Medical device and snoring detection method
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- FUKUDA DENSHI CO LTD
- Filing Date
- 2024-03-29
- Publication Date
- 2026-07-30
Smart Images
Figure 0007897889000001 
Figure 0007897889000002 
Figure 0007897889000003
Abstract
Description
Technical Field
[0001] The present invention relates to a medical device and a snoring detection method.
Background Art
[0002] As a treatment method for sleep apnea syndrome (SAS), nasal continuous positive airway pressure therapy (CPAP) is known. CPAP is a treatment method that continuously delivers air at an appropriate pressure to the patient's airway through a nasal mask, preventing airway obstruction during sleep and suppressing the occurrence of apnea and hypopnea. Snoring may occur as a precursor to the occurrence of apnea and hypopnea. In Patent Document 1, the occurrence of snoring is detected when the value obtained by subtracting the flow rate during the exhalation period from the flow rate during the inhalation period exceeds a threshold value.
Prior Art Documents
Patent Documents
[0003]
Patent Document 1
Summary of the Invention
Problems to be Solved by the Invention
[0004] With the method of Patent Document 1, it is difficult to accurately detect snoring. Some aspects of the present invention aim to provide a technique for accurately detecting the occurrence of snoring.
Means for Solving the Problems
[0005] According to some embodiments, a medical device comprises: generation means for generating high-frequency data from flow rate data representing a time series of the patient's respiratory flow rate; determination means for determining an evaluation value of the high-frequency data for each of a plurality of periods; selection means for selecting one of the plurality of periods based on the evaluation values of the plurality of periods; and detection means for detecting the occurrence of snoring in the patient by analyzing the portion of the flow rate data for the selected one period, wherein in the high-frequency data, frequency components below a specific frequency lower than the frequency component caused by snoring are blocked from the flow rate data. The aforementioned evaluation value is a value used to assess the likelihood that the patient is snoring. Medical equipment will be provided. [Effects of the Invention]
[0006] According to some embodiments, snoring can be detected with high accuracy. [Brief explanation of the drawing]
[0007] [Figure 1] A block diagram illustrating an example configuration of a CPAP device in some embodiments. [Figure 2] A flowchart illustrating an example of a snoring detection method in some embodiments. [Figure 3] A schematic diagram illustrating an example of a method for determining the analysis period in some embodiments. [Figure 4] A schematic diagram illustrating an example of detection conditions in some embodiments. [Figure 5] A flowchart illustrating an example of a method for changing treatment pressure in some embodiments. [Modes for carrying out the invention]
[0008] The embodiments will be described in detail below with reference to the attached drawings. Note that the following embodiments do not limit the invention as defined in the claims, and not all combinations of features described in the embodiments are essential to the invention. Two or more of the features described in the embodiments may be combined in any way. Furthermore, identical or similar configurations will be given the same reference numeral, and redundant descriptions will be omitted.
[0009] Referring to Figure 1, an example configuration of a continuous positive airway pressure (CPAP) device (hereinafter referred to as a CPAP device) 100 according to one embodiment will be described. The CPAP device 100 is an example of a medical device having a snoring detection function. The CPAP device 100 has a main unit 101, a mask 125, and a tube 122 connecting the main unit 101 and the mask 125. The operation of the CPAP device 100 is realized by the CPU (Central Processing Unit) 112 reading a program stored in ROM (Read Only Memory) 113 into RAM (Random Access Memory) 114 and executing it. Thus, a device having a CPU 112, ROM 113 and RAM 114 may be considered a computer. The function blocks 117 to 121 described in the CPU 112 schematically represent the main functions among the various functions realized by the CPU 112 executing a program. Therefore, the operation described mainly using function blocks 117 to 121 is actually realized by the CPU 112 executing a program. Alternatively, one or more functional blocks may be implemented using hardware circuits other than the CPU 112.
[0010] First, let's describe the components present in the airflow path. The filter 102 is installed at the air intake and removes pollen, bacteria, dust, etc. The temperature sensor 103 measures the temperature of the incoming air. The value measured by the temperature sensor 103 is supplied to the temperature control unit 119. The humidity sensor 104 measures the humidity of the incoming air. The value measured by the humidity sensor 104 is supplied to the temperature control unit 119.
[0011] The flow (differential pressure) sensor 105 (hereinafter simply referred to as flow sensor 105) is, for example, a differential pressure type flow sensor that measures the flow rate of air in the flow path based on the pressure difference between the upstream and downstream sides. Here, a positive measurement value is obtained when the upstream pressure is higher than the downstream pressure, and a negative measurement value is obtained when the downstream pressure is higher than the upstream pressure. Therefore, the direction of airflow in the flow path can also be determined from the measurement value of the flow sensor 105. The measurement value of the flow sensor 105 is supplied to the respiratory analysis unit 117.
[0012] The blower 106 has an impeller and a motor that drives the impeller inside. The treatment pressure control unit 118 can adjust the flow rate and supply pressure of air supplied to the patient by controlling the rotation speed of the motor through the motor driver 108.
[0013] The pressure sensor 107 is located downstream of the blower 106 in the flow path and measures the pressure in the flow path. The value measured by the pressure sensor 107 is supplied to the treatment pressure control unit 118. The treatment pressure control unit 118 controls the supply pressure, assuming that air is being supplied to the patient at the pressure measured by the pressure sensor 107.
[0014] The humidifier 109 has a water storage tank and humidifies the air supplied to the tube 122. Here, the temperature control unit 119 controls the amount of water vaporized from the water storage tank, i.e., the degree of humidification, by controlling the temperature of the heater 110 provided in the humidifier 109. The temperature sensor 111 measures the temperature of the heater 110 and supplies it to the temperature control unit 119. In this embodiment, by using the heater 110 in the humidifier 109, the temperature of the air supplied to the patient is adjusted along with humidification. Note that the temperature adjustment of the air can also be achieved by other methods, such as using a heater 123 provided in the tube 122. If humidification and temperature adjustment are performed separately, the heater 110 and temperature sensor 111 do not need to be provided in the humidifier 109. Alternatively, the air may be blown onto the water surface in the water storage tank, or the flow path may be arranged so that it passes through the water.
[0015] The tube 122 connects the main body 101 and the mask 125. The tube 122 has elasticity and flexibility so that it can easily follow the movement of the mask 125. A temperature sensor 124 for measuring the temperature of the air supplied to the patient is provided in the tube 122. The measured value of the temperature sensor 124 is supplied to the temperature control unit 119.
[0016] The mask 125 has a size and shape that cover the patient's nose and mouth, and is worn on the patient by strings or bands with adjustable lengths.
[0017] The display unit 115 is, for example, a display provided on the housing of the main body 101, and displays messages regarding the operation of the CPAP device 100, various menu screens for setting the CPAP device 100, measured values of various sensors, and the like. The display of the display unit 115 is controlled by the input / output control unit 120.
[0018] The operation unit 116 is a general term for input devices that can be operated by a user, such as buttons and switches provided on the housing of the main body 101, for example. When the display unit 115 is a touch display, the display unit 115 and the operation unit 116 are integrally configured. The input / output control unit 120 detects an operation on the operation unit 116, and the CPU 112 executes an operation corresponding to the detected operation.
[0019] The respiration analysis unit 117 detects the occurrence of a predetermined event based on the flow rate measured by the flow sensor 105. When the respiration analysis unit 117 detects the occurrence of a predetermined event, it notifies the treatment pressure control unit 118. In the present embodiment, the respiration analysis unit 117 detects the occurrence of the patient's snoring.
[0020] The treatment pressure control unit 118 controls the operation of the blower 106 based on the measured value of the pressure sensor 107 so that the supply air pressure becomes the target value. Also, the treatment pressure control unit 118 switches between the control of the supply air pressure during the patient's inhalation and the control of the supply air pressure during exhalation in response to a notification from the respiration analysis unit 117. The treatment pressure control unit 118 controls the supply air pressure by, for example, giving the duty ratio of the pulsed voltage applied to the motor to the motor driver 108 to control the rotational speed of the impeller of the blower 106. The treatment pressure control unit 118 notifies the temperature control unit 119 of the currently set supply air pressure.
[0021] The temperature control unit 119 controls the operation of the heater 110 according to the measured values of the temperature sensor 103, the humidity sensor 104, the temperature sensor 111, and the temperature sensor 124 and the supply air pressure notified from the treatment pressure control unit 118, thereby controlling the temperature and humidity of the air supplied to the patient. The temperature and humidity of the air supplied to the patient may be the temperature and humidity set by the user through the operation unit 116. The user setting through the operation unit 116 is notified to the temperature control unit 119 through, for example, the input / output control unit 120. The supply air pressure is taken into account because, even if the temperature of the heater 110 is constant, the degree of increase in the air temperature is lower when the flow rate is high than when it is low.
[0022] The communication control unit 121 executes processes related to the communication between the main body 101 and the external system 130. The communication control unit 121 can execute communication with the external system 130 in accordance with, for example, one or more known wireless and / or wired communication standards. The external system 130 may be, for example, a management system for in-hospital medical treatment and treatment data or a remote management system for the CPAP device 100.
[0023] Referring to Figure 2, a method performed by the CPAP device 100 to detect snoring will be described. Snoring can be a vibrational sound produced from the nasopharynx as air passes through a narrowed airway during a patient's sleep. In the following description, each step of the method in Figure 2 is performed by the CPU 112 (e.g., its respiratory analysis unit 117). Specifically, each step is performed by the CPU 112 executing a program read from the RAM 114. Alternatively, at least some steps of the method in Figure 2 may be performed by a dedicated circuit such as an ASIC (Application Specific Integrated Circuit). The method in Figure 2 may be initiated when the patient instructs the CPAP device 100 to start operation (e.g., therapeutic operation during sleep), when the CPAP device 100 detects the patient falling asleep, or when triggered by other events.
[0024] During the execution of the method shown in Figure 2, the values measured by the flow sensor 105 are continuously supplied to the CPU 112 (for example, its respiratory analysis unit 117). As described above, the values measured by the flow sensor 105 represent the patient's respiratory flow rate. The CPU 112 samples the flow rate supplied by the flow sensor 105 at a predetermined sampling interval (for example, 2 ms) and stores the flow rate at each time point as flow rate data in the RAM 114. This flow rate data represents the time series of the patient's respiratory flow rate. As will be described in detail below, the CPU 112 detects the occurrence of snoring by analyzing this flow rate data.
[0025] In S201, the CPU 112 determines whether the patient's current breath has finished. If the CPU 112 determines that the patient's current breath has finished (YES in S201), it proceeds to S202; otherwise (NO in S201), it repeats S201. In this way, the CPAP device 100 waits until the patient's current breath has finished.
[0026] The CPU 112 may detect the end of a current breath based on flow rate data. The start and end points of a breath may be arbitrary. In the following description, the start of a patient's inspiration is considered the start of a breath, and the end of a patient's exhalation is considered the end of a breath. In this case, the CPU 112 may detect the end of a breath based on the flow rate changing from a value less than a predetermined threshold to that threshold. The predetermined threshold may be zero or a non-zero value (positive or negative). Alternatively, the start of a patient's exhalation may be considered the start of a breath, and the end of a patient's inspiration may be considered the end of a breath. In this case, the CPU 112 may detect the end of a breath based on the flow rate changing from a value greater than a predetermined threshold to that threshold. The predetermined threshold may be zero or a non-zero value (positive or negative).
[0027] The duration of a single breath is called the respiratory period. The period during which inspiration occurs is called the inspiratory period. The inspiratory period may be the period during which the flow rate data is greater than a predetermined threshold (e.g., zero or a positive value). The period during which exhalation occurs is called the expiratory period. The expiratory period may be the period during which the flow rate data is less than a predetermined threshold (e.g., zero or a negative value).
[0028] In S202, CPU112 determines the period within the most recently completed respiratory cycle for which flow data will be analyzed to detect snoring. In the following explanation, the period for which flow data is analyzed will be referred to as the analysis period. Details on how the analysis period is determined will be described later.
[0029] In S203, the CPU 112 analyzes the portion of the flow rate data corresponding to the analysis period to determine whether the detection conditions for detecting snoring are met. If the CPU 112 determines that the detection conditions are met (YES in S203), it proceeds to S204; otherwise (NO in S203), it proceeds to S207. Details of the detection conditions will be described later.
[0030] In S204, the CPU 112 increments a counter by 1 to count the number of breaths that have been determined to be consecutive when the detection conditions are met. In the following description, this counter will be referred to as the continuous detection counter. The continuous detection counter is initialized to zero at the start of Figure 2. In S207, the CPU 112 resets the continuous detection counter to zero.
[0031] In S205, the CPU 112 determines whether the value of the continuous detection counter is equal to or greater than a predetermined threshold number of times. If the CPU 112 determines that the value of the continuous detection counter is equal to or greater than the predetermined threshold number of times (YES in S205), it proceeds to S206; otherwise (NO in S205), it proceeds to S201.
[0032] In S206, the CPU 112 detects that snoring occurred during the most recent breath. The CPU 112 may store the duration of this breath in the RAM 114 as the snoring duration. If snoring is detected during multiple consecutive breaths, the CPU 112 may store the total duration of these breaths as the snoring duration. Alternatively, the CPU 112 may store the total duration of multiple breaths determined to be consecutive when the detection conditions are met as the snoring duration. The snoring duration stored in the RAM 114 may be presented to the patient or physician after the patient wakes up.
[0033] The threshold count used in S205 may be, for example, one. In this case, the CPU 112 detects snoring when one breath that satisfies the detection condition occurs. Alternatively, the threshold count used in S205 may be, for example, two or more (for example, three). In this case, the CPU 112 detects snoring when two or more (for example, three) breaths that satisfy the detection condition occur consecutively. By detecting snoring when the detection condition is met multiple times consecutively in this way, false detections caused by body movement or mask displacement can be suppressed.
[0034] As described above, in the method shown in Figure 2, the CPU 112 executes steps S202 to S206 each time a patient's breath is completed. Alternatively, the CPU 112 may execute steps S202 onwards each time a predetermined number of breaths (two or more) are completed. For example, the CPU 112 (e.g., the respiratory analysis unit 117) counts the number of breaths the patient has taken since the most recent execution of steps S202 onwards. When the respiratory count reaches a predetermined number, the CPU 112 executes steps S202 onwards. The CPU 112 may also execute steps S203 to S206 sequentially for each of the most recent predetermined number of breaths, starting from the past breath. Even in this case, the same results can be obtained as when steps S202 to S206 are executed each time a patient's breath is completed. Alternatively, the CPU 112 may execute steps S203 to S206 for a portion of the most recent predetermined number of breaths. For example, the CPU 112 may execute steps S203 to S206 for the most recent breath out of a predetermined number of breaths. In this case, steps S203 to S206 are executed for non-consecutive breaths.
[0035] Next, referring to Figure 3, we will explain in detail how to determine the analysis period for S202 in Figure 2. Graph 301 in Figure 3 shows an example of flow rate data. The horizontal axis of Graph 301 represents time, and the vertical axis represents flow rate. Although Figure 3 shows the flow rate data continuously, as mentioned above, the flow rate data may also be digital data. Graph 301 focuses on the duration of a single breath from the flow rate data.
[0036] In the example in Graph 301, inhalation begins at time t1. Inhalation ends and exhalation begins at time t4. Graph 301 includes high-frequency components due to snoring between times t2 and t3. Generally, snoring has a frequency of 10 Hz or higher. High-frequency components of 10 Hz or higher are superimposed on the airflow when snoring occurs.
[0037] The CPU 112 may determine the inspiratory period 303 (times t1 to t4) of a single respiratory cycle as the analysis period. Snoring is known to occur more frequently during inspiration than during exhalation. Exhalation may contain noise caused by air turbulence resulting from the patient's exhalation (for example, air turbulence caused by the collision of air supplied from the CPAP device 100 with the patient's exhaled breath). Therefore, attempting to detect snoring based on the difference between the amount of noise in inspiration and the amount of noise in exhalation may not accurately detect snoring. In this embodiment, the CPU 112 can accurately detect snoring by including the inspiratory period 303 and excluding the exhalation period in the analysis period.
[0038] The CPU 112 may determine the entire inspiratory period 303 as the analysis period. Alternatively, the CPU 112 may determine only the portion of the inspiratory period 303 in which snoring is thought to occur as the analysis period. This reduces computational costs and improves the accuracy of snoring detection compared to the case where the entire inspiratory period 303 is analyzed.
[0039] The following describes a method for determining that only a portion of the intake period 303 will be used for analysis. The CPU 112 applies a filter to the flow rate data that blocks low-frequency components. The data generated by this filtering is referred to as high-frequency data. The filter that blocks low-frequency components may be a high-pass filter or a band-pass filter. For example, the filter that blocks low-frequency components may block frequency components below a specific frequency (e.g., 4.3 Hz) within the range of 4 to 10 Hz, and allow frequency components above this frequency to pass through.
[0040] If the flow rate data is digital data, the CPU 112 may perform filtering by subtracting its moving average from the flow rate data. For example, if the sampling interval of the flow rate data is 2ms, the CPU 112 may generate high-frequency data by subtracting a 51-point moving average from the flow rate data.
[0041] Graph 302 in Figure 3 shows high-frequency data generated by applying a high-pass filter with a cutoff frequency of 4.3 Hz to the flow rate data shown in Graph 301. The horizontal axis of Graph 302 represents time, and the vertical axis represents flow rate.
[0042] The CPU 112 may determine the analysis period based on the high-frequency data represented by the graph 302. For example, the CPU 112 may determine the analysis period to be a period centered on the time when the absolute value of the high-frequency data is maximum and having a predetermined time length (a portion of the inspiratory period 303 of one breath, for example, a value in the range of 400ms to 1500ms, for example, 1024ms). If the period centered on the time when the absolute value of the high-frequency data is maximum and having a predetermined time length includes a portion that is not included in the inspiratory period 303, the CPU 112 may shift this period so that the entire period is included in the inspiratory period 303. In this case, the CPU 112 may shift the analysis period while maintaining the predetermined time length so that the start point of the analysis period coincides with the start point of the inspiratory period 303, or so that the end point of the analysis period coincides with the end point of the inspiratory period 303. Furthermore, the CPU 112 may not maintain a predetermined time length and may modify (shorten or extend) the analysis period, including a portion before the start of the inspiratory period 303, or a portion after the end of the inspiratory period 303, so that the start of the analysis period coincides with the start of the inspiratory period 303, or the end of the analysis period coincides with the end of the inspiratory period 303.
[0043] Alternatively, the CPU 112 may determine one of several candidate periods 304_1 to 304_n included in the inspiratory period 303 of a single breath as the period to be analyzed. The multiple candidate periods 304_1 to 304_n are collectively referred to as candidate period 304. The following description of one candidate period 304 applies to any of the multiple candidate periods 304_1 to 304_n. The subscript of candidate period 304 has a smaller value the closer the starting point of candidate period 304 is to the starting point of the inspiratory period 303 (i.e., the earlier the time). Hereinafter, the multiple candidate periods 304_1 to 304_n will simply be referred to as multiple candidate periods 304.
[0044] In the example in Figure 3, multiple candidate periods 304 with the same duration are arranged with the same spacing to cover the entire intake period 303. In other words, all of the multiple candidate periods 304 have the same duration. The duration of a candidate period 304 may be a value within the range of, for example, 400ms to 1500ms, or it may be 1024ms. The starting points of the multiple candidate periods 304 are aligned at the same time interval. This interval may be wider than the sampling interval for the flow rate data (for example, 2ms), or it may be equal to this sampling interval. This interval may be a value within the range of, for example, 30ms to 150ms, or it may be 64ms. If this interval is too short, the number of candidate periods 304 will increase, and the computational cost of the processing described later will increase. On the other hand, if this interval is too long, the number of candidate periods 304 will decrease, and the options for the period to be analyzed will be limited. The spacing between the starting points of the multiple candidate periods 304 may be a divisor of the period to be analyzed as described above. The ending points of the multiple candidate periods 304 are also aligned at the same time interval. The start of the first candidate period 304_1 coincides with the start of the inspiratory period 303. Alternatively, the start of the first candidate period 304_1 may be after the start of the inspiratory period 303. The end of the last candidate period 304_n coincides with the end of the inspiratory period 303. Alternatively, the end of the last candidate period 304_n may be before the end of the inspiratory period 303.
[0045] In the example in Figure 3, the intervals between multiple candidate periods 304 are shorter than the duration of a single candidate period 304. Therefore, each of the multiple candidate periods 304 partially overlaps with another candidate period of the multiple candidate periods 304. For example, candidate interval 304_1 partially overlaps with candidate periods 304_2, 304_3, etc. This allows more candidate periods 304 to be included in the inspiratory period 303, thus enabling more accurate analysis of the inspiratory period 303. Alternatively, the intervals between multiple candidate periods 304 may be equal to or longer than the duration of a single candidate period 304. In this case, each of the multiple candidate periods 304 does not overlap with another candidate period of the multiple candidate periods 304.
[0046] Instead of the example in Figure 3, the multiple candidate periods 304 may have different durations. Also, the starting points of the multiple candidate periods 304 do not have to be aligned at the same time interval. Snoring is thought to be more likely to occur near the middle of inhalation than at the end of inhalation. Therefore, for example, shorter candidate periods 304 may be densely arranged near the middle of the period under analysis than those at the ends of the period under analysis.
[0047] In the example above, multiple candidate periods 304 are included in the inspiratory period 303 of a single breath. Alternatively, the multiple candidate periods 304 may include portions not included in the inspiratory period 303 of a single breath. For example, the multiple candidate periods 304 may include candidate periods 304 included in the expiratory period. Furthermore, the multiple candidate periods 304 may be distributed across the inspiratory periods 303 of multiple breaths.
[0048] Next, we will explain how to select one candidate period 304 to be analyzed from among multiple candidate periods 304. The CPU 112 determines the evaluation value of the high-frequency data for each of the multiple candidate periods 304. Hereinafter, the evaluation value of the high-frequency data for one candidate period will simply be referred to as the evaluation value for this candidate period. The evaluation value may be a value used to assess the likelihood that the patient is snoring during one candidate period 304. Specifically, the evaluation value for one candidate period 304 will be larger the more likely it is that the patient is snoring during this candidate period 304.
[0049] The CPU 112 may determine the evaluation value of a candidate period 304 by taking the maximum absolute value of the high-frequency data in that candidate period 304. Alternatively, the CPU 112 may determine the evaluation value based on the integral value of the absolute value of the high-frequency data in a candidate period 304. Specifically, the CPU 112 may determine the evaluation value based on the integral value of the absolute value of the high-frequency data in a candidate period 304. Snoring is known to last for a certain period of time during a single breath. Therefore, by determining the evaluation value based on the integral value of the absolute value of the high-frequency data, the occurrence of snoring can be detected with high accuracy. If multiple candidate periods 304 include different time lengths, the evaluation value may be determined by dividing the integral value of the absolute value of the high-frequency data by the time length of the candidate period 304.
[0050] This section describes an example of how to determine the evaluation value of high-frequency data for each of multiple candidate periods 304 when determining the integral of the absolute value of high-frequency data in one candidate period 304 as the evaluation value. First, the CPU 112 calculates the integral of the absolute value of high-frequency data for candidate period 304_1. This integral is the evaluation value for candidate period 304_1. Next, the CPU 112 calculates the integral of the absolute value of high-frequency data from the start of candidate period 304_1 to the start of candidate period 304_2 (hereinafter referred to as the start point integral) and the integral of the absolute value of high-frequency data from the end of candidate period 304_1 to the end of candidate period 304_2 (hereinafter referred to as the end point integral). Then, the CPU 112 determines the evaluation value for candidate period 304_2 by subtracting the start point integral and adding the end point integral from the evaluation value of candidate period 304_1. Subsequently, the CPU 112 similarly determines the evaluation values for each of candidate periods 304_3 to 304_n. Instead of calculating only the difference as described above, the CPU 112 may individually determine the integral value of the absolute value of the high-frequency data for each of the multiple candidate periods 304.
[0051] After determining the evaluation value for each of the multiple candidate periods 304, the CPU 112 selects one candidate period from among the multiple candidate periods 304 by comparing the evaluation values of the multiple candidate periods 304.
[0052] In one example, the CPU 112 may select one candidate period 304 from among several candidate periods 304 that has the highest evaluation value. If there are multiple candidate periods 304 that have the highest evaluation value, the CPU 112 may select one candidate period 304 that is located in the middle of the time axis from among these multiple candidate periods 304. In another example, the CPU 112 may select a predetermined number (for example, 5) from among several candidate periods 304 in descending order of evaluation value, and then select one candidate period 304 that is located in the middle of the time axis from among them.
[0053] Alternatively, the CPU 112 may select one candidate period 304 from among the multiple candidate periods 304 without comparing the evaluation values of the multiple candidate periods 304. For example, the CPU 112 may identify one or more candidate periods 304 from among the multiple candidate periods 304 that have evaluation values exceeding a predetermined threshold, and then arbitrarily select one candidate period 304 from these one or more candidate periods 304. If none of the evaluation values of the multiple candidate periods 304 exceed the predetermined threshold, the CPU 112 may determine that snoring did not occur during this breathing.
[0054] Next, the detection conditions for S203 in Figure 2 will be explained. In one example, the detection condition may be that the evaluation value mentioned above for the analysis period is greater than or equal to a predetermined threshold. In another example, the detection condition may be that the result of frequency analysis of the flow rate data for the analysis period satisfies a predetermined condition. By detecting the occurrence of snoring using the results of frequency analysis in this way, the occurrence of snoring can be detected with high accuracy.
[0055] Referring to Figure 4, an example of frequency analysis of flow rate data will be explained. Graphs 301 and 302 in Figure 4 are identical to graphs 301 and 302 in Figure 3. Assuming that the period 304_t, from time t5 to time t6, is determined to be the analysis period based on the results of S202 in Figure 2,
[0056] CPU 112 generates frequency data by performing signal processing, including a Fourier transform, on the portion of the flow rate data corresponding to period 304_t. This signal processing may include filtering before the Fourier transform. For example, CPU 112 may directly perform a Fourier transform on the portion of the flow rate data corresponding to period 304_t. Alternatively, CPU 112 may generate high-frequency data by applying a filter to the flow rate data to block low-frequency components, and then perform a Fourier transform on the portion of this high-frequency data corresponding to period 304_t. The Fourier transform may be performed using the Fast Fourier Transform (FFT).
[0057] Graph 401 in Figure 4 represents frequency data generated by the Fourier transform. The horizontal axis of Graph 401 represents frequency, and the vertical axis represents spectral amplitude. The detection conditions may also be conditions relating to the frequency data and reference data. For example, the detection conditions may be conditions relating to the relationship between the frequency data and reference data. Reference data is data that has a frequency distribution, and has the frequency distribution that the frequency data can take when it is assumed that snoring is not occurring. Graph 402 in Figure 4 represents the reference data.
[0058] Reference data may be stored in ROM 113 before the start of the method in Figure 2 (for example, before the patient starts sleeping). For example, the manufacturer of the CPAP device 100 may generate reference data based on respiratory flow rate data during sleep when snoring is not occurring from multiple subjects, and store this reference data in ROM 113 of the CPAP device 100. For example, the manufacturer may generate a frequency spectrum by performing a Fourier transform on the respiratory flow rate data (or its high-frequency data) of multiple subjects, and use the average of the spectrum plus twice the standard deviation as the reference data.
[0059] Alternatively, the reference data may be generated during the patient's sleep and stored in RAM114. It is considered unlikely that snoring will occur immediately after the patient falls asleep. Therefore, CPU112 may perform a Fourier transform on the flow rate data (or its high-frequency data) of each breath of the patient for a predetermined period (e.g., 15 or 30 minutes) from the time the patient falls asleep to generate a frequency spectrum, and use the average of this spectrum plus twice the standard deviation as the reference data.
[0060] The detection conditions may include conditions relating to the integral value of the difference between the frequency data and the reference data in one or more frequency bands where the frequency data exceeds the reference data in a specific frequency band. This condition is referred to as the integral value condition. The specific frequency band in the integral value condition may be a frequency band that includes air vibrations caused by snoring, and may be, for example, 10 Hz or higher. There may be no upper limit to the specific frequency band, or it may be 100 Hz or 200 Hz. Alternatively, the specific frequency band may be an upper limit determined by the sampling theorem. For example, if flow rate data is sampled at 500 Hz, the upper limit of the frequency band is 250 Hz.
[0061] In the example in Figure 4, in the frequency band above 10 Hz, the frequency data exceeds the reference data in the four bandwidths 403_1 to 403_4. The multiple bandwidths 403_1 to 403_4 are collectively referred to as bandwidth 403. The following explanation for one bandwidth 403 applies to any of the multiple bandwidths 403_1 to 403_4. Hereafter, the multiple bandwidths 403_1 to 403_4 will simply be referred to as multiple bandwidths 403.
[0062] The integral value of the difference between frequency data and reference data in multiple bandwidths 403 represents the sum of the areas of regions 404_1 to 404_4. The integral value condition may include the condition that the above integral value is greater than or equal to a predetermined threshold. This threshold may be stored in ROM 113 before the start of the method in Figure 2 (for example, when the CPAP device 100 is manufactured). It is known that snoring frequencies are distributed over a certain range. Therefore, by including an integral value condition in the detection conditions, the occurrence of snoring can be detected with high accuracy.
[0063] The detection conditions may include conditions relating to the relative value of the frequency data to the reference data in one or more bands in a specific frequency range where the frequency data exceeds the reference data. The relative value in each of these one or more bands may be the maximum difference of the frequency data to the reference data in one band (i.e., the value obtained by subtracting the reference data from the frequency data), or it may be the maximum ratio of the frequency data to the reference data in that band (i.e., the value obtained by dividing the frequency data by the reference data).
[0064] The detection conditions may include a condition relating to the relative value in one of one or more frequency bands in a specific frequency range where the frequency data exceeds the reference data. This condition is referred to as the primary relative value condition. The specific frequency band in the primary relative value condition may be the same as the specific frequency band in the integral value condition.
[0065] The single bandwidth used in the primary relative value condition may be the bandwidth with the largest relative value among one or more bandwidths in a particular frequency band where the frequency data exceeds the reference data. For example, in the primary relative value condition, bandwidth 403_4 may be used among a plurality of bandwidths 403. The primary relative value condition may include the condition that the relative value of this bandwidth is greater than or equal to a predetermined threshold. This threshold may be stored in ROM 113 before the start of the method in Figure 2 (for example, when the CPAP device 100 is manufactured).
[0066] The detection conditions may include a condition relating to the relative value in one of the bands in a specific frequency range where the frequency data exceeds the reference data, but which is different from the band used in the primary relative value condition. This condition is referred to as the secondary relative value condition. The specific frequency band in the secondary relative value condition may be the same as the specific frequency band in the integral value condition.
[0067] One of the bandwidths used in the secondary relative value condition may be the bandwidth with the second largest relative value among one or more bandwidths in a particular frequency band where the frequency data exceeds the reference data. For example, in the secondary relative value condition, bandwidth 403_2 may be used among a plurality of bandwidths 403. The secondary relative value condition may include the condition that the relative value of this bandwidth is greater than or equal to a predetermined threshold. This threshold may be stored in ROM 113 before the start of the method in Figure 2 (for example, when the CPAP device 100 is manufactured).
[0068] CPU112 may use only one of the integral value condition, the primary relative value condition, and the secondary relative value condition as a detection condition, or it may use a combination of two or more of these three conditions as a detection condition. For example, CPU112 may determine whether each of the integral value condition, the primary relative value condition, and the secondary relative value condition is satisfied, and determine that the detection condition is satisfied based on the fact that at least one of these three conditions is satisfied. For example, CPU112 may determine whether each of the integral value condition and the primary relative value condition is satisfied, and determine that the detection condition is satisfied based on the fact that at least one of these two conditions is satisfied.
[0069] In the determination at S203 in Figure 2, the CPU 112 may adjust the value of the reference data and use the adjusted reference data to determine whether the detection conditions are met. It is known that the magnitude of the frequency data (i.e., spectral amplitude) is proportional to the flow rate. Therefore, the CPU 112 may adjust the value of the reference data based on the inspiratory flow rate of each breath.
[0070] For example, the CPU 112 determines a representative value of the flow rate data during the inspiratory period of each breath. The representative value may be the maximum value of the flow rate data during the inspiratory period, the average value of the flow rate data during the inspiratory period, the integral value of the flow rate data over the inspiratory period, or any other value. The ROM 113 or RAM 114 stores the reference flow rate in association with the reference data. The CPU 112 may adjust the value of the reference data by multiplying the value obtained by dividing the representative value of the flow rate data during the inspiratory period of each breath by the reference flow rate (i.e., multiplying by the spectral amplitude at each frequency of the reference data).
[0071] Referring to Figure 5, a method performed by the CPAP device 100 to adjust the therapeutic pressure is described. Detection of snoring may be used, in addition to or instead of adjusting the therapeutic pressure, to adjust the temperature and humidity supplied to the patient, or for other purposes. In the following description, each step of the method in Figure 5 is performed by the CPU 112 (e.g., its therapeutic pressure control unit 118). Specifically, each step is performed by the CPU 112 executing a program read from the RAM 114. Alternatively, at least some steps of the method in Figure 5 may be performed by a dedicated circuit such as an ASIC. The method in Figure 5 may be initiated when the patient instructs the CPAP device 100 to start operation (e.g., therapeutic operation during sleep), when the CPAP device 100 detects the patient falling asleep, or in response to other events. In the method in Figure 5, the therapeutic pressure is changed in response to the detection of snoring. In addition, the therapeutic pressure may be changed in response to the detection of other events (e.g., the passage of a predetermined time, the occurrence of apnea).
[0072] In S501, the CPU 112 determines whether it has detected snoring. If the CPU 112 determines that it has detected snoring (YES in S501), it proceeds to S502; otherwise, it repeats S501. In this way, the CPAP device 100 waits until it detects snoring. The occurrence of snoring in the patient is detected in real time while the patient is sleeping.
[0073] In S502, CPU112 increases the current therapeutic pressure by a predetermined value or ratio. Therapeutic pressure may refer to the pressure of the air supplied to the patient.
[0074] In S503, CPU112 determines whether the snoring is continuing. If CPU112 determines that the snoring is continuing (YES in S503), it proceeds to S504; otherwise (NO in S503), it proceeds to S501.
[0075] If snoring ceases (i.e., stops), it is considered that the airway obstruction has been resolved by increasing the therapeutic pressure. Therefore, CPU112 maintains this therapeutic pressure. On the other hand, if snoring continues even after increasing the therapeutic pressure, it is considered that the cause of snoring is not airway obstruction. Therefore, in S504, CPU112 performs a predetermined process. For example, CPU112 may decrease the therapeutic pressure. For example, CPU112 may return the therapeutic pressure to the value before it was increased in S502. In addition to decreasing the therapeutic pressure, or instead, CPU112 may output an alert indicating that snoring not caused by airway obstruction has occurred. The alert may be output to a doctor or other medical professional. Snoring not caused by airway obstruction can be caused by nasal congestion or mouth breathing, etc. Based on this alert, the doctor or medical professional may provide guidance on nasal congestion or mask use.
[0076] The treatment pressure may have an upper limit set by a physician, for example. The CPU 112 does not need to increase the treatment pressure if it detects snoring in S501 when the treatment pressure has reached the upper limit. The treatment pressure may be reset to an initial value after each treatment (for example, every day). This initial value may be the minimum treatment pressure set by a physician, for example. The CPU 112 may also decrease the treatment pressure if no respiratory disturbance events such as snoring, apnea, or hypopnea are detected for a certain period of time or longer.
[0077] In the embodiment described above, the snoring detection method shown in Figure 2 is performed by a CPAP device 100. Alternatively, the snoring detection method shown in Figure 2 may be performed by a sleep apnea syndrome (SAS) testing device. Such a testing device does not have the function of supplying compressed air to the patient. The SAS testing device that performs the snoring detection method shown in Figure 2 is also a medical device that has a snoring detection function.
[0078] In the embodiment described above, the occurrence of snoring in the patient is detected during the patient's sleep. Alternatively, the snoring detection method shown in Figure 2 may be performed after the patient wakes up, using flow data acquired during the patient's sleep. That is, the occurrence of snoring in the patient may be detected after the patient wakes up. This makes it possible to identify the period during which snoring occurs, which can then be used for the diagnosis of the patient.
[0079] <Summary of Embodiments> [Item 1] It is a medical device, A generation means for generating high-frequency data from flow rate data representing the time series of the patient's respiratory flow rate, A determination means for determining the evaluation value of the high-frequency data for each of the multiple periods, A selection means for selecting one of the multiple periods based on the evaluation values for the multiple periods, A medical device comprising: detection means for detecting the occurrence of snoring in the patient by analyzing a portion of the flow rate data for one selected period. According to this section, by analyzing flow rate data for the selected period, snoring can be detected with high accuracy. [Item 2] The medical device according to item 1, wherein the selection means selects one of the multiple periods by comparing the evaluation values of the multiple periods. According to this section, by comparing evaluation values, an appropriate period can be selected, allowing for accurate detection of snoring. [Item 3] The aforementioned multiple periods are included in the inspiratory period, as described in item 1 or 2. According to this section, snoring can be detected with high accuracy by analyzing flow rate data during the inspiratory period, which is likely to include snoring. [Item 4] Each of the aforementioned periods partially overlaps with another of the aforementioned periods, and the medical device is one of the medical devices described in any one of items 1 to 3. According to this item, it is possible to increase the number of periods within a given time length, thus enabling the determination of multiple evaluation values at a finer granularity. [Item 5] A medical device as described in any one of items 1 to 4, wherein the aforementioned multiple periods are all of the same duration. According to this item, evaluation values can be determined using the same scale. [Item 6] A medical device as described in any one of items 1 to 5, wherein the starting points of the aforementioned multiple periods are aligned at the same time interval. According to this section, you can select multiple periods from the target period in a balanced manner. [Item 7] The aforementioned flow rate data is digital data. The medical device described in item 6, wherein the interval between the starting points of the aforementioned multiple periods is wider than the sampling interval of the flow rate data. According to this item, the processing load required to detect snoring can be reduced. [Item 8] The aforementioned evaluation value has a larger value the higher the probability that snoring is occurring. The medical device according to any one of items 1 to 7, wherein the selection means selects one period from the plurality of periods in which the evaluation value is the maximum. This item allows for the accurate selection of periods when snoring is most likely to occur. [Item 9] The medical device according to any one of items 1 to 8, wherein the determination means determines the evaluation value based on the integral value of the absolute value of the high-frequency data during each period. According to this item, the evaluation value can be determined with high accuracy based on the characteristic of snoring that it continues for a certain period of time. [Item 10] The detection means is a medical device according to item 1, which detects the occurrence of snoring in the patient in real time while the patient is sleeping. According to this section, measures to address snoring can be taken in real time while you are sleeping. [Item 11] The aforementioned medical device is a continuous positive airway pressure device. The medical device according to item 10, further comprising pressure control means for increasing the pressure of air supplied to the patient based on the detection of the occurrence of snoring by the patient. According to this item, it is possible to prevent the occurrence of respiratory problems. [Item 12] A program for causing a computer to function as one of the means of a medical device as described in any one of items 1 through 11. According to this section, a program is provided to realize a medical device having the effects described above. [Item 13] A snoring detection method performed by a computer, A generation process that generates high-frequency data from flow rate data representing the time series of the patient's respiratory flow rate, A determination step for each of multiple periods, in which the evaluation value of the high-frequency data in each period is determined, A selection step of selecting one of the multiple periods based on the evaluation values of the multiple periods, A snoring detection method comprising: a detection step of detecting the occurrence of snoring in the patient by analyzing a portion of the flow rate data for one selected period. According to this section, by analyzing flow rate data for the selected period, snoring can be detected with high accuracy.
[0080] The invention is not limited to the embodiments described above, and various modifications and changes are possible within the scope of the gist of the invention. [Explanation of Symbols]
[0081] 100 CPAP device, 303 Inspiratory period, 304 Candidate period, 403 Bandwidth
Claims
1. It is a medical device, A generation means for generating high-frequency data from flow rate data representing the time series of the patient's respiratory flow rate, A determination means for determining the evaluation value of the high-frequency data for each of the multiple periods, A selection means for selecting one of the multiple periods based on the evaluation values for the multiple periods, The system includes a detection means for detecting the occurrence of snoring in the patient by analyzing a portion of the flow rate data for one selected period, In the aforementioned high-frequency data, frequency components below a certain frequency, which are lower than the frequency components caused by snoring, are blocked from the flow rate data. The aforementioned evaluation value is a value used to assess the likelihood that the patient is snoring, in a medical device.
2. The medical device according to claim 1, wherein the detection means performs frequency analysis on a portion of the flow rate data for one selected period.
3. The medical device according to claim 1, wherein the aforementioned plurality of periods are included in the inhalation period.
4. The medical device according to claim 3, wherein the evaluation value has a larger value the higher the probability that snoring is occurring.
5. The aforementioned flow rate data is digital data. The medical device according to claim 4, wherein the interval between the starting points of the plurality of periods is wider than the sampling interval of the flow rate data.
6. The medical device according to claim 5, wherein the detection means detects the occurrence of snoring in the patient in real time while the patient is sleeping.
7. The aforementioned medical device is a continuous positive airway pressure device. The medical device according to claim 6, further comprising pressure control means for increasing the pressure of air supplied to the patient based on the detection of the occurrence of snoring by the patient.
8. The medical device according to claim 7, wherein the selection means selects one period from the plurality of periods in which the evaluation value is the maximum.
9. The medical device according to claim 7, wherein the determination means determines the evaluation value based on the integral value of the absolute value of the high-frequency data during each period.
10. The medical device according to claim 7, wherein the selection means selects one of the plurality of periods by comparing the evaluation values of the plurality of periods.
11. The medical device according to claim 7, wherein each of the aforementioned multiple periods partially overlaps with another of the aforementioned multiple periods.
12. The medical device according to claim 7, wherein all of the aforementioned periods are of the same duration.
13. The medical device according to claim 7, wherein the starting points of the aforementioned multiple periods are aligned at the same time interval.
14. The medical device according to claim 3, wherein the aforementioned multiple periods are included only in the inhalation period.
15. A program for causing a computer to function as one of the means of a medical device according to any one of claims 1 to 14.
16. A snoring detection method performed by a computer, A generation process that generates high-frequency data from flow rate data representing the time series of the patient's respiratory flow rate, A determination step for each of multiple periods, in which the evaluation value of the high-frequency data in each period is determined, A selection step of selecting one of the multiple periods based on the evaluation values of the multiple periods, The system includes a detection step of detecting the occurrence of snoring in the patient by analyzing a portion of the flow rate data for one selected period, In the aforementioned high-frequency data, frequency components below a certain frequency, which are lower than the frequency components caused by snoring, are blocked from the flow rate data. A snoring detection method wherein the evaluation value is a value for evaluating the likelihood that the patient is snoring.