Central apnea detection

Abstract

Existing apnea monitors are unable to detect clinically significant apnea events because they cannot distinguish cardiac artifacts from the thoracic impedance signal and therefore are unable to detect pauses in breathing. A system and method are disclosed that provides improved apnea detection, particularly in the neonatal setting. The disclosed technique filters out cardiac artifacts from the thoracic impedance signal and allows the probability of an apnea event to be determined. Detection of an apnea event can then be used to trigger an alarm, initiate automatic physical stimulation of the patient, or both.

Description

[Technical field]

[0001] [CROSS-REFERENCE TO RELATED APPLICATIONS] This patent application claims priority to U.S. Provisional Patent Application No. 63 / 265,817, entitled "Central Apnea Detection," filed December 21, 2021. The disclosure of the prior application is considered part of this patent application and is incorporated by reference into this patent application.

[0002] [Technical field] The present invention relates to the technical field of apnea detection, in particular to central apnea detection. [Background technology]

[0003] Apnea is a temporary cessation of breathing. Apnea may occur at any stage of development and is particularly common in premature infants, where it is sometimes called apnea of ​​prematurity (AOP). AOP is quite distinct from adult sleep apnea. AOP is a developmental disorder, but the reasons behind the propensity of apnea in immature infants are not entirely clear. The pathogenesis of AOP is poorly understood, but immature pulmonary reflexes and breathing responses to hypoxia and hypercapnia may contribute to the occurrence or severity of AOP. It may also be exacerbated by coexisting factors or conditions.

[0004] AOP is a significant and common clinical problem and is often the rate-limiting process in Neonatal Intensive Care Unit (NICU) discharge. Accurate detection of clinically significant neonatal apnea episodes using existing chest impedance monitoring is clinically essential.

[0005] Apnea is a serious clinical event that requires immediate medical attention. However, existing monitoring for apnea, particularly for neonatal infants, is inadequate in that it misses many serious events. For this reason, they often fail to recognize apnea and therefore fail to provide a warning signal to alert NICU staff to the fact that the infant is not breathing.

[0006] Thus, there is a need in the art for techniques that provide improved detection of apnea. [Brief description of the drawings]

[0007] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate implementations of apparatus and methods consistent with the invention and, together with the description of the invention, serve to explain advantages and principles consistent with the invention.

[0008] [Figure 1] 1 is a screenshot illustrating a graphical user interface for displaying indicators of apnea events according to one embodiment. [Diagram 2]1 is a screenshot illustrating a graphical user interface for displaying indicators of apnea events according to one embodiment. [Diagram 3] 1 is a flow chart illustrating a technique for detecting apnea events according to one embodiment. [Figure 4] FIG. 1 is a block diagram illustrating a system for collecting, archiving, and processing arbitrary data in a medical environment according to one embodiment. [Diagram 5] 5 is a block diagram illustrating a computer server for use in the system of FIG. 4. [Figure 6] 1 is a screenshot illustrating a graphical user interface for tracking apnea events according to one embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] In the following description, for the purpose of explanation, numerous specific details are set forth in order to provide a thorough understanding of the present invention. However, it will be apparent to one skilled in the art that it is possible to practice the present invention without the use of these specific details. In other instances, structures and devices are shown in block diagram form to avoid obscuring the present invention. Reference to a number without a subscript is understood to refer to all of the subscripted examples corresponding to the referenced number. Furthermore, the language used in this disclosure has been selected primarily for readability and instructional purposes, and may not be selected to define or limit the subject matter of the invention, which must be determined by reliance on the claims. When "an embodiment" or "an embodiment" is mentioned in this specification, it means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the present invention, and multiple references to "an embodiment" or "an embodiment" should not be understood to all necessarily be referring to the same embodiment.

[0010] Although some of the following description is written in terms relating to software or firmware, embodiments may implement the features and functionality described herein through software, firmware, or hardware, including any combination of software, firmware, and hardware, as appropriate. Reference to a daemon, driver, engine, module, or routine should not be construed as implying a limitation of the embodiment to any type of implementation.

[0011] Apnea is a serious clinical event that requires medical attention within seconds and is very common in premature infants. Newborns tend to have irregular breathing patterns and stop breathing for several seconds at a time. Physicians vary in how they determine which events are clinically significant. A general rule of thumb is to consider apnea a clinical event if (a) the cessation of breathing lasts longer than 20 seconds, or (b) the cessation of breathing lasts longer than 10 seconds and is accompanied by either bradycardia (a slowing of the heart) or oxygen desaturation. In neonates, bradycardia is typically defined as a heart rate less than 100 beats per minute, and oxygen desaturation is typically defined as an SpO2 less than 80%. Bradycardia and desaturation may occur without respiratory pauses, and further deterioration is likely occurring when these vital thresholds are crossed within the time period following the onset or end of a significant neonatal respiratory pause event.

[0012] Clinical apneas occur in more than half of infants weighing less than 1500 g at birth and in nearly all infants weighing less than 1000 g at birth. Apnea may be the cause or result of a number of other clinical problems, such as hypoxemia, hypoglycemia, neurological injury, or sepsis. Apnea is also a common manifestation of an immature neurological system in infants who were born very preterm but otherwise have no clinical pathology. Three types of apnea are common in premature infants: obstructive apnea, central apnea, and mixed apnea. Obstructive apnea is a blockage of the airway, typically accompanied by struggling or thrashing movements of the infant. Central apnea is a cessation of respiratory drive, in which the infant usually remains extremely still, without any effort to breathe. Mixed apnea typically begins with an obstructive event, then transforms into a central apnea. In contrast, there is evidence that central apnea can also become obstructive apnea, so some sources suggest that the distinction between the two types should not be made too strictly. In many cases, apnea is combined with or induces bradycardia (a significant decrease in heart rate).When accompanied by bradycardia, oxygen desaturation, or both, these apneas are serious clinical events that require immediate medical attention and investigation for associated pathology. Furthermore, central apneas are considered to indicate immaturity of control of respiration, and discharge from the NICU is typically delayed until apnea has been absent for between three and eight days. As a result of these regulations, infants can be safely discharged home without monitoring devices and clinicians, decreasing the likelihood that they will experience an apneic episode.

[0013] Bedside monitors attempt to detect apnea based on continuous monitoring of chest impedance (CI) physiological data. Using electrodes used to monitor the electrocardiogram (ECG), a small high frequency voltage (e.g., 52.6 kHz) is applied to the chest and the resulting high frequency current is measured. The measured impedance Z is equal to the applied voltage V divided by the observed current. The measured impedance is typically in the range of 50 to 300 ohms and is related to the conductivity of the muscles, skin, other tissues, and the contact between the electrodes and the skin. When the infant is breathing, the impedance varies with each breath. In the event of a central apnea, the change in chest impedance decreases as breathing pauses. However, the beating of the heart also causes impedance fluctuations each time the heartbeat pumps blood from the thorax. In this way, even during an apnea event, thoracic impedance may fluctuate and reach a threshold for detection as breathing, foiling the monitor's apnea alarm.

[0014] Detecting respiratory pauses in neonates is particularly difficult because the neonate's more malleable rib cage moves with the heartbeat. Chest impedance leads can mistake the resulting change in rib cage diameter for respiration. Current apnea detectors do not address this problem, so in these cases clinicians must rely on alarms that indicate subsequent deterioration of the patient, such as bradycardia.

[0015] As a result of these cardiac artifacts, some episodes of central neonatal apnea are missed entirely, and some of these episodes are severe. Some studies have found that conventional apnea monitors miss more than 10% of even extreme apnea events due to cardiac artifacts in the thoracic impedance signal.

[0016] Techniques described below remove these cardiac artifacts from the chest impedance waveforms used to measure respiration, and although described below with respect to real-time physiological data, the same techniques may be used on historical physiological data for retrospective analysis.

[0017] Due to increased detection of apnea events, in certain embodiments, automated dialogue may be utilized to stimulate premature infants during apnea events.

[0018] The following disclosure describes techniques for improving the detection of central apnea, more specifically, analyzing a standard CI signal used to monitor respiration rate to filter out the contribution to the thoracic impedance signal arising from the beating heart and provide indications of respiratory pause events.

[0019] FIG. 1 is a screenshot illustrating a graphical user interface 100 in which an indication of central apnea is displayed according to one embodiment. In the graphical user interface 100, two waveform lanes are illustrated: a CI waveform 110 and an ECG waveform 120. The exemplary waveform data in FIG. 1 is not intended to be medically accurate. In this example, the patient is experiencing an apnea condition, as shown by line 130, even though the CI waveform 110 is not flat. The CI waveform is experiencing variations as a result of the heartbeat, as shown by ECG waveform 120. This is known to be because both waveforms periodically repeat at the same rate. In some of the embodiments, one or both of an audible alarm indication and a visual alarm indication may be generated in response to a determination that an apnea condition exists, in addition to the indication shown in line 130. In other embodiments, an automatic response may be triggered to physically stimulate the infant, which can cause a cessation of the apnea event.

[0020] 2 is a screenshot illustrating a graphical user interface 200 according to an embodiment similar to the graphical user interface 100. In this example, graph 220 is an exemplary CI waveform, graph 230 shows heart rate determined from photoplethysmogram data, graph 240 shows heart rate determined from ECG data, graph 250 shows SpO2 data determined from photoplethysmogram data, and graph 260 shows respiration rate calculated based solely on the CI waveform. In this example, apnea events indicating that the patient has ceased breathing and meets the bradycardia and oxygen desaturation criteria are shown as indicated by horizontal line 210. In an embodiment, apnea events are shown to begin only after a threshold time (e.g., 20 seconds) during which pauses in breathing have been identified and bradycardia and oxygen desaturation have occurred. This reduces nuisance alarm indications for brief apnea events that terminate without intervention.

[0021] 1 and 2 both illustrate a real-time display of information indicating the presence of an apnea event, some of the embodiments in which the real-time data is archived for research or clinical purposes may provide the ability to investigate historical patient data, analyze that data, and indicate the presence of apnea events in that historical data. Such historical investigations may be useful to detect unnoticed prior apneas and provide clinicians with better information for assessing the health of a patient.

[0022] For clarity, the following description focuses on cessation of breathing events that are apnea events. However, different medical personnel may define an apnea event as a cessation of breathing without other factors or in combination with one or more other factors, such as bradycardia or a drop in oxygen saturation. The system may be configured for any combination of these additional factors as needed. For example, a patient may stop breathing for longer than a threshold time but maintain a normal heart rate and oxygen saturation, in which case some of the configurations of the system described below may not indicate an apnea event. An embodiment of the system described below may allow the clinician to select the combination of criteria used to determine an apnea event.

[0023] Inputs to the central apnea detector may include the CI waveform, the ECG waveform, and vitals of heart rate and oxygen saturation (SpO2). Each of the signals can be subjected to cleaning to remove data artifacts, if required.

[0024] As described below, it is possible to identify the cardiac artifact component of thoracic impedance based on its shared phase (frequency progression) with any signal (such as an ECG signal) that captures heart movement. The first step is to model the progression of the cardiac phase. This modeling may be done in several ways, using any signal that captures heart movement. For example, the cardiac phase may be calculated by utilizing the phase of the Hilbert transform applied to the photoplethysmogram waveform.

[0025] In one embodiment, to determine the cardiac phase, standard peak / trough detection is performed to find the R peaks in the ECG signal that occur in each cardiac cycle. A cardiac phase signal φ is created to model the heart's frequency at any given time by setting the phase to 0 at each R peak time and having the phase increase linearly with time to 2π=0 radians for samples in between each set of consecutive R peaks. In this way, the cardiac phase signal φ for a segment of data containing N samples is calculated as an N×1 vector

number

[0026] A measured chest impedance signal (CI) corresponding to the cardiac phase init (i.e., cardiac artifacts) are then determined as a function of cardiac phase, CI init may be extracted using a Fourier series approximation of: Although the following description is written in terms of Fourier series, other frequency or time domain analysis and approximation techniques may be used.

[0027] The Fourier series approximation uses M harmonics stored as a 1×M vector H=[1 2 … M].

[0028] In some embodiments, M=3 adequately captures the frequency content contained in the CI being cleaned, although other numbers of harmonics may be used if desired.

[0029] This is the N × M cardiac phase harmonic matrix φH

number

[0030] The frequency content F of the cardiac artifact Fourier series approximation can be calculated from the cardiac phase and is given by the N × 2M matrix

number

[0031] The cardiac artifact frequency components and the chest impedance signal can then be used to determine the Fourier series coefficients of the cardiac artifact. These Fourier series coefficients are stored in a 2M×1 vector

number

[0032] The Fourier series coefficients of the cardiac artifact are

number

[0033] Cardiac Artifact CI art The approximation formula for is then:

number

[0034] Cardiac Artifact CI art may then be removed from the measured thoracic impedance signal, yielding a filtered thoracic impedance signal CI filt =CI init -CI art Generate.

[0035] Once cardiac artifacts have been removed from the thoracic impedance, it is possible to identify events of respiratory pauses, bradycardia, and desaturation using standard signal processing techniques of the filtered thoracic impedance and thresholding of the vitals of heart rate and oxygen saturation. When there is little pulmonary movement (small standard deviation) in the filtered thoracic impedance signal, the neonate is considered not breathing and events of respiratory pauses are identified.

[0036] The apnea event probability is:

number

[0037] In the above formula, σ is the CI filt The above Fermi function parameters are illustrative and are only one example of the present invention. Other Fermi functions or other types of functions may be used to calculate the CI filtIn some of the embodiments, when the probability of an apnea event exceeds a predetermined threshold, a system is triggered that displays an alarm indication of an apnea event in a graphical user interface, such as the graphical user interface illustrated in Figures 1 and 2. Furthermore, a set of alarm criteria may be developed based on the relative probability of an apnea event over a defined period of time. Additionally, in embodiments where one or more of bradycardia and oxygen desaturation are considered apnea criteria, the results of measuring heart rate and SpO2 data may be combined with the probabilities calculated above to trigger an alarm indication.

[0038] Respiratory pauses accompanied by bradycardia and oxygen desaturation events are identified when these vital thresholds are passed within a defined period of time at the beginning or end of the respiratory pause event. The techniques described herein may be used to determine apnea events in children and adults, although the specific criteria, such as heart rate and oxygen desaturation, may differ in non-neonatal populations.

[0039] The resulting improved system for detecting apnea events avoids missing apnea events that are masked by the patient's heartbeat in existing apnea detection systems.

[0040] 3 is a flow chart illustrating a technique 300 for detecting and automatically acting on apnea events according to one embodiment. In block 310, the system receives real-time CI and cardiac data from multiple sensors attached to the patient. In block 320, a phase analysis of the cardiac data (e.g., ECG) is performed as described above to generate a phase signal Φ and a cardiac artifact phase harmonic matrix ΦH. In block 330, the cardiac artifact component CI of the chest impedance is calculated using the algorithm described above. art (cardiac artifact component CI art ) is then calculated in block 340. init (measured chest impedance CI init ) from cardiac artifact component CI art (cardiac artifact component CI art ) is subtracted to obtain the filtered thoracic impedance CI filt (filtered chest impedance CI filt ) is calculated in block 350. filtindicates a respiratory pause, an alarm may be triggered in a graphical user interface, such as the graphical user interface illustrated in Figures 1 and 2, as an audible alarm, or any other method known in the art for generating an alarm. In some of the embodiments, a respiratory pause may be detected when the probability of an apnea event, calculated as described above, reaches or exceeds a predefined threshold or meets a set of criteria for a predefined duration. In some of the implementations, rather than using a strict time threshold, an overall period that meets the threshold is evaluated. For example, if there are two 12-second periods that meet the threshold, separated by one second, the two segments may be combined into one respiratory pause event that spans 25 seconds.

[0041] If no pause in breathing is detected in block 350, the technology continues to receive CI and cardiac data in block 310. If a pause in breathing is detected in block 350, an alarm is signaled in block 360. If available, an automatic action may be taken to physically stimulate the patient in response to an apnea event, such as generating vibrations, blowing air, or initiating other physical stimuli that may cause the patient to resume breathing, in block 370. If a system implementing the technology 300 is configured to use one or more of bradycardia and desaturation as part of the evaluation of thresholds for an apnea event, the technology 300 may include other actions not shown in FIG. 3 for detecting bradycardia and desaturation from heart rate and SpO2 data received from a patient monitor. If automatic physical stimulation of the patient is not available, a clinical staff member may approach the patient and manually provide a tactile stimulus, such as tapping the sole of the foot or rubbing the back. If spontaneous breathing is not initiated by physical stimulation, the apneic event may require additional intervention, such as the administration of oxygen.

[0042] 4 is a block diagram illustrating a system 400 for collecting, archiving, and processing arbitrary data in a healthcare environment according to one embodiment. System 400 is described in detail in U.S. Patent No. 10,8892,045, entitled "Distributed Grid Computing Platform for Collecting, Archiving, and Processing Arbitrary Data in a Healthcare Environment," which is incorporated by reference in its entirety for all purposes.

[0043] As shown, there are five types of servers: data acquisition (DAQ) server 487, informatics server(s) 480, database server 485, HL7 server 483, and web server 490. Any number of any type of servers may be deployed as desired. All of the servers 480-490 connect to each other and to bedside monitors via one or more hospital networks 430. Although shown as a single hospital Ethernet network 430, any number of interconnected networks may be used using any desired network protocols and technologies.

[0044] Also connected to the hospital network 430 are a number of bedside monitors for monitoring physiological data of the patient in bed 410. The bedside monitors may include a network-connected monitor 420A capable of delivering digital physiological data to the hospital network 430, a serial device 420B that is not directly connected to the network but generates digital data, and an analog device 420C that is not directly connected to the network but generates analog data. Communication boxes 440A and 440B typically allow the serial device 420B and the analog device 420C to be connected to the hospital network 430, respectively, via a network switch 450. In addition, a branch office 460 may also be connected to the network 430 via the network switch 450 to perform data manipulation and time synchronization, as described below. Any number of bedside monitors 420 may be used for the patient in bed 410 as determined by the physician and other clinical staff as desired.

[0045] FIG. 4 illustrates bedside monitors and associated communication devices that are directly or indirectly connected to a hospital network 430; as part of the system 400, remote bedside monitoring devices, such as home monitoring devices that are indirectly connected to the hospital network 430 by the Internet or other communication technologies, may also be used.

[0046] Additionally, one or more research computers 470 may be connected directly or indirectly to the hospital network 430 to allow researchers to access aggregate data collected from the bedside monitors 420 for performing analysis and development.

[0047] Web server 490 is configured to communicate with personal devices such as laptop 495A, tablet 495B, or smartphone 495C via a web browser interface using the HyperText Transport Protocol (HTTP).

[0048] 5, an exemplary system 500 for use as one of the servers 480-490 is illustrated in block diagram form. The exemplary computer 500 includes a system unit 510 that may be optionally connected to an input device or system 560 (e.g., keyboard, mouse, touch screen, etc.) and a display 570. A program storage device (PSD) 580 (sometimes referred to as a hard disk) is included in the system unit 510. A network interface 540 for communicating over a network with other computing devices and enterprise infrastructure devices (not shown) is also included in the system unit. The network interface 540 may be included in the system unit 510 or may be external to the system unit 510. In either case, the system unit 510 will be communicatively coupled to the network interface 540. The program storage device 580 represents any medium of non-volatile storage including, but not limited to, all forms of optical and magnetic including solid state storage elements including removable media, and may be contained within the system unit 510 or may be external to the system unit 510. The program storage device 580 may be used for storing software that, when executed, controls the system unit 510, data used by the computer 500, or both.

[0049] The system unit 510 may be programmed to execute a method according to the present disclosure (an example of which is shown in FIG. 3). The system unit 510 includes a processor unit (PU) 520, an input / output (I / O) interface 550, and a memory 530. The processor unit 520 may include any programmable controller device, such as a microprocessor available from Intel Corporation and other manufacturers. The memory 530 may include one or more memory modules and may include random access memory (RAM), read only memory (ROM), programmable read only memory (PROM), programmable read and write memory, and solid state memory. Those skilled in the art will recognize that the PU 520 may also include some internal memory, including, for example, cache memory.

[0050] FIG. 6 is a screenshot illustrating one exemplary graphical user interface 600 for monitoring historical information regarding a patient's apnea events according to one embodiment. Each apnea event may be detected and recorded using the systems and techniques described above and then transmitted to the graphical user interface 600 for display. Aggregate information regarding the apnea events may also be transmitted to the graphical user interface 600 for display. In this exemplary graphical user interface 600, a first region 610 provides historical information regarding the number of recent apnea events, in this example, the time from the last apnea event and the number of apnea events in the last 24 hours and the last 5 days. The displays of "last 24 hours" and "last 5 days" are merely examples and may provide other useful historical periods in addition to or instead of the historical periods illustrated in FIG. 6. A second region 620 provides historical information regarding the total length of apnea events in the last 12 hours. In this example, recordings related to two types of oxygen saturation (SpO2), SpO2 > Max SpO2 Threshold and SpO2 < Min SpO2 Threshold, are displayed. Similarly, recordings related to two types of heart rate (HR), HR > Max HR Threshold and HR < Min HR Threshold, are displayed. In addition to or instead of the information illustrated in FIG. 6, other types of historical information may be displayed in the second region 620.

[0051] Below the first region 610 and the second region 620 is a region that provides historical information regarding individual apnea events. The timeline 650 in this example provides an indication of apnea events in the 24-hour period up to the selected "time zero". A pair of calipers 640 may be placed on the timeline to select a specific window for a detailed display within the region 660 below the timeline 650. The export button 630 enables exporting information regarding the selected time window to an electronic medical record (EMR) system.

[0052] Details about each apnea event may be displayed in the detailed information area 660. As shown in FIG. 6, the information displayed for each event may include the length of the respiratory pause, the minimum HR occurring during a specified time period associated with the event, the minimum oxygen saturation level occurring during a specified time period associated with the event, whether the event detected by the system has been determined by the clinician to be an actual apnea event, the type of clinical intervention being performed in response to the apnea event, such as compressions, foot taps, or no intervention, a designation of the type of apnea event, such as obstructive, central, mixed, or unknown, and the start time of the event. Other detailed information may be provided in the user interface 600 as needed.

[0053] The regions and arrangements of regions in the exemplary graphical user interface 600 of FIG. 6 are exemplary and illustrative only, and other regions and arrangements of regions may be provided as desired. Some of the elements of the graphical user interface are illustrated as drop-down type user interaction elements, but other types of user interaction elements may be used as desired. Color may be used to improve user usability. For example, in some implementations, an indicator in the timeline 650 of a particular apnea event may be displayed in a different color than other indicators, while the corresponding row in the details area 660 may be highlighted or shown in a different color.

[0054] In clinical settings without the disclosed patient monitoring system, nurses or other clinical staff manually record apnea events occurring during a shift on a flow sheet by handwriting the data on a piece of paper or typing the data into an electronic form. In either case, the flow sheet requires clinical attention and is subject to human transcription errors. The disclosed system allows for automatically populating clinical flow sheets using apnea event information, which can reduce the possibility of human error in recording the information and significantly reduce manual work by nurses or other clinical staff. The clinical flow sheet may include respiratory pause length, start time, SpO2, and HR information. Other items such as clinical adjudication, type of apnea, and type of clinical intervention may still be entered by the nurse, but in a standardized form by selecting from a list of possible entries. For example, clinical staff may report that it may be difficult to detect the end of an apnea event and therefore calculate the length of the apnea event. By automatically detecting both the start and end of an apnea event, the disclosed system is able to automatically calculate the length of the event, further reducing the burden on clinical staff and reducing manual data entry errors.

[0055] By collecting objective data on apnea events and improving existing alarm systems rather than relying on detection by clinical personnel, it is possible to provide better tracking of apnea events. This improved tracking would result in fewer false alarms as well as detection of apnea events that are not currently detected in clinical practice. One consequence of such better data would be reduced clinical stays for infants, reduced childcare costs for both the clinical facility and the parents, and improved quality of life for both infants and parents. For example, reducing false alarms for apnea events would allow patients to be released sooner, since patients are often not released until a period of time has passed without apnea.

[0056] Although certain exemplary embodiments have been described in detail and illustrated in the accompanying drawings, it is to be understood that such embodiments are exemplary only and that no departure from the basic scope thereof, as determined by the following claims, is to be construed as falling within the spirit and scope of the present invention.

Claims

1. 1. A method for detecting apnea in a patient, the method comprising: receiving thoracic impedance physiological data and cardiac physiological data from a bedside monitoring device associated with the patient; modeling cardiac phase using said cardiac physiological data; calculating a frequency domain or a time domain approximation of a cardiac artifact based on the representation of the cardiac phase and the thoracic impedance physiological data; removing the cardiac artifacts from the thoracic impedance physiological data to generate filtered thoracic impedance data; calculating a probability of an apnea event based on the filtered thoracic impedance data; generating an alert indication responsive to meeting a set of criteria based on the probability of the apnea event over a time period. method.

2. The method of claim 1 , further comprising triggering an automatic physical stimulation of the patient responsive to the apnea event.

3. The method of claim 1 , further comprising the step of calculating a length of the apnea event.

4. recording apnea events for the patient; transmitting historical information regarding apnea events to a graphical user interface; calculating aggregate information regarding the recorded apnea events for the patient; 10. The method of claim 1, further comprising the step of: transmitting the aggregate information regarding the recorded apnea events to the graphical user interface.

5. 5. The method of claim 4, wherein recording apnea events for the patient includes recording oxygen saturation and heart rate data associated with the apnea events.

6. receiving a heart rate measurement and an oxygen saturation measurement from the bedside monitoring device; The method of claim 1 , wherein the set of criteria includes heart rate and oxygen saturation criteria.

7. The method of claim 1, wherein the patient is a neonate.

8. 1. A patient monitoring system for detecting an apnea event in a patient, the patient monitoring system comprising: a bedside monitoring device associated with the patient and generating thoracic impedance physiological data and cardiac physiological data; a computer system, the computer system comprising: receiving the thoracic impedance physiological data and the cardiac physiological data; modeling cardiac phase using said cardiac physiological data; calculating a frequency domain or a time domain approximation of a cardiac artifact based on the representation of the cardiac phase and the thoracic impedance physiological data; removing the cardiac artifacts from the thoracic impedance physiological data to generate filtered thoracic impedance data; calculating a probability of an apnea event based on the filtered thoracic impedance data; and generating an alert indication responsive to meeting a set of criteria based on the probability of the apnea event over a time period. Patient monitoring systems.

9. The computer system further comprises:

10. The patient monitoring system of claim 8, programmed to trigger automatic physical stimulation of the patient.

10. The computer system further comprises:

10. The patient monitoring system of claim 8, programmed to calculate a length of the apnea event.

11. The computer system further comprises: recording apnea events for said patient; Transmitting historical information regarding apnea events to a user interface; Calculating aggregate information regarding the recorded apnea events for the patient; and 10. The patient monitoring system of claim 8, programmed to transmit the aggregate information regarding the recorded apnea events to the user interface.

12. 12. The patient monitoring system of claim 11, wherein the computer system is further programmed to record oxygen saturation and heart rate data associated with the apnea events.

13. The computer system is further programmed to receive heart rate measurements and oxygen saturation measurements from the bedside monitoring device; 9. The patient monitoring system of claim 8, wherein the set of criteria includes heart rate and oxygen saturation criteria.

14. 10. The patient monitoring system of claim 8, wherein the patient is a neonate.

15. 1. A non-transitory storage medium having stored thereon software for detecting apnea events in a patient, the software, when executed, causing a computer system to: receiving thoracic impedance physiological data and cardiac physiological data from a bedside monitoring device associated with the patient; modeling cardiac phase using said cardiac physiological data; calculating a frequency domain or a time domain approximation of a cardiac artifact based on the representation of the cardiac phase and the thoracic impedance physiological data; removing the cardiac artifacts from the thoracic impedance physiological data to generate filtered thoracic impedance data; calculating a probability of an apnea event based on the filtered thoracic impedance data; and generating an alert indication responsive to meeting a set of criteria based on the probability of the apnea event over a time period. Non-transitory storage media.

16. The software, when executed, further causes the computer system to:

16. The non-transitory storage medium of claim 15, further comprising: a memory for storing a plurality of memory devices for storing a plurality of memory devices;

17. The software, when executed, further causes the computer system to:

16. The non-transitory storage medium of claim 15, causing a length of the apnea event to be calculated.

18. The software, when executed, further causes the computer system to: recording apnea events for said patient; Transmitting historical information regarding apnea events to a user interface; Calculating aggregate information regarding the recorded apnea events for the patient; and 16. The non-transitory storage medium of claim 15, further configured to cause the user interface to transmit the aggregate information regarding the recorded apnea events.

19. The software, when executed, further comprises: the computer system determining an SpO 2 and heart rate data.

20. The software, when executed, further causes the computer system to: receiving a heart rate measurement and an oxygen saturation measurement from the bedside monitoring device; 16. The non-transitory storage medium of claim 15, wherein the set of criteria includes heart rate and oxygen saturation criteria.

Images