the window width of the rolling window.

[0101] Step a2: If the segmented result includes the end of inspiration or the end of expiration, and the duration of the end of inspiration is greater than half of the duration of the entire end of inspiration, and the duration of the end of expiration is greater than half of the duration of the entire end of expiration, then extract the maximum or minimum value of the segmented result.

[0102] Since the end-inhalation process corresponds to the maximum amplitude and the end-expiration process corresponds to the minimum amplitude, if the segmentation result includes the end-inhalation stage, the maximum value is extracted from the segmentation result, and the maximum value is the maximum amplitude; if the segmentation result includes the end-expiration stage, the minimum value is extracted from the segmentation result, and the minimum value is the minimum amplitude.

[0103] Step a3: If the maximum value or the minimum value does not meet the set amplitude condition with the maximum value or the minimum value of the set period, the maximum value or the minimum value is marked as an amplitude abnormality.

[0104] The set period may be a reference period or a previous period. The reference period is a period selected by the user from the first respiratory curve, or any period within a period with good repeatability in the first respiratory curve.

[0105] In one embodiment, if the current maximum value and the maximum value of the previous cycle do not meet the first set amplitude condition, the maximum value is marked as a maximum amplitude anomaly; if the current minimum value and the minimum value of the previous cycle do not meet the second set amplitude condition, the minimum value is marked as a minimum amplitude anomaly.

[0106] The first set amplitude condition and the second set amplitude condition can be difference amplitude conditions or ratio amplitude conditions. If the first set amplitude condition is a difference amplitude condition, it is necessary to determine whether the difference between the current maximum value and the maximum value of the previous cycle meets the difference amplitude condition. If the first set amplitude condition is a ratio amplitude condition, it is necessary to determine whether the ratio between the current maximum value and the maximum value of the previous cycle meets the corresponding ratio amplitude condition.

[0107] Step a4: If the amplitude abnormality occurs N times consecutively, it is determined that the first respiratory curve has an amplitude abnormality.

[0108] Here, N can be set to 2, 3 or 4 based on experience.

[0109] Taking N as 3 as an example, if the first respiratory curve has abnormal amplitude three times in a row, it is determined that the first respiratory curve has abnormal amplitude.

[0110] In one embodiment, determining whether the first respiratory curve has periodic abnormality is performed by the following steps:

[0111] Step b1: determining a correlation measure between the current respiratory curve segment and a set respiratory curve segment based on an autocorrelation operation when a current respiratory curve segment is determined on the first respiratory curve based on a second segment duration, wherein the second segment duration is greater than one respiratory cycle and less than two respiratory cycles.

[0112] The respiratory curve segment is set to be a reference respiratory curve segment or a previous respiratory curve segment. The reference respiratory curve segment is a respiratory curve segment selected by the user from the first respiratory curve, or any respiratory curve segment in a period with good periodicity in the first respiratory curve.

[0113] Autocorrelation, also known as serial correlation, is the cross-correlation between a signal and itself at different points in time. Informally speaking, autocorrelation compares two observations of the same signal at different times to assess their similarity. The autocorrelation function is the average of the product of a signal x(t) and its time-shifted counterpart x(t-τ). It is a function of the time-shift variable τ.

[0114] In one embodiment, the second segment duration may be 1.5 respiratory cycles.

[0115] b2. If the correlation metric value does not meet the set metric matching condition, the current respiratory curve segment is recorded as a repetitive anomaly.

[0116] Simply put, if the correlation metric value does not meet the set metric matching condition, it is considered that the similarity between the current respiratory curve segment and the previous respiratory curve segment is low, that is, the repeatability between the first respiratory curve in the previous respiratory curve segment and the current respiratory curve segment is poor, so the current respiratory curve segment is recorded as a repeatability anomaly.

[0117] Step b3: If the abnormal repeatability occurs N times consecutively, it is determined that the first respiratory curve has a periodic repeatability abnormality.

[0118] Taking N as 3 as an example, if the first respiratory curve has periodic repetitive abnormalities three times in a row, it is determined that the first respiratory curve has periodic abnormalities.

[0119] In one embodiment, the exhalation phase is set to the breathing phase expected by the user. For the chest and abdominal puncture scenario, the user expects the end of exhalation phase, so the exhalation phase is set to the end of exhalation phase.

[0120] In one embodiment, the abnormal duration of the end-expiratory phase is determined by the following steps:

[0121] Step c1: If the breathing stage corresponding to the current drawing point in the first breathing curve is the end-expiratory stage, the total number of drawing points corresponding to the end-expiratory stage is increased by 1 until the current breathing stage is the start moment of the inspiratory stage.

[0122] During the breathing process, the inhalation phase automatically begins after the end of the exhalation phase. Therefore, this embodiment ends the accumulation of the total number of plotted points corresponding to the end of the exhalation phase when the start of the inhalation phase is detected. This step is intended to count the total number of plotted points corresponding to the latest end of the exhalation phase.

[0123] Step c2: determining the duration of the end-expiration stage according to the total number of plotted points corresponding to the end-expiration stage.

[0124] Since the plotting points of the first respiratory curve are collected based on a set time interval, the duration corresponding to the end-expiration stage can be determined by counting the total number of plotting points corresponding to the end-expiration stage.

[0125] Step c3: If the duration of the end-expiratory phase is abnormal, mark the end-expiratory phase as an abnormal end-expiratory phase.

[0126] If the duration of the end-expiratory phase does not meet the set duration condition, the current end-expiratory phase is marked as an abnormal end-expiratory phase. In one embodiment, if the difference between the duration of the end-expiratory phase and the target duration is greater than a set duration threshold, the current end-expiratory phase is marked as an abnormal end-expiratory phase. The target duration is usually different for different target subjects, so before performing breathing pattern detection on a target subject, the target duration for that target subject must be determined.

[0127] In one embodiment, multiple breathing process data of the target object can be collected in advance, and an initial breathing curve can be determined based on the multiple breathing process data. The duration distribution of the target object's end-expiration stage can be determined based on the initial breathing curve, and the target duration can be determined based on the duration distribution of the end-expiration stage. For example, the singular duration can be eliminated from the duration distribution result of the end-expiration stage, and the mean of the remaining durations of the end-expiration stage can be used as the target duration. Alternatively, the initial exhalation curve can be analyzed using a set model to obtain the duration of the target object's end-expiration stage.

[0128] Step c4: If the abnormal end-expiratory phase occurs N times consecutively, it is determined that the duration of the end-expiratory phase in the first respiratory curve is abnormal.

[0129] Taking N as 3 as an example, if the first respiratory curve has abnormal duration of the end-expiratory phase three times in a row, it is determined that the first respiratory curve has abnormal duration of the end-expiratory phase.

[0130] In one embodiment, when the first prompt information is detected, the hidden Markov model is controlled to enter a learning phase; when it is detected that the learning of the hidden Markov model is completed, the process returns to the step of outputting the first prompt information indicating abnormal breathing regularity if it is determined that the first abnormal event occurs N times in a row on the first breathing curve. That is to say, if abnormal breathing regularity is detected on the first breathing curve, the monitoring of breathing regularity is stopped, and the hidden Markov model is re-learned. After the learning is completed, the monitoring of breathing regularity is re-performed. This embodiment is suitable for when the patient's state changes and a new prediction is required, such as when the patient changes from a tense state to a relaxed state, or vice versa.

[0131] In one embodiment, when a baseline drift abnormality is determined according to the first respiratory curve, second prompt information indicating the movement of the target object is output.

[0132] In one embodiment, upon detecting a second prompt, the hidden Markov model is controlled to enter a learning phase; upon detecting that the hidden Markov model has completed learning, the current hidden Markov model is used to monitor respiratory regularity. That is, if the second prompt is detected, it indicates that the target subject has moved, at which point respiratory regularity monitoring must be stopped, and the hidden Markov model must be relearned to determine new baseline drift data. After learning is complete, respiratory regularity monitoring is resumed using the current hidden Markov model. This embodiment is suitable for situations where the patient's chest or abdomen is displaced.

[0133] The technical solution provided in the embodiment of the present application realizes unsupervised learning of the hidden Markov model through the Baum-Welch algorithm, realizes tailor-made hidden Markov model for each target object, and estimates the set exhalation phase and the target exhalation phase through the combination of the hidden Markov model and the first respiratory curve, thereby improving the accuracy of the respiratory phase estimation, the accuracy of the amplitude abnormality determined based on the maximum or minimum value of the target respiratory phase, and the accuracy of the abnormality of the duration of the set respiratory phase, thereby improving the accuracy of respiratory regularity monitoring.

[0134] Figure 5 is a flow chart of a method for estimating respiratory stages according to an embodiment of the present application, which is used to refine the steps of estimating respiratory stages based on a hidden Markov model according to the above embodiment.

[0135] S410: Determine an observation sequence corresponding to each drawing point of the segmentation result, where the observation sequence includes a respiratory phase corresponding to each drawing point.

[0136] Among them, the drawing points can be understood as sampling points.

[0137] In the first respiratory curve, the respiratory phase of a point can be understood as the phase angle of the point, such as a phase angle of 90 degrees or a phase angle of 0 degrees.

[0138] In one embodiment, at least one observation patch is set on the chest surface of the target object; three-dimensional position information of the corresponding body surface observation point at the current moment is collected through the at least one observation patch; based on the at least one three-dimensional position information, a first respiratory curve for reflecting the patient's respiratory movement is determined; and the respiratory phase corresponding to the drawn point in the first respiratory curve is determined.

[0139] In one embodiment, a first respiratory curve is determined by the following steps: Based on the at least one three-dimensional position information, an initial respiratory curve is determined (see FIG. 6A ), and the initial respiratory curve is preprocessed to obtain the first respiratory curve, such as removing low-frequency baseline drift, performing data dimensionality reduction, and performing data smoothing. The data dimensionality reduction includes obtaining a projection direction by principal component analysis, where the projection direction is the main motion direction.

[0140] In one embodiment, to remove low-frequency baseline drift, the following steps are performed on multiple-dimensional position information sequences in an initial respiratory curve: morphological dilation and morphological erosion are performed on the initial respiratory curve using a structuring element with a length greater than one respiratory cycle, and the average of the dilation and erosion results is used as the low-frequency baseline drift; and the low-frequency baseline drift is removed from the initial respiratory curve. The structuring elements corresponding to the multiple-dimensional position information sequences are the same.

[0141] In one embodiment, the data smoothing process is a projection signal trajectory smoothing process. A moving window smoothing process can be used. To avoid the phase difference caused by the window length L, the data of the K actual plotting points after the current respiratory phase in the previous respiratory cycle in the first respiratory curve are used to predict the data of the future K predicted plotting points. The data smoothing process result is determined based on the data of the K actual plotting points and the data of the K predicted plotting points to update the first respiratory curve. Exemplarily, the corresponding weighted average result of the data of the K actual plotting points and the data of the K predicted plotting points is used as the data of the future K plotting points in the first respiratory curve to update the first respiratory curve (see Figure 6B). K is greater than or equal to half of the window width.

[0142] In one embodiment, the first respiratory curve is used to reflect changes in the motion component of the target subject's respiratory motion in a principal motion direction, where the principal motion direction is the direction of the covariance matrix eigenvector corresponding to the first principal component of the respiratory motion signal. Because the motion component of the respiratory motion in the principal motion direction accurately reflects the patient's respiratory process, the respiratory phase corresponding to each plotted point can be accurately determined based on the first respiratory curve.

[0143] Among them, the observation sequence corresponding to each plotted point is based on the normal distribution φ(O|μ Oj ,σ Oj ) and the prior probability φ(θ based on the respiratory phase t |μ θi ,σ θi ) is determined, and the observation items in the observation sequence are:

[0144] Among them, μ θi ,σ θi is an empirical value, i is greater than or equal to 1 and less than or equal to the total number of breathing stages, for example, 1≤i≤4. The observation sequence is recorded as B, which can be expressed as, B={μ Oj ,σ Oj}, where j is greater than or equal to 1 and less than or equal to the total number of breathing stages, such as 1≤j≤4.

[0145] Where π is the initial state probability. When the total number of breathing stages is 4, π can be expressed as: π = {π1, π2, π3, π4}, where π j =P{q1=S j},1≤j≤4.

[0146] The model parameters of a hidden Markov model can be expressed as λ = {A, B, π}. The Baum-Welch algorithm, also known as the Em algorithm, is an unsupervised learning method used to solve hidden Markov model learning problems. During the learning phase, the hidden Markov model requires learning the aforementioned model parameters.

[0147] The respiratory phase is set as a hidden state. Accordingly, the hidden sequence of the hidden Markov model includes three or four respiratory phase identifiers. The respiratory phase identifiers included in the hidden sequence are related to the division method of the respiratory cycle.

[0148] In one embodiment, a complete respiratory cycle can be divided into four phases, namely, the exhalation phase, the end-expiration phase, the inspiration phase and the end-inspiration phase; which can be expressed as S={S1, S2, S3, S4}.

[0149] In one embodiment, a complete respiratory cycle can be divided into three phases: an exhalation phase, an end-expiration phase, and an inhalation phase.

[0150] Assume that the hidden state is converted in the order of S1→S2→S3→S4→S1→…; the state transition probability matrix between hidden states can be defined as:

[0151] Among them, a ij =P{q t+1 =S j |qt =S i},1≤i,j≤4, that is, from state S i Transfer to state S j probability.

[0152] S420: Determine the respiratory phase corresponding to the observation sequence corresponding to each plotted point under the hidden Markov model.

[0153] After the model parameters of the hidden Markov model are determined, the breathing phase corresponding to the current observation sequence under the hidden Markov model is determined based on the Viterbi algorithm. The Viterbi algorithm is an algorithm in related art that actually uses dynamic programming to solve the prediction problem of the hidden Markov model.

[0154] If the hidden sequence includes four breathing stage identifiers, the corresponding hidden Markov model can be used to estimate the four breathing stages; if the hidden sequence includes three breathing stage identifiers, the corresponding hidden Markov model can be used to estimate the three breathing stages.

[0155] The breathing process of different patients is different. For example, different patients have different inhalation processes and / or different exhalation processes. Reflected in the respiratory curve, the time span of the partial respiratory curve of the exhalation phase of some patients is relatively large, while the time span of the partial respiratory curve of the exhalation phase of some patients is relatively small. Even if the time span of the partial respiratory curve of the exhalation phase of different patients is the same, some patients may begin to enter the end-expiration phase at phase C, while some patients may begin to enter the end-expiration phase at phase D, and phase D is greater than phase C. Therefore, compared with determining the end-expiration phase based only on a fixed amplitude threshold or phase threshold, the present application determines the model parameters of the hidden Markov model through the Baum-Welch algorithm, and establishes a hidden correspondence between the observation sequence and the respiratory phase for each patient; since the observation sequence is determined based on the prior probability of the respiratory phase, it can be regarded as establishing a correspondence between the respiratory phase and the respiratory phase for each patient, thereby achieving the technical effect of accurately estimating the patient's respiratory phase based on the respiratory phase.

[0156] S430: Determine the respiratory stages included in the segmentation result according to the respiratory stages corresponding to the observation sequence corresponding to each drawing point.

[0157] After the respiratory stage corresponding to each drawing point is determined, the respiratory stage corresponding to the drawing point included in each segmentation result is determined, and the respiratory stage included in each drawing point is also determined.

[0158] The embodiments of the present application achieve the technical effect of accurately estimating the respiratory stage based on the respiratory phase and the hidden Markov model.

[0159] FIG7A is a schematic diagram of the structure of a respiratory data processing device provided in an embodiment of the present application. As shown in FIG7A , the device includes: a first respiratory curve determination module 310 , a respiratory phase module 320 , an observation sequence determination module 330 , and an estimation module 340 .

[0160] A first breathing curve determining module 310 is configured to determine a first breathing curve reflecting the breathing process of the target object based on three-dimensional motion information of at least one observation point set on the chest and / or abdomen of the target object;

[0161] A respiratory phase module 320 is configured to determine a current respiratory phase of the target subject based on the first respiratory curve;

[0162] An observation sequence determination module 330 is configured to determine a current observation sequence according to a priori probability corresponding to the current respiratory phase;

[0163] The estimation module 340 is configured to determine the respiratory phase corresponding to the current observation sequence under a hidden Markov model, wherein the model parameters of the hidden Markov model are determined based on a Baum-Welch algorithm.

[0164] In one embodiment, the Hidden Markov Model is used to estimate the inspiratory phase, the end-inspiratory phase, the expiratory phase, and the end-expiratory phase.

[0165] The device also includes an output module, which is configured to perform a setting operation when the breathing stage is a target stage; wherein the target stage is the end-expiration stage.

[0166] In one embodiment, the setting operation is outputting prompt information or outputting a control signal to the puncture device, and the control signal is used to control the puncture device to perform the set puncture operation.

[0167] The device also includes an information determination module, a change rate determination module and an update module.

[0168] an information determination module configured to determine an average of a rate of change of a portion of the respiratory curve of the first respiratory curve in the respiratory stage at a current moment, a rate of change at a previous moment, and a difference between the amplitude at the current moment and the amplitude at the previous moment;

[0169] a change rate determination module, configured to determine a change rate of the second respiratory curve at the current moment based on the change rate mean, the change rate at the previous moment, and the difference;

[0170] The updating module is configured to update the second breathing curve according to a changing rate of the second breathing curve at the current moment.

[0171] The change rate determination module is configured to perform weighted averaging on the change rate mean, the change rate at the previous moment and the difference to obtain a weighted average result; and use the weighted average result as the change rate of the second respiratory curve at the current moment.

[0172] In one embodiment, the first respiratory curve is used to reflect changes in the motion component of the target object's respiratory motion in a main motion direction; the main motion direction is the direction of the covariance matrix eigenvector corresponding to the first principal component of the respiratory motion signal.

[0173] The device also includes a first prompt information determination module, which is configured to output first prompt information indicating abnormal breathing regularity when it is determined that the same first abnormal event occurs N times consecutively in the first breathing curve, wherein the first abnormal event is an amplitude abnormality, a periodic abnormality, or an abnormal duration of a set breathing stage; wherein N is greater than or equal to 2, the periodic abnormality is determined based on serial correlation, the amplitude abnormality is determined based on the maximum or minimum value of the target breathing stage, and the set breathing stage and the target breathing stage are both determined based on the hidden Markov model.

[0174] The device also includes a second prompt information determination module, which is configured to output second prompt information indicating the movement of the target object when a baseline drift abnormality is determined according to the first respiratory curve.

[0175] In one embodiment, determining whether the first respiratory curve has an abnormal amplitude is performed by the following steps:

[0176] Segmentation processing is performed on the first breathing curve based on a first segmentation duration to obtain a segmentation result, wherein the first segmentation duration is greater than 0.75 breathing cycles and less than one breathing cycle; when the segmentation result includes the end-inhalation stage and the duration of the end-inhalation stage is greater than half of the duration corresponding to the entire end-inhalation stage, the maximum value of the segmentation result is extracted, or when the segmentation result includes the end-expiration stage and the duration of the end-expiration stage is greater than half of the duration corresponding to the entire end-expiration stage, the minimum value of the segmentation result is extracted; when the maximum value or the minimum value does not meet the set amplitude condition with the maximum value or the minimum value of the set period, the maximum value or the minimum value is marked as an amplitude abnormality; when the amplitude abnormality occurs N times consecutively, it is determined that the first breathing curve has an amplitude abnormality.

[0177] In one embodiment, the respiratory phase included in the segmentation result is determined by the following steps, including:

[0178] Determine an observation sequence corresponding to each drawing point of the segmentation result, wherein the observation sequence includes a respiratory phase corresponding to each drawing point; determine the respiratory stage corresponding to the observation sequence corresponding to each drawing point under the hidden Markov model; and determine the respiratory stage included in the segmentation result based on the respiratory stage corresponding to the observation sequence corresponding to each drawing point.

[0179] In one embodiment, determining whether the first respiratory curve has periodic abnormality is performed by the following steps:

[0180] When a current respiratory curve segment is determined on the first respiratory curve based on a second segment duration, a correlation metric value between the current respiratory curve segment and a set respiratory curve segment is determined based on an autocorrelation operation, wherein the second segment duration is greater than one respiratory cycle and less than two respiratory cycles; when the correlation metric value does not meet a set metric matching condition, the current respiratory curve segment is recorded as a repetitive anomaly; when the repetitive anomaly occurs N times consecutively, it is determined that a periodic repetitive anomaly occurs in the first respiratory curve.

[0181] In one embodiment, the set breathing phase is the end-expiratory phase;

[0182] The abnormal duration of the end-expiratory phase is determined by the following steps:

[0183] In the first respiratory curve, when the respiratory stage corresponding to the current plotted point is the end-expiratory stage, the total number of plotted points corresponding to the end-expiratory stage is increased by 1 until the current respiratory stage is the start moment of the inspiratory stage;

[0184] Determining the duration of the end-expiratory phase according to the total number of plotted points corresponding to the end-expiratory phase;

[0185] If the duration of the end-expiratory phase is abnormal, marking the end-expiratory phase as an abnormal end-expiratory phase;

[0186] When the abnormal end-expiratory phase occurs N times consecutively, it is determined that the duration of the end-expiratory phase in the first respiratory curve is abnormal.

[0187] In one embodiment, the device also includes a learning module, which is configured to control the hidden Markov model to enter a learning phase when the first prompt information is detected; and return to execution when it is determined that the same first abnormal event occurs N times consecutively in the first breathing curve, to output a first prompt information indicating abnormal breathing regularity.

[0188] FIG7B is a schematic structural diagram of a respiratory data processing device provided by an embodiment of the present application. As shown in FIG7B , the device includes: a respiratory phase module 31, an observation sequence determination module 32, an estimation module 33, and an output module 34. The respiratory phase module 31 is configured to determine the current respiratory phase of the target object based on a first respiratory curve, wherein the first respiratory curve is determined based on three-dimensional motion information of at least one observation site set on the chest and / or abdomen of the target object; the observation sequence determination module 32 is configured to determine the current observation sequence based on the prior probability corresponding to the current respiratory phase; the estimation module 33 is configured to determine the respiratory stage corresponding to the current observation sequence under a hidden Markov model, wherein the model parameters of the hidden Markov model are determined based on the Baum-Welch algorithm; and the output module 34 is configured to perform a setting operation when the respiratory stage is the target stage.

[0189] In one embodiment, the Hidden Markov Model may be used to estimate the inspiratory phase, the end-inspiratory phase, the expiratory phase, and the end-expiratory phase.

[0190] In one embodiment, the target phase is the end-expiratory phase.

[0191] In one embodiment, the estimation module 33 is configured to determine the respiratory phase corresponding to the current observation sequence under the Hidden Markov Model based on the Viterbi algorithm.

[0192] In one embodiment, the setting operation is outputting prompt information or outputting a control signal to the puncture device, and the control signal is used to control the puncture device to perform the set puncture operation.

[0193] In one embodiment, as shown in FIG7C , the apparatus further includes a breathing curve updating module 35 , wherein the breathing curve updating module 35 includes:

[0194] The first unit is configured to determine an average of a rate of change of a portion of the respiratory curve of the first respiratory curve in the respiratory stage at a current moment, a rate of change at a previous moment, and a difference between an amplitude at the current moment and an amplitude at the previous moment;

[0195] The second unit is configured to determine the change rate of the second respiratory curve at the current moment according to the change rate mean, the change rate at the previous moment and the difference;

[0196] The third unit is configured to update the second breathing curve according to a change rate of the second breathing curve at a current moment.

[0197] In one embodiment, the first respiratory curve is used to reflect changes in the motion component of the target object's respiratory motion in the main motion direction;

[0198] The main motion direction is the direction of the covariance matrix eigenvector corresponding to the first principal component of the respiratory motion signal.

[0199] In one embodiment, the second unit is configured as:

[0200] Performing a weighted average on the change rate mean, the change rate at the previous moment, and the difference to obtain a weighted average result;

[0201] The weighted average result is used as the change rate of the second respiratory curve at the current moment.

[0202] The technical solution of the embodiment of the present application is that since the model parameters of the hidden Markov model are determined based on the Baum-Welch algorithm, the determination of the model parameters can be completed quickly; since the first respiratory curve is used to reflect the change of the motion component of the respiratory motion of the target object in the main motion direction, the respiratory phase of the target object at the current moment can be accurately determined according to the first respiratory curve, and the current observation sequence can be accurately determined according to the prior probability corresponding to the respiratory phase; thereby accurately determining the respiratory stage corresponding to the current observation sequence under the hidden Markov model, achieving the technical effect of accurately estimating the patient's respiratory stage at the current moment based on the respiratory phase and the hidden correspondence between the respiratory phase and the respiratory stage unique to the target object.

[0203] The respiratory data processing device provided in the embodiments of the present application can execute the respiratory data processing method provided in any embodiment of the present application, and has the corresponding functional modules and effects of the execution method.

[0204] FIG8 is a schematic diagram of the structure of a respiratory data processing device provided in an embodiment of the present application. As shown in FIG8 , the device includes: a determination module 41 configured to determine a first respiratory curve reflecting the respiratory process of a target subject; a first monitoring module 42 configured to output a first prompt message indicating abnormal respiratory regularity if it is determined that the first respiratory curve has the same first abnormal event N times consecutively, wherein the first abnormal event is an amplitude abnormality, a periodic abnormality, or an abnormal duration of a set respiratory phase; wherein N is greater than or equal to 2, the periodic abnormality is determined based on serial correlation, the amplitude abnormality is determined based on the maximum or minimum value of the target respiratory phase, and the set respiratory phase and the target respiratory phase are determined based on a hidden Markov model.

[0205] In one embodiment, as shown in FIG9 , the device further includes a second monitoring module 43 , which is configured to:

[0206] When it is determined that the baseline drift is abnormal according to the first respiratory curve, second prompt information for indicating the movement of the target object is output.

[0207] In one embodiment, the amplitude abnormality of the first respiratory curve is determined by an amplitude unit, and the amplitude unit is configured to:

[0208] performing segmentation processing on the first respiratory curve based on a first segmentation duration to obtain a segmentation result, wherein the first segmentation duration is greater than 0.75 respiratory cycles and less than one respiratory cycle;

[0209] If the segmented result includes the end-inhalation phase or the end-expiration phase, and the duration of the end-inhalation phase is greater than half of the duration of the entire end-inhalation phase, and the duration of the end-expiration phase is greater than half of the duration of the entire end-expiration phase, extracting the maximum value or the minimum value of the segmented result;

[0210] If the maximum value or minimum value does not meet the set amplitude condition with the maximum value or minimum value of the set period, the maximum value or minimum value is marked as an amplitude abnormality;

[0211] If the amplitude abnormality occurs N times consecutively, it is determined that the first respiratory curve has an amplitude abnormality.

[0212] In one embodiment, the respiratory stage included in the segmentation result is determined by a respiratory stage estimation unit, and the respiratory stage estimation unit is configured to:

[0213] Determine an observation sequence corresponding to each plotted point of the segmentation result, wherein the observation sequence includes a respiratory phase corresponding to each plotted point;

[0214] determining a respiratory phase corresponding to the observation sequence corresponding to each plotted point under a hidden Markov model;

[0215] The respiratory stages included in the segmentation result are determined according to the respiratory stages corresponding to the observation sequence corresponding to each of the plotted points.

[0216] In one embodiment, the periodic abnormality of the first respiratory curve is determined by a periodic unit, and the periodic unit is configured to:

[0217] determining a correlation measure between the current respiratory curve segment and a set respiratory curve segment based on an autocorrelation operation when a current respiratory curve segment is determined on the first respiratory curve based on a second segment duration, wherein the second segment duration is greater than one respiratory cycle and less than two respiratory cycles;

[0218] If the correlation metric value does not meet the set metric matching condition, the current respiratory curve segment is recorded as a repetitive anomaly;

[0219] If the repetitive abnormality occurs N times consecutively, it is determined that the first respiratory curve has a periodic repetitive abnormality.

[0220] In one embodiment, the set breathing phase is the end-expiratory phase;

[0221] The duration of the end-expiratory phase is determined by the duration unit, which is set to:

[0222] If the respiratory phase corresponding to the current plotted point in the first respiratory curve is the end-expiratory phase, the total number of plotted points corresponding to the end-expiratory phase is increased by 1 until the current respiratory phase is the start of the inspiratory phase;

[0223] Determining the duration of the end-expiratory phase according to the total number of plotted points corresponding to the end-expiratory phase;

[0224] If the duration of the end-expiratory phase is abnormal, marking the end-expiratory phase as an abnormal end-expiratory phase;

[0225] If the abnormal end-expiratory phase occurs N times consecutively, it is determined that the duration of the end-expiratory phase in the first respiratory curve is abnormal.

[0226] In one embodiment, the apparatus further comprises a circulation module, wherein the circulation module is configured to:

[0227] When the first prompt information is detected, controlling the hidden Markov model to enter a learning phase;

[0228] When it is detected that the hidden Markov model learning is completed, the method returns to the step of outputting first prompt information indicating abnormal breathing regularity if it is determined that the same first abnormal event occurs N times consecutively in the first breathing curve.

[0229] The technical solution provided in the embodiment of the present application realizes unsupervised learning of the hidden Markov model through the Baum-Welch algorithm, realizes tailor-made hidden Markov model for each target object, and estimates the set exhalation phase and the target exhalation phase through the combination of the hidden Markov model and the first respiratory curve, thereby improving the accuracy of the respiratory phase estimation, the accuracy of the amplitude abnormality determined based on the maximum or minimum value of the target respiratory phase, and the accuracy of the abnormality of the duration of the set respiratory phase, thereby improving the accuracy of respiratory regularity monitoring.

[0230] The respiratory data processing device provided in the embodiments of the present application can execute the respiratory data processing method provided in any embodiment of the present application, and has the corresponding functional modules and effects of the execution method.

[0231] FIG10 shows a block diagram of an electronic device 10 that can be used to implement an embodiment of the present application. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present application described and / or required herein.

[0232] As shown in Figure 10, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by the at least one processor 11, and the processor 11 can perform a variety of appropriate actions and processes according to the computer program stored in the ROM 12 or the computer program loaded from the storage unit 18 into the RAM 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, ROM 12 and RAM 13 are connected to each other via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0233] Multiple components in the electronic device 10 are connected to the I / O interface 15, including an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a magnetic disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0234] The processor 11 can be a variety of general-purpose and / or specialized processing components with processing and computing capabilities. Some examples of the processor 11 include a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The processor 11 performs the various methods and processes described above, such as the respiratory data processing method.

[0235] In some embodiments, the respiratory data processing method can be implemented as a computer program tangibly embodied in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the respiratory data processing described above can be performed. Alternatively, in other embodiments, processor 11 can be configured to perform the respiratory data processing method in any other suitable manner (e.g., via firmware).

[0236] Various embodiments of the systems and techniques described herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard parts (ASSPs), system on chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system that includes at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0237] Computer programs for implementing the methods of the present application may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when the computer program is executed by the processor, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The computer program may be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0238] In the context of the present application, a computer-readable storage medium can be a tangible medium that can contain or store a computer program for use by an instruction execution system, device or equipment or used in combination with an instruction execution system, device or equipment. A computer-readable storage medium can include electronic, magnetic, optical, electromagnetic, infrared or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium can be a machine-readable signal medium. A machine-readable storage medium includes an electrical connection based on one or more lines, a portable computer disk, a hard disk, RAM, ROM, an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device or any suitable combination of the foregoing. The storage medium can be a non-transitory storage medium.

[0239] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a cathode ray tube (CRT) or a liquid crystal display (LCD) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0240] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: Local Area Networks (LANs), Wide Area Networks (WANs), blockchain networks, and the Internet.

[0241] A computing system may include a client and a server. The client and server are generally remote from each other and typically interact via a communication network. The client-server relationship arises through computer programs running on the respective computers and establishing a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, a host product within a cloud computing service ecosystem that addresses the management difficulties and limited scalability of physical hosts and virtual private server (VPS) services.

[0242] The various forms of processes shown above can be used to reorder, add, or delete steps. For example, the multiple steps described in this application can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solution of this application can be achieved. This document is not limited here.

Claims

1. A respiratory data processing method, comprising: Determine a first breathing curve reflecting the breathing process of the target object based on three-dimensional motion information of at least one observation site set on the chest and / or abdomen of the target object; determining a current breathing phase of the target object according to the first breathing curve; Determine a current observation sequence according to a priori probability corresponding to the current respiratory phase; The breathing phase corresponding to the current observation sequence under a hidden Markov model is determined, wherein the model parameters of the hidden Markov model are determined based on a Baum-Welch algorithm.

2. The method according to claim 1, wherein: The hidden Markov model is used to estimate the inspiratory phase, the end-inspiratory phase, the expiratory phase and the end-expiratory phase.

3. The method according to claim 1, further comprising: When the breathing phase is a target phase, a setting operation is performed; wherein the target phase is an end-expiratory phase.

4. The method according to claim 3, wherein: The setting operation is outputting prompt information or outputting a control signal to the puncture device, and the control signal is used to control the puncture device to perform the set puncture operation.

5. The method according to claim 1, further comprising: Determine the average of the change rates of the partial breathing curves of the breathing stage at the current moment in the first breathing curve, the change rate at the previous moment, and the difference between the amplitude at the current moment and the amplitude at the previous moment; Determine the change rate of the second respiratory curve at the current moment according to the change rate mean, the change rate at the previous moment and the difference; The second breathing curve is updated according to a rate of change of the second breathing curve at the current moment.

6. The method according to claim 5, wherein: The determining the change rate of the second respiratory curve at the current moment according to the change rate mean, the change rate at the previous moment and the difference comprises: Performing weighted averaging on the change rate mean, the change rate at the previous moment and the difference to obtain a weighted average result; The weighted average result is used as the change rate of the second respiratory curve at the current moment.

7. The method according to claim 1, wherein: The first breathing curve is used to reflect the change of the motion component of the target object's respiratory motion in the main motion direction; The main motion direction is the direction of the eigenvector of the covariance matrix corresponding to the first principal component of the respiratory motion signal.

8. The method according to claim 1, further comprising: When it is determined that the first breathing curve has the same first abnormal event N times in succession, outputting first prompt information indicating abnormal breathing regularity, wherein the first abnormal event is abnormal amplitude, abnormal periodicity, or abnormal duration of a set breathing phase; Among them, N is greater than or equal to 2, the periodic anomaly is determined based on sequence correlation, the amplitude anomaly is determined based on the maximum or minimum value of the target breathing stage, and the set breathing stage and the target breathing stage are both determined based on the hidden Markov model.

9. The method according to claim 8, further comprising: When it is determined that the baseline drift is abnormal according to the first respiratory curve, second prompt information for indicating the movement of the target object is output.

10. The method according to claim 8, wherein: Determine that the first respiratory curve has an abnormal amplitude by the following steps: Performing segmentation processing on the first breathing curve based on a first segmentation duration to obtain a segmentation result, wherein the first segmentation duration is greater than 0.75 breathing cycles and less than one breathing cycle; When the segmentation result includes the end-inhalation stage and the duration of the end-inhalation stage is greater than half of the duration of the entire end-inhalation stage, extract the maximum value of the segmentation result; or, when the segmentation result includes the end-expiration stage and the duration of the end-expiration stage is greater than half of the duration of the entire end-expiration stage, extract the minimum value of the segmentation result; When the maximum value or the minimum value does not meet the set amplitude condition with the maximum value or the minimum value of the set period, the maximum value or the minimum value is marked as an amplitude abnormality; When the amplitude abnormality occurs N times continuously, it is determined that the first respiratory curve has an amplitude abnormality.

11. The method according to claim 10, wherein: The respiratory phase included in the segmented result is determined by the following steps, including: Determine an observation sequence corresponding to each drawing point of the segmentation result, wherein the observation sequence includes a respiratory phase corresponding to each drawing point; determining a breathing phase corresponding to the observation sequence corresponding to each plotted point under the hidden Markov model; The respiratory stage included in the segmentation result is determined according to the respiratory stage corresponding to the observation sequence corresponding to each drawing point.

12. The method according to claim 8, wherein: Determine that the first respiratory curve has periodic abnormality by the following steps: In a case where a current breathing curve segment is determined on the first breathing curve based on a second segment duration, determining a correlation measure value between the current breathing curve segment and a set breathing curve segment based on an autocorrelation operation, wherein the second segment duration is greater than one breathing cycle and less than two breathing cycles; When the correlation metric value does not meet the set metric matching condition, recording the current respiratory curve segment as a repetitive anomaly; When the abnormal repetitiveness occurs N times continuously, it is determined that the first respiratory curve has a periodic abnormal repetitiveness.

13. The method according to claim 8, wherein: The set breathing stage is the end-expiratory stage; The abnormal duration of the end-expiratory phase is determined by the following steps: In the first respiratory curve, when the respiratory stage corresponding to the current drawing point is the end-expiratory stage, the total number of drawing points corresponding to the end-expiratory stage is increased by 1 until the current respiratory stage is the starting moment of the inhalation stage; Determining the duration of the end-expiratory phase according to the total number of plotted points corresponding to the end-expiratory phase; When the duration of the end-expiratory phase is abnormal, marking the end-expiratory phase as an abnormal end-expiratory phase; When the abnormal end-expiration stage occurs N times in succession, it is determined that the duration of the end-expiration stage in the first breathing curve is abnormal.

14. The method according to claim 8, further comprising: When the first prompt information is detected, controlling the hidden Markov model to enter a learning phase; When it is detected that the hidden Markov model learning is finished, the method returns to executing the method of outputting first prompt information indicating abnormal breathing regularity when it is determined that the same first abnormal event occurs N times continuously in the first breathing curve.

15. A respiratory data processing device, comprising: A first breathing curve determining module is configured to determine a first breathing curve reflecting the breathing process of the target object based on three-dimensional motion information of at least one observation site set on the chest and / or abdomen of the target object; a breathing phase module, configured to determine a current breathing phase of the target object according to the first breathing curve; An observation sequence determination module, configured to determine a current observation sequence according to a priori probability corresponding to the current respiratory phase; The estimation module is configured to determine the breathing phase corresponding to the current observation sequence under a hidden Markov model, wherein the model parameters of the hidden Markov model are determined based on a Baum-Welch algorithm.

16. An electronic device, comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so as to enable the at least one processor to perform the respiratory data processing method according to any one of claims 1 to 14.

17. A computer-readable storage medium, wherein: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the respiratory data processing method according to any one of claims 1 to 14 when executed.

Images