Sleep disordered breathing (SDB) patient care path including prediction of collapse pattern and / or therapy outcome

Abstract

A method of sleep disordered breathing care using specific patient information for a first patient to determine a probability of a particular collapse pattern for an upper airway and / or a therapy outcome, and based on the determination, performing further evaluative procedures and / or implementing therapy.

Description

1618.322 1111SLEEP DISORDERED BREATHING (SDB) PATIENT CARE PATH INCLUDING PREDICTION OF COLLAPSE PATTERN AND / OR THERAPYOUTCOMECross-Reference to Related Applications

[0001] This application claims benefit to U.S. Provisional Application Serial No. 63 / 699,138, filed on September 25, 2024, the entire teachings of which is incorporated herein by reference.Background

[0002] A significant portion of the population suffers from various forms of sleep- related issues, some of which may involve sleep disordered breathing (SDB) and / or other conditions. Patients seeking care for these conditions often navigate a myriad of options, which may be confusing as they seek to sort out the causes and / or best solutions for their conditions.Brief Description of the Drawings

[0003] FIG. 1 is a block diagram schematically representing an example method of SDB patient care.

[0004] FIG 2 is a diagram schematically representing an example sleep disordered breathing (SDB) patient care path (or workflow) according to example methods of the present disclosure.

[0005] FIGS. 3-4 are each a block diagram schematically representing an example patient information portions.

[0006] FIG. 5 is a diagram including graph representing a respiratory waveform including at least some respiratory cycles corresponding to a concentric collapse pattern of the upper airway.

[0007] FIG. 6 is a diagram including graph representing a respiratory waveform including at least some respiratory cycles corresponding to a non-concentric collapse pattern of the upper airway.1618.322 1112

[0008] FIG. 7 is a diagram including an enlarged portion of FIG. 5 further representing respiratory cycles corresponding to a concentric collapse pattern.

[0009] FIG. 8 is a diagram including an enlarged portion of FIG. 6 further representing respiratory cycles corresponding to a non-concentric collapse pattern.

[0010] FIG. 9A is a block diagram representing an example scoring tool based on application of a data model relative to patient information.

[0011] FIG. 9B is a flow diagram representing an example SDB patient care path in relation to determining a probability of a particular collapse pattern.

[0012] FIG. 10A is a diagram including a graph and FIG. 10B is a chart, each of which represent a probability of a collapse pattern according to an example data model.

[0013] FIG. 11 is a block diagram representing an example stimulation portion.

[0014] FIG. 12 is a block diagram representing an example method of determining a probability of a collapse pattern.

[0015] FIG. 13 is a block diagram schematically representing an example data model type portion.

[0016] FIG. 14 is a diagram schematically representing an example method of constructing a data model for identifying a probability of a collapse pattern.

[0017] FIG. 15 is a diagram schematically representing an example method of using a constructed data model for identifying for identifying a probability of a collapse pattern.

[0018] FIGS. 16A-16D are diagrams including a top view illustrating example upper airway collapse patterns.

[0019] FIG. 16E is a diagram illustrating a side sectional view of example patient anatomy including an upper airway and related tissues.

[0020] FIGS. 16F-16G are diagrams illustrating example patient anatomy designators including example upper airway collapse patterns and locations.

[0021] FIG. 17 is a diagram schematically representing an example patient care environment.

[0022] FIG. 18A is a block diagram schematically representing an example control portion.1618.322 1113

[0023] FIG. 18B is a diagram schematically illustrating at least some example arrangements of a control portion.

[0024] FIG. 18C is a block diagram schematically representing a user interface.

[0025] FIG. 19 is a block diagram which schematically represents some example implementations by which an IMD may communicate wirelessly with external devices outside the patient.Detailed Description

[0026] In the following detailed description, reference is made to the accompanying drawings which form a part hereof, and in which is shown by way of illustration specific examples in which the disclosure may be practiced. It is to be understood that other examples may be utilized and structural or logical changes may be made without departing from the scope of the present disclosure. The following detailed description, therefore, is not to be taken in a limiting sense. It is to be understood that features of the various examples described herein may be combined, in part or whole, with each other, unless specifically noted otherwise.

[0027] Some current patient care pathways for those exhibiting sleep disordered breathing may include attempts to determine a location and / or degree of collapse within the upper airway which obstructs normal breathing. Among other techniques, some patient care pathways may include participating in a drug- induced sleep endoscopy (DISE) procedure in which a patient is placed in a state of simulated sleep during which a surgeon aims to observe at what level the upper airway collapses, a degree of collapse (e.g. partial, complete), and / or a pattern of collapse. In some examples, imaging sensors are used to make the observations such as (but not limited to) a camera being inserted into the mouth and upper airway to facilitate visualizing the anatomy, functions, etc. This collapse information may be used to determine which types of treatment are expected to provide positive therapeutic results or whether available treatments may be unsuccessful. For example, based on the DISE procedure, a surgeon may determine that a given patient should have surgery to remove their tonsils and / or remove other tissues which may structurally crowd the upper airway. Via the1618.322 1114DISE procedure, a surgeon also might observe that the upper airway becomes obstructed from the tongue falling back into the oropharynx portion of the upper airway. For some patients, preventing such rearward tongue movement during sleep may promote patency of the upper airway, thereby reducing a frequency of obstructive sleep apnea (OSA) events. In some examples, applying electrical stimulation to the hypoglossal nerve (and / or genioglossus muscle) while the patient is sleeping will result in tongue protrusion, thereby maintaining patency of the oropharynx portion of the upper airway. As further described later, other nerves and / or muscles may be stimulated to promote upper airway patency.

[0028] However, such electrical stimulation therapy to cause tongue protrusion may be less effective for some patients which exhibit certain collapse patterns such as, but not limited to, a concentric collapse pattern (e.g. at the velum (soft palate) of the upper airway because such tongue protrusion does not move those tissues responsible for the partial or complete obstruction of the upper airway.

[0029] With this in mind, at least some example methods / devices of the present disclosure may identify a likelihood (e.g. probability) of a particular collapse pattern(s) (e.g. concentric collapse) based on available patient information to help determine a preferred course of action. For patients still on a path of determining their type and / or degree of sleep disordered breathing, as well as suitable treatment options, the identification of a probability of a particular collapse pattern (e.g. concentric collapse) may be used to advise a patient to forgo some types of invasive evaluative procedures such as (but not limited to) a DISE procedure, advise a patient whether implementation of certain types of SDB therapy may be effective, and / or other courses of action. For instance, if the example method identified a low likelihood (e.g. low probability) of concentric collapse pattern for a particular patient, then the patient may be advised to forgo a DISE procedure and / or advised that electrical stimulation of certain upper airway patency-related (LIAPR) tissue likely would be effective. In some examples, such electrical stimulation of LIAPR tissues may comprise electrical stimulation of a hypoglossal nerve, genioglossus muscle, infrahyoid muscle (e.g. strap muscles), and / or infrahyoid muscle (IHM)-innervating nerves, among other potential target tissues which may indirectly or directly promote upper airway patency.1618.322 1115

[0030] Advising a patient to forgo a drug-induced sleep endoscopy (DISE) procedure in some instances may enhance patient care, benefiting both the patient and care teams. In particular, a DISE procedure is expensive and can be difficult to obtain because the DISE procedure requires access to an operating room (OR), for which access is limited because of other surgeries, procedures, etc., which have higher priority than a DISE procedure. Moreover, a DISE procedure typically involves an ear, nose, throat (ENT) surgeon, anesthetist, and nurse, which also are typically in higher demand for other procedures, tasks, duties, etc. In some instances, the results of a DISE procedure for identifying a collapse pattern, particularly in relation to diagnosing sleep disordered breathing (SDB), may sometimes be viewed as being subjective and / or inconsistent depending on the experience (or philosophy) of the ENT surgeon, particular patient anatomy, insertion techniques for (or available positioning of) the imaging tools, type / size of imaging tools, etc.

[0031] In some examples, patient access to a DISE procedure may vary widely by socio-economic status, cultural expectations, geographic location, ethnicity, etc., such that reliance on a DISE procedure as the main or sole screening tool for a therapeutic device (e.g. implantable neurostimulator) may prevent or deter a significant number of patients who could benefit from such a therapeutic device from receiving appropriate SDB patient care.

[0032] Accordingly, via at least some example methods, devices, and / or systems of the present disclosure, available patient information may be used according to example data models to identify a low probability of a particular collapse pattern (e.g. concentric collapse). With this identification, the patient may bypass having a DISE procedure (with its attendant disadvantages) as part of receiving a faster determination that the particular patient is a good stimulation therapy candidate for some UAPR tissues, which in turn may expedite implementing (e.g. implanting) the stimulation therapy. On the other hand, in some instances the example methods also may use patient information, data models, etc., to identify a relatively higher probability of a particular collapse pattern (e.g. concentric collapse). With this identification, the patient also may bypass a DISE procedure (with its attendant disadvantages) as part of receiving a faster determination that1618.322 1116 the particular patient is a poor stimulation therapy candidate for some UAPR tissues, which in turn may expedite foregoing stimulation therapy.

[0033] Moreover, as will be apparent from these example arrangements that the identification of a probability of a collapse pattern and / or therapy outcome may be implemented in multiple ways, as will be further described later in various examples of the present disclosure.

[0034] In some examples, a method of sleep disordered breathing care comprises receiving, at least one of a clinician portal and a service provider resource, specific patient information for a first patient. In some examples, the method also comprises determining via the service provider resource which is in communication with the clinician portal, based on the specific patient information, a numerical score representing a probability of a complete concentric collapse (CCC) pattern for an upper airway and / or a therapy outcome. In some examples, the method also comprises comparing, via the service provider resource, the numerical score to a selectable first criterion, and upon the service provider resource determining that the numerical score fails to meet the selectable first criterion, the service provider resource communicating to the clinician portal for display at the clinician portal: a first recommendation to forego a drug-induced sleep endoscopy procedure for the first patient; and a second recommendation to implement electrical stimulation of an upper airway patency-related tissue for a non-complete concentric collapse pattern of the upper airway.

[0035] In some examples, the method comprises, upon the service provider resource determining that the numerical score meets the selectable first criterion, the service provider resource communicating to the clinician portal for display at the clinician portal a third recommendation to perform a drug-induced sleep endoscopy procedure on the first patient. This method also comprises, the service provider resource receiving, from the clinician portal, a result of the DISE procedure confirming the first patient having a CCC pattern, and the service provider resource communicating to the clinician portal for display at the clinician portal a fourth recommendation to: forego implementing electrical stimulation of the upper airway patency-related tissue including the hypoglossal nerve as the sole target tissue; or implement electrical stimulation of the upper airway patency-1618.322 1117 related tissue, which includes both a hypoglossal nerve and a non-hypoglossal nerve.

[0036] In some examples, these examples methods the non-hypoglossal nerve comprises an infrahyoid muscle-related nerve and / or an infrahyoid muscle. The infrahyoid muscle may comprise a sternothyroid muscle and the infrahyoid muscle-related nerve innervates the sternothyroid muscle.

[0037] In some examples, in these example methods, determining the numerical score comprises submitting the patient specific information as an input to a constructed data model to determine, as an output, the numerical score.

[0038] In some examples, the specific patient information comprises at least one of gender and BMI, while in some examples, the specific patient information comprises a sensed respiratory waveform of the patient including inspiratory phase information including at least one sensed inspiratory phase in which a negative effort dependence parameter meets a selectable first criterion, a negative skewness meets a selectable second criterion, and / or a substantially reduced amplitude meets a selectable third criterion.

[0039] In some examples, the specific patient information comprises behavioral information comprising at least one of: an arousal threshold parameter; a snoring profile parameter; a consumption habits parameter; a loop gain parameter; a neuromuscular collapse response parameter; or a cortical sensorimotor-muscle coherence parameter.

[0040] In some examples, the specific patient information comprises sleep study information, while in some examples, the specific patient information comprises electronic health record information, patient survey information, demographic information, and / or behavioral information.

[0041] In some examples, a device implements the method of sleep disordered breathing care, wherein the device comprises: a memory configured to store machine readable instructions; and at least one processor coupled with the memory and configured to execute the machine readable instructions to: receive, at at least one of the clinician portal and the service provider resource, specific patient information for the first patient; determine via the service provider resource which is in communication with the clinician portal, based on the specific1618.322 1118 patient information, a numerical score representing the probability of a complete concentric collapse (CCC) pattern for an upper airway and / or the therapy outcome; compare, via the service provider resource, the numerical score to the selectable first criterion; and upon the service provider resource determining that the numerical score fails to meet the selectable first criterion, the service provider resource to communicate to the clinician portal for display at the clinician portal: the first recommendation to forego a drug-induced sleep endoscopy procedure for the first patient; and the second recommendation to implement electrical stimulation of an upper airway patency-related tissue for a non-com plete concentric collapse pattern of the upper airway.

[0042] These examples, and additional examples, are further described below in association with at least FIGS. 1-19.

[0043] As shown at 200 in FIG. 1 , one example method comprises identifying a probability of a collapse pattern and / or therapy outcome. Various aspects of this method and / or related methods are further described below.

[0044] FIG. 2 is a diagram schematically representing an example method 205 implementing a sleep disordered breathing (SDB) patient care path (e.g. workflow) and may comprise one example implementation of method 200 of FIG. 1. As shown at 210 in FIG. 2, an example method 205 comprises identifying a probability of a collapse pattern of an upper airway based on patient information. In some examples, the patient information may comprise specific patient information 500 (FIG. 3) relating to a particular patient and / or general patient information 800 (FIG. 4) relating to patient populations.

[0045] In some examples, further details regarding identifying the probability of a collapse pattern according to a score and / or other indicator(s) are described extensively in association with at least FIGS. 9A-10B. Meanwhile, a range of different types and / or locations of collapse patterns is further described later in association with at least FIGS. 16A-16G, while FIGS. 5-8 provide examples of using respiratory waveform features to help identify different collapse patterns of the upper airway.

[0046] Once the probability of a particular collapse pattern (e.g. complete concentric collapse (CCC)) has been identified, then method 205 in FIG. 21618.322 1119 comprises using this information to guide the patient further along an example SDB patient care path.

[0047] In some example implementations of method 205, upon identifying a relatively high probability of a particular collapse pattern (e.g. a complete concentric collapse (CCC), the patient may sometimes be referred to as “a CCC patient”. Because of this relatively high probability, via method 205, the CCC patient may avoid path 222, effectively foregoing a further evaluative procedure 292, such as a DISE procedure 294, because the identification was made reliably using objective patient information (e.g. 500 in FIG. 3, 800 in FIG. 4), as further described later. In some examples, this CCC patient may take path 230 to 232 at which they forego some types of electrical stimulation therapy (e.g. certain types of tissue stimulation (e.g. certain types of stimulating solely certain portions of the hypoglossal nerve, genioglossus muscle, and / or other nerves / muscles)) because such stimulation therapy may be relatively ineffective in treating a complete concentric collapse pattern. It will be understood that such types of tissue stimulation may comprise implantable stimulation elements and / or external stimulation elements.

[0048] However, there are some circumstances in which a CCC patient may proceed along path 220 to implement a stimulation therapy at 240, as further described below.

[0049] In some example implementations of method 205, either upon identifying a relatively low probability of a particular collapse pattern (e.g. concentric collapse pattern) at 210 or upon identifying a relatively high probability of a particular collapse pattern (e.g. tongue-based collapse pattern) at 210, the patient may sometimes be referred to as “a non-CCC patient”. Because of this relatively high probability of being a non-CCC patient (or expressed as being a relatively low probability of being a CCC patient), via method 205, the CCC patient also may avoid path 222, effectively bypassing a further evaluative procedure 292 (e.g. DISE procedure 294) on their way via path 220 to receive new stimulation therapy 260 via path 244. Per path 270, this new stimulation therapy 260 may comprise implementing (e.g. implanting) new device components 272, which may be implantable 274 and / or external 276. In some examples in which the non-CCC1618.322 11110 patient’s collapse pattern is tongue-based, then the new stimulation therapy may target the hypoglossal nerve (and / or genioglossus muscle), in some examples. In some such examples, the prediction of a lack of complete concentric collapse (CCC) may also serve as a prediction (e.g. identification of a relatively high probability) that a stimulation therapy targeting the hypoglossal nerve (HGN) will provide a successful outcome of reducing disease burden via promoting upper airway patency.

[0050] In some examples, a patient on the SDB patient care path 205 already has implemented (e.g. received) stimulation therapy such as via a therapeutic device (e.g. implantable) and also may have already participated in a DISE procedure 294. However, this patient experienced poor therapy outcome, and they are seeking a better outcome. Accordingly, the patient enters (or re-enters) the SDB patient care path, at 205 in FIG. 2, at which an identification is made of a probability of a collapse pattern based on patient information (e.g. 500 in FIG. 3, 800 in FIG. 4 as further described later). In some examples in which this patient is identified (at 210) as being a CCC patient, the patient may take path 220, 242, to 250 in which settings (e.g. stimulation 252, sensing 254) may be adjusted for an already implemented (e.g. implanted) stimulation therapy in an attempt to achieve a more successful therapeutic outcome despite the apparent confirmation of being a CCC patient as the explanation for the previously poor therapeutic outcome.

[0051] Moreover, in some examples in which this patient (having an already implemented stimulation therapy for which there was a poor outcome) has been indicated or confirmed to be a CCC patient, then some example implementations may comprise taking path 220, 244 to 260 at which new stimulation therapy is implemented via path 262 in which a new target tissue is engaged (264). For instance, the engagement of new target tissue at 264 may comprise electrically stimulating nerves and / or muscles which promote upper airway patency via increasing a tone of, and / or causing full contraction of, muscles forming at least a portion of the upper airway (e.g. soft palate / velum) in order to lessen or prevent the complete concentric collapse. In some such examples, this remedy may be achieved via stimulating solely the new target tissue. In some examples, the new1618.322 11111 target tissue may comprise other nerves / muscles which also promote upper airway patency, as further described below in association with at least FIGS. 11 , 16A-16G, etc.

[0052] However, in some examples, the new target tissue (264) may be stimulated in complementary relation (e.g. in addition) to the prior (originally targeted) target stimulation tissue. In some such examples, the SDB patient care path also comprises taking path 265 to 250 at which the settings (e.g. stimulation 252 and / or sensing 254) of the already implemented stimulation therapy are adjusted in order to better promote upper airway patency because the combination of therapy of multiple nerve targets (and / or muscles targets) may produce a successful therapeutic outcome despite the indication of being a CCC patient.

[0053] For patients which have an already implemented therapy and the method 205 directs that a new target tissue be stimulated (e.g. at 264), as described above, in some examples a new separate SDB therapeutic device (e.g. IMD and / or EMD) (to engage the new target tissue (264) may be implemented in addition to or instead of the already implemented SDB therapeutic device. The new separate SDB therapeutic device may activate new target tissue 264 using device components 272 separate from the already implemented SDB therapeutic device and / or using at least some device components of the already implemented SDB therapeutic device. In some such examples, upon implementing new device components and engaging new tissue at 264, method 205 may also comprise, at 250, adjusting the settings (e.g. stimulation and / or sensing) of the already implemented SDB therapeutic device, which may in some examples enhance complementary operation of the new SDB therapeutic device and the already implemented SDB therapeutic device. However, in some examples, the new SDB therapeutic device may operate independently of the already implemented SDB therapeutic device, which in some examples may comprise the new SDB therapeutic device not being in communication with the already implemented SDB therapeutic device.

[0054] Among other aspects of method 205, in some examples at 290 additional patient information may be obtained via performing an evaluative procedure on1618.322 11112 the patient (292). In some examples, the evaluative procedure (292) may be invasive such as placing a sensor within a body cavity, such as but not limited to placement of a sensor within a mouth and / or upper airway of the patient. In some such examples, the sensor (and / or tools associated with the sensor) do not puncture or pass through the skin and / or other surface tissues. However, in some examples, the invasive nature of the evaluative procedure may comprise puncturing or penetrating a surface tissue, skin, etc. In some examples, the evaluative procedure may comprise a drug-induced sleep endoscopy (DISE) procedure 294, as previously described. The DISE procedure 294 may be used to help determine which, if any, particular collapse pattern (e.g. concentric collapse) is occurring for the particular patient, the location along the upper airway at which the collapse occurs, the degree of collapse (e.g. obstruction), and / or which tissues are causing the collapse, etc.

[0055] Other evaluative procedure 296 and / or other ways 298 of obtaining patient information may yield some of the patient information (e.g. 500 in FIG. 3; 800 in FIG. 4).

[0056] With this in mind, in some examples in which the identification at 210 of method 205 yields an indication that a patient is not clearly a CCC patient and this patient also is not clearly a non-CCC patient, then path 222 may be taken to obtain the previously described additional patient information via a further evaluative procedure 292 such as (but not limited to) a DISE procedure 294. In some instances, this further evaluation may help clarify a level, type, and / or degree of a collapse pattern, which may then be used to direct the patient along path 220 (including further paths 242, 244) or path 230. With this in mind, as will be further described later in association with at least FIG. 9B, in some examples the identification of a probability of a particular collapse pattern (and / or therapy outcome) may be expressed as a numerical score (at 1781), which then may be compared to criteria (at 1782) to further determine whether a further evaluative procedure (1786 in FIG. 9B, e.g. DISE 294 in FIG. 2) is recommended. In some such examples, a first threshold and / or second threshold (1794, 1795) may be used to help determine which values of a score will fall clearly into a CCC patient category (and for which DISE may be skipped), fall clearly into a non-CCC patient1618.322 11113 category (and for which DISE may be skipped), or fall into a category in which it may be less clear whether the patient is a CCC patient or a non-CCC patient (and for which DISE may be of greater benefit to help determine whether CCC is present). In some examples, the score 1783 (at 1782 in FIG. 9B) is based on multiple parameters while in some examples, the score is based on a single parameter.

[0057] While several previously described examples of method 205 direct CCC patients (identified at 210) to forego or bypass a further evaluative procedure 292 (e.g. DISE 294), in some examples the method 205 may still direct this CCC patient along path 222, 290 to the further evaluative procedure 292 (e.g. DISE 294) as part of an effort toward achieving a successful therapeutic outcome via learning more about the patient’s upper airway anatomy, physiologic function, etc. For instance, it may be desirable to seek visual confirmation that a non-responder to an implanted electrical stimulation therapy (e.g. targeting the hypoglossal nerve) does in fact exhibit a complete concentric collapse pattern, which provides an at least partial explanation for their non-response.

[0058] While the above-described examples for the method 205 (FIG. 2) focused primarily on example implementations involving identifying concentric collapse patterns, it will be understood that the same principles of operation may be applied to direct a patient along a SDB patient care path while considering collapse patterns other than a complete concentric collapse (CCC) and / or other than tongue-based collapse patterns. For instance, method 205 may be performed in a manner in which a probability of an epiglottis-based (e.g. epiglottis- dominated) collapse pattern and / or a pharyngeal wall-based (e.g. lateral and / or posterior) collapse pattern is used (with or without regard to tongue-based collapse information) to direct a patient along the paths (220, 222, and / or 230) of method 205 in order to successfully predict therapeutic outcomes, patient eligibility, etc.

[0059] Finally, it will be understood that in some example the method 205 (e.g. SDB patient care path / workflow) in FIG. 2 may be implemented via (e.g. within) a device environment such as but not limited to example arrangement(s) further described later in association with at least FIGS. 17, 18B-19.1618.322 11114

[0060] FIG. 3 is a block diagram schematically representing an example specific patient information portion 500, which receives, obtains, tracks, etc. information (e.g. physiologic, other) which is specific to a particular patient. As shown in FIG. 3, in some examples specific patient information portion 500 may comprise demographic portion 520, behavioral portion 535, electronic health record (EHR) portion 540, survey portion 550, and / or sensing portion 600.

[0061] The specific patient information portion 500 may receive, obtain, and / or track the specific patient information that includes demographic information, physiological information, and / or behavioral information, etc. from a variety of sources, including but not limited to: sensor(s); memory storing sensed information received from the sensor(s); memory storing patient medical records (e.g., EHR as shown by 540); memory storing inputs from surveys (as shown by 550); and / or memory storing behavioral information 535. In some examples, the sensor(s) may be external to a patient and / or may be implantable. In some examples, one or more of the external sensors may comprise a non-contact sensor 687, wearable sensor 689, and / or other forms of external sensing. In some examples, the non-contact sensor 687 may comprise a nearable sensor 688. In some examples, the sensing parameters represented in specific patient information portion 500 may comprise actual sensors of the respective sensing modalities, formats, sources, etc. identified in FIG. 3 and / or may comprise sensed information received from such actual sensors. As further described below, in some examples, the sensor(s) may comprise a sleep study such as (but not limited to) a formal sleep study via polysomnography (PSG) 612, 620 or informal / home sleep study (e.g. 614), or other format 616 of sleep study.

[0062] As further described below, sources of the patient information may thereby include sleep studies including, but not limited to, information associated with sleep disturbances, percent of time in different sleeping positions (per 641 in FIG. 3), endogenous signals such as electromyography (EMG, 626 in FIG. 3), electroencephalogram (EEG, 622 in FIG. 3), respiration effort (634 in FIG. 3), heart rate (630 in FIG. 3), and body temperature (640 in FIG. 3). Potential derived signals from such information may include, but is not limited to, delta heart rate (630), heart rate variability (630), negative-effort dependence (678), coherence1618.322 11115 metrics (, respiratory effort (634), respiratory arousal thresholds (536), loop gain metrics (539A), and latent wakefulness through a combination of the signals obtained from the sleep study.

[0063] In some examples, the patient information may include markers which indicate a higher probability of a particular collapse pattern (and / or a particular therapy outcome). For instance, one example marker indicative of a higher probability of a non-complete concentric collapse (CCC) pattern includes a coherence parameter 539C. In some examples, the coherence parameter 539C may be associated with behavioral portion 535 of specific patient information portion 500 in FIG. 3, while in some examples, the coherence parameter 539C may be associated with other specific portions of (or generally associated with) the specific patient information portion 500.

[0064] In some examples, the coherence parameter comprises identification of a coherence between an electroencephalography (EEG) parameter (e.g. 622 in FIG.3) and an electromyographic (EMG) parameter (e.g. 626 in FIG. 3), and may be referred to as an EEG-EMG coherence parameter in some examples. In some examples, the EEG information may be EEG signals / information associated with the neck region (e.g. C3, C4) while in some examples, the EMG information may be associated with muscles in the head-neck region (e.g. chin). In some examples, this EEG information and / or EMG information may be obtained during a sleep study and / or other techniques.

[0065] With regard to the coherence parameter 539C, in some examples, the EEG information comprises sleep stage information such as (but not limited to) the REM sleep stage and each NREM sleep stage (e.g. N1 , N2, N3). In some examples, in context with respiratory-specific events (e.g. behavior) a high degree of coherence between active cortical sensorimotor areas (e.g. sensed via EEG at C3 / C4 of spine) and activation of lower facial motor units (e.g. chin muscles sensed via EMG) corresponds to lower obstructive sleep apnea (OSA) severity, while a low degree of such coherence corresponds to a higher OSA severity. In particular, in some examples, a significantly higher predictive value in identifying patient’s responsive to neurostimulation therapy (e.g. electrically stimulating LIAPR tissue (e.g. hypoglossal nerve)) associated with such coherence1618.322 11116 corresponds to determining the coherence at the N3 NREM sleep stage. Accordingly, in some examples, when high coherence is identified, it may be used to identify a non-CCC patient which may be treated via electrical neurostimulation of UAPR tissue (e.g. hypoglossal nerve). However, it will be understood that in some examples, reasonably predictive values for the coherence parameter 539C may be obtained when making such determinations at other sleep stages (e.g. N1 , N2, REM).

[0066] In some examples, one marker example (e.g. a marker indicative of a higher probability of a particular collapse pattern) may comprise airflow shape, which may be obtained and / or tracked in association with airflow shape parameter 632 of sensing portion 600 in FIG. 3, in some examples. In some examples, the airflow shape information may be obtained from sleep studies, which may be used to predict patient outcomes using machine learning model(s) (e.g. deep learning) that identify latent features on the airflow shape waveform to discriminate responders versus non-responders. In some examples, first-order features of airflow shape of significance include negative-effort dependence, skewness, peak-width, peak-height, peak-to-trough-ratio, and second-order linear approximation shapes. At least some of these airflow shape features are discussed in association with at least FIGS. 5-8. In some examples, at least some airflow shape features may be obtained or include at least some of substantially the same features and / or attributes as described in: Aittokallio, et al., “Inspiratory flow shape clustering: An automated method to monitor upper airway performance during sleep”, Computer Methods and Programs in Biomedicine, 85:1 , 2007, pages 8-18, ISSN 0169-2607; and Op de Beeck, et al., “Polysomnographic airflow shapes and site of collapse during drug-induced sleep endoscope”, Eur Respir J 2024; 63: 2400261 , each of which are incorporated herein in their entirety for their teaching.

[0067] In some examples, one marker example (e.g. a marker indicative of a higher probability of a particular collapse pattern) may include respiratory effort that is associated with an obstruction event. For instance, in some examples a patient may predominantly have obstructive sleep events leading to increased respiratory effort. In some instances, some such patients may have a higher1618.322 11117 probability of a non-CCC collapse pattern, and therefore such patients likely would respond well to electrical stimulation of at least the hypoglossal nerve (e.g. stimulation therapy 240 in FIG. 2). In some such examples, this sensed respiratory effort may be derived from a combination of sleep study signals including sleep belts (which sense respiration), airflow sensors, and / or other respiratory information sensing. In some such examples, this respiratory effort information may be obtained via, and / or is associated with, respiratory effort parameter 634 of sensing portion 600 in FIG. 3.

[0068] Any of the above information may be obtained via a sleep study, such as a formal sleep study (e.g. 612 in FIG. 3) and / or informal at-home sleep study (e.g. 614 in FIG. 3). In some examples, the information may be obtained from patient medical records (e.g. EHR 540 in FIG. 3), such as from a previously conducted sleep study. Moreover, the patient information may be obtained from sources other than a sleep study, including but not limited to patient medical records (e.g. EHR 540 in FIG. 3) and surveys (e.g. 550 in FIG. 3). Such information may include demographic information (e.g. 520 in FIG. 3) and / or behavioral information (e.g. 535 in FIG. 3), among other information.

[0069] In some examples, in general terms demographic information 520 may comprise any type of demographic information regarding a patient, which in some examples may comprise gender 522, age 524, comorbidities 525, body mass index (BMI) 526, neck size (e.g. circumference) 527, supine pharyngeal width (SPW) 528, pharyngeal opening pressure (POP) 529, and / or other attributes 532. In some examples, the pharyngeal opening pressure (POP), which may correspond to a 90 / 95% Auto Positive Airway (PAP) pressure.

[0070] In general terms, a lower value of supine pharyngeal width (SPW), such as lower than a selectable threshold (e.g. 17mm, 18mm, 19mm, 20 mm), may correspond to a higher probability of a collapse pattern (e.g. complete concentric collapse (CCC)), and therefore in some examples this attribute (alone or with other attributes) may be useful in patient candidacy screening, predicting therapy outcomes, and / or in determining whether further evaluative procedures (e.g. DISE) should be performed. In other words, a SPW lower than the selectable threshold, without or without other attributes, may indicate that the further1618.322 11118 evaluative procedure (e.g. DISE) can be omitted because the SPW measurement indicates a high probability of a collapse pattern (e.g. CCC).

[0071] This type of analysis, screening, and prediction may be implemented for other easily obtained patient information, such as neck circumference alone and / or in combination with the supine pharyngeal width (SPW) attribute (and / or other attributes).

[0072] Of course, some attributes such as age, gender, and BMI also are easily obtained and may form the basis for such analysis, screening, and prediction with (or without) at least some of the other attributes of patient information (e.g. FIG. 3) which are easily obtained without a formal sleep study or further evaluative procedure (e.g. 1786 in FIG 9C such as DISE).

[0073] In some examples, this type of analysis, screening, and prediction may be performed in a regular clinic setting saving clinicians and patient a considerable amount of time and money while increasing the accuracy of predictions regarding successful therapeutic outcomes, including which therapies may be most effective, if at all.

[0074] In some examples, this type of analysis, screening, and prediction may be implemented in conjunction with less easily obtained information, such as at least some patient information from a sleep study (e.g. PSG 612, home 614, other 616 in FIG. 3).

[0075] In some examples, at least some aspects of such analysis, screening, and prediction for decision-making along a SDB patient care path are at least partially embodied or implemented in example methods 205 (FIG. 2), 1780 (FIG. 9B), and / or other example methods within the present disclosure. In some such examples, at least some aspects of such analysis, screening, and prediction for decision-making (e.g. within a clinic setting) may comprise at least a portion of identifying a probability of a particular collapse pattern as implemented at 210 in FIG. 2, at 1781 in FIG. 9B, and / or other action points within example methods of the present disclosure.

[0076] In some examples, at least some of the patient-specific demographic information may sometimes comprise (or be referred to as) morphometric information (identified via dotted box 534), which comprises quantitative data1618.322 11119 expressing information about an absolute (or relative) size, shape, volume, proportional distribution, etc. of a patient’s entire body or portion thereof. In some examples, the morphometric information 532 may comprise BMI 526, neck size 527, supine pharyngeal width (SPW) 528, pharyngeal opening pressure (POP) 529, and / or other physical attributes 532.

[0077] In some examples, the demographic information within portion 520 may comprise at least some specific patient information from at least the sensing portion 600, the EHR portion 540, and / or survey portion 550. For instance, at least some respective parameters of the demographic information portion 520 may be sensed during a sleep study, form part of a patient medical record, and / or be input by the patient or another person as part of a survey.

[0078] In some examples, in general terms behavioral information portion 535 may comprise any type of behavioral information parameter regarding a patient, which in some examples may comprise the following parameters: arousal threshold 536; snoring profile 537; consumption habits 538; loop gain 539A; neuromuscular response 539B (e.g. neuromuscular response to upper airway collapse); and / or coherence parameter 539C. In some examples, arousal threshold parameter 536, the loop gain parameter 539A, neuromuscular response 539B, and / or coherence parameter 539C may be considered as non- anatomical factors (affecting upper airway collapsibility) in comparison to anatomical factors such as a generally narrow upper airway arising from the general anatomical structural features of the patient. In some examples, each non-anatomical factor may sometimes be referred to as a patient endotype.

[0079] Example consumption habits 538 include drinking alcohol, smoking, illicit drugs, certain prescription drugs, and / or dietary information, among others. Viewed individually or in various combinations, this information may indicate susceptibility of upper airway collapse. For instance, a patient which regularly consumes alcohol, which is known to relax upper airway musculature, may be more prone to obstructive sleep apnea. Moreover, at least some consumption habits (538) may contribute to BMI (526, 547 in FIG. 3).

[0080] In some examples, the arousal threshold parameter 536 corresponds to a relative propensity to awaken. For instance, a patient that awakens easily in1618.322 11120 relation to sleep disordered breathing behavior (e.g. apnea events and / or precursors to such events) may be deemed as having a low respiratory arousal threshold. In some such examples, the arousal threshold parameter 536 may be evaluated relative to selectable threshold such as (but not limited to) a rate and / or severity of sleep disordered breathing events.

[0081] In some examples, in one aspect the loop gain parameter 539A corresponds to sensitivity of the patient’s ventilatory control system, wherein a high loop gain corresponds to a patient exhibiting an overly-sensitive ventilatory control system. In some such examples, such high loop gain may lead to a higher likelihood of periodic breathing including cycles of apnea events. In some such examples, the loop gain parameter 539A may be evaluated relative to a selectable threshold such as (but not limited to) a nominal loop gain (e.g. 1 ).

[0082] In some examples, the neuromuscular response parameter 539B corresponds to how completely and / or how quickly the patient’s central nervous system and / or peripheral nerves / muscles implement a general breathing reflex response to overcome a collapsed upper airway). In some such examples, a patient with a reduced (e.g. poor) neuromuscular response (to upper airway collapse) may take longer to invoke and / or implement a patient’s general reflex response to overcome upper airway collapse (e.g. obstruction).

[0083] In some examples, at least the arousal threshold parameter 536, the loop gain parameter 539A and / or neuromuscular response 539B comprise parameters which may correspond to patient characteristics which are responsive to drugs (parameter 548 in FIG. 3) which help alleviate sleep disordered breathing. The response to drugs (parameter 548) may be recorded in a patient’s EHR (e.g. portion 540) and / or be determined via performing a trial on the patient.

[0084] As with the demographic information portion 520, the behavioral information portion 535 may comprise at least some specific patient information from at least the sensing portion 600, the EHR portion 540, and / or survey portion 550 of the specific patient information portion 500 of FIG. 3. For instance, at least some respective portions of the behavioral information 535 may be sensed during a sleep study (e.g. 612, 620 in FIG. 3), form part of a patient medical record (e.g.1618.322 11121EHR 540), may be input by the patient or another person as part of a survey (e.g. 550 in FIG. 3), and / or arise from other sources.

[0085] In some examples, the electronic health record portion 540 may comprise specific patient information available via a patient’s electronic health record (e.g. patient medical record), which may comprise at least some specific patient information from the demographic portion 520, sensing portion 600, and / or survey portion 550. For example, the electronic health record portion 540 may include information such as reported snoring parameter 543, tonsil grade parameter 545, and / or BMI parameter 547. Other potential information includes disease burden (e.g., AHI) response to drugs parameter 548, (such as a patient response to Acetazolamide and / or dronabinol), and CPAP pressure level parameter 549. For example, the occurrence of snoring 543 may be reported by a bed partner and form part of the electronic health record. In some examples, a combination of reported snoring 543, lower tonsil grade 545, and lower BMI 547 may indicate a higher probability of a non-CCC collapse pattern. In some examples, a patients disease burden response to drug(s) (548) may be used to identify the collapse pattern, such as (but not limited to) using at least some of substantially the same features and / or attributes as described in Thomson, et al., “The impact of acetazolamide and dronabinol on the physiological endotypes responsible for obstructive sleep apnea”, Sleep Med. 2025 Aug; 132: 106542, which is incorporated herein in its entirety for its teaching.

[0086] In some examples, a patient’s CPAP pressure levels (549) may be used to identify collapse information such as (but not limited to) the collapse pattern.

[0087] As previously described, in some examples, a patient with a higher probability of a non-CCC collapse pattern may indicate that electrical stimulation of certain UAPR tissue (e.g. hypoglossal nerve) would likely be effective in treating the patient, and the stimulation therapy (e.g. 240 in FIG. 2) may be implemented, in some examples, without obtaining a further evaluation procedure (290, 292 in FIG. 2), such as a DISE procedure (294 in FIG. 2).

[0088] In some examples, the reported snoring parameter 543 corresponds to reporting by a bed-partner of a relative degree of snoring by the patient. In some examples, a patient which exhibits a lesser degree of snoring while still exhibiting1618.322 11122SDB behavior (e.g. OSA events) may respond well to electrical stimulation therapy of UAPR tissue (e.g. 240 in FIG. 2), where the LIAPR tissue may comprise the hypoglossal nerve. In some such examples, the lesser degree of snoring may correspond to a non-CCC collapse pattern.

[0089] In some examples, the tonsil grade parameter 545 corresponds to a grading scope of the patient’s tonsils, which may be based on factors such as (but not limited to) a type, size, topographic shapes, degree of protrusion into airway, and / or other features of a patient’s tonsils. In some examples, a patient which exhibits a lower grade of tonsil (e.g. assigned a relatively normal score for a tonsil) while still exhibiting SDB behavior (e.g. OSA events) may respond well to electrical stimulation therapy of UAPR tissue (e.g. 240 in FIG. 2), where the UAPR tissue may comprise the hypoglossal nerve. In some such examples, the lower tonsil grade may correspond to a non-CCC collapse pattern.

[0090] In some me examples, the body-mass index (BMI) parameter 545 corresponds to BMI parameter 526 in demographic information portion 520 in FIG. 3. In some examples, a patient which exhibits a lower BMI (e.g. lower than a BMI of a typical patient receiving electrical stimulation therapy to treat OSA) while still exhibiting SDB behavior (e.g. OSA events) may respond well to electrical stimulation therapy of UAPR tissue (e.g. 240 in FIG. 2), where the UAPR tissue may comprise the hypoglossal nerve. In some such examples, the lower tonsil grade may correspond to a non-CCC collapse pattern.

[0091] In some examples, survey portion 550 may comprise an Epworth Sleep Scale (ESS) portion 552 which provides specific patient information regarding sleep, disease burden, etc. according to a patient parameter 554, and / or clinician parameter 556.

[0092] In some examples, as shown in FIG. 3, sensing portion 600 may comprise one or more of format portion 610, polysomnography (PSG) portion 620, respiratory waveform portion 650, as well as other parameters and / or portions such as drug induced sleep endoscopy (DISE) parameter 688, imaging portion 680, impedance parameter 690, pressure parameter 692, vibration parameter 694 and / or other sensing modality parameter.1618.322 11123

[0093] In some examples, in general terms, format portion 610 may comprise a format (e.g. source, environment, context, etc.) by which specific patient information is obtained, received, etc. For instance, the format portion 610 may comprise a polysomnography (PSG) parameter 612 by which patient information is obtained from a polysomnography (PSG) or other formal sleep study, while in some examples, the format portion 610 may comprise a home parameter 614 (e.g. home sleep test (HST)), which may provide a smaller range or fewer types of sensed information than a formal PSG. For example, via home parameter 614, a home sleep test may comprise respiratory airflow information (e.g. via parameter 632, 654) and / or respiratory effort information (via parameter 634, 656), which may indicate absence of breathing, pauses in breathing, relative depth (or shallowness) of breathing. In some examples, the airflow information (e.g. parameter 632, 654) is obtained via measuring airflow at the nostrils while respiratory effort information (e.g. 634) is obtained via sensors (e.g. impedance, piezoelectric, accelerometer) in sensing relation to the chest and / or abdomen during sleep with sensors removably secured to the body via belts or adhesive patches, sensors embedded into a sleep mat on which the patient lies, sensors in a nearby camera or other non-contact sensor (e.g. radiofrequency, etc.) which may be present in a phone or other personal / mobile electronic device. In some examples, a home sleep study may obtain cardiac information (630), such as heart rate, and other information, such as body temperature 640.

[0094] In some examples, the PSG portion 620 may comprise information sensed via parameters: EEG 622, EOG 624 (eye movement), EMG 626, cardiac 630 (e.g. ECG, heart rate (HR), heart rate variability (HRV), other); respiratory 631 ; oxygen saturation 636, activity and / or motion 638, body temperature 640, and / or other parameters 641.

[0095] In some examples, disease burden parameter 642 may be determined from one or more of parameters 622-641 and / or other parameters. For instance, from one or more of these parameters, one can determine arousal information, breathing interruptions (e.g. pauses, obstructions, cessation, etc.), and / other behaviors indicative of sleep disordered breathing, which may in some examples comprise one or more indices (e.g. apnea-hypopnea index (AHI), respiratory1618.322 11124 disturbance index (RDI ) , etc.) of disease burden. In some examples, the disease burden parameter 642 may comprise an AHI prior to any intervention (e.g. surgical, positive airway pressure, electrical nerve stimulation (e.g. hypoglossal nerve), etc.) while in some examples, the disease burden parameter 642 may comprise an AHI after such intervention. For purposes of identifying a probability of a collapse pattern (e.g. complete concentric collapse) (e.g. 1781 in FIG. 9B; 210 in FIG. 2) in some examples, such as based on patient information prior to or without performing further evaluative procedures (e.g. DISE), the disease burden parameter 642 may comprise an AHI without or prior to: electrical nerve stimulation therapy; positive airway pressure therapy; oral appliance therapy; and / or other interventions.

[0096] In some examples, the respiratory parameter 631 comprises parameters (e.g. sensed information, sensing modalities, etc.) regarding airflow 632, effort 634, and / or other aspects of respiration from which information about respiratory waveforms, respiratory phase (e.g. inspiration, expiration, expiratory pause), respiratory fiducials, etc. may be obtained.

[0097] In some examples, the other parameter 641 may comprise sleep position such as (but not limited to) a percentage of time spent in each respective sleep position (e.g. supine, right / left lateral decubitus, etc.), which also may be further correlated with other parameters in patient specific information portion 500 such as (but not limited to) disease burden parameter 642.

[0098] As further shown in FIG. 3, sensing portion 600 may comprise a respiratory waveform portion 650, which in general terms provides more detailed information about the inspiratory phase information, expiratory phase information (including expiratory pause), and specific features (e.g. amplitude, duration, slope, variation, rate, etc.). For instance, respiratory waveform portion 650 may comprise parameters regarding source 652, phase 658, skewness 660, peak 670, and / or negative effort dependence (NED) 678, among other parameters (e.g. slope). In some examples, the NED parameter 678 may sometimes be referred to as a scoopiness parameter, which may correspond to a depth and shape (e.g. degree of curvature) of a concave-shaped top portion of an inspiration phase of a respiratory cycle of a respiratory (e.g. airflow) waveform. In some examples,1618.322 11125 the NED parameter 678 (e.g. scoopiness) may comprise a lowest value of a middle third of an inspiratory phase (inspiration) in proportion to a highest value of a last third of the same inspiratory phase. As further described later in association with at least FIGS. 5 and 7, this relationship corresponds to a concave-shaped portion (e.g. at B2, B3 in FIGS. 5, 7), where the lowest value (at 1135A in FIG. 7) of the middle third of the inspiratory phase (1111 B in FIG. 7) of the respiratory waveform has a lower amplitude than the highest value (at second peak 1136) in the last third of the same inspiratory phase (1111 B in FIG. 7). The greater the difference between the lowest value (at 1135A) of the middle third of the inspiratory phase (111 B in FIG. 7) of the respiratory waveform and the highest value (e.g. second peak 1136) of the last third of the same inspiratory phase (e.g. 111 B in FIG. 7) corresponds to a higher degree of scoopiness (negative effort dependence 678). In some such examples, higher degrees of scoopiness may be indicative of a complete concentric collapse (CCC) pattern. In some examples, the degree of scoopiness may be evaluated via comparing a quantitative scoopiness score relative to a selectable quantitative threshold. In some examples, the degree of scoopiness may be determined automatically via computerized geometric (e.g. shape) analysis of the morphology of the sensed respiratory waveform.

[0099] In some examples, the skewness parameter 660 identifies and / or tracks an airflow shape of a top portion (e.g. 1131 A in FIG. 7) of an inspiratory phase (e.g. 1111 B in FIG. 7) in which an amplitude of an initial peak (e.g. 1132 in FIG. 7) is compared to other amplitudes of later segments (e.g. valley portion 1134A in FIG. 7) of the top portion of the remainder of the same inspiratory phase (e.g. 1111 B in FIG. 7), such as the first two-thirds of the same inspiratory phase (e.g. 1111 B in FIG. 7). In some such examples, the greater the difference between the amplitude of the initial peak (e.g. 1132 in FIG. 7) and the amplitude of the later segments (e.g. valley portion 1134A in FIG. 7) of the top portion (e.g. 1131 A) of the same inspiratory phase (e.g. 11 11 B) corresponds to a greater degree of skewness (per degree parameter 665 in FIG. 3). In some examples in which the amplitude of the initial peak (e.g. 1132 in FIG. 7) is greater than the general amplitude of the valley portion 1 134A and / or the lowest amplitude (e.g. 1135A in1618.322 11126FIG. 7) of the top portion (e.g. 1131 A in FIG. 7) of the inspiratory phase (e.g. 1111 B in FIG. 7), at least the top portion (1131 A) of the inspiratory phase 1111 B may be deemed as having a negative skew 664 (e.g. being skewed with a negative slope. In this way, a greater degree of skewness (e.g. negative skew) corresponds to a relatively higher degree 665 (e.g. proportion) of inspiratory flow occurring earlier in the same inspiratory phase of a given breath than would otherwise occur during normal, stable breathing.

[0100] In a similar manner, the skewness parameter 660 would be similarly applicable to other example respiratory cycles, such as respiratory cycle 1110C in FIG. 7 to identify and / or track an airflow shape of a top portion (e.g. 1131 B) of an inspiratory phase (1111 C) in which an amplitude of an initial peak (e.g. 1132 in FIG. 7) is compared to other amplitudes of later segments (e.g. valley portion 1134B in FIG. 7) of the top portion of the remainder of the same inspiratory phase 1111 C, such as the first two-thirds of the same inspiratory phase (e.g. 1111 C in FIG. 7). In some such examples, the greater the difference between the amplitude of the initial peak (e.g. 1 132 in FIG. 7) and the amplitude of the later segments (e.g. valley portion 1134B in FIG. 7) of the top portion (e.g. 1131 B) of the same inspiratory phase (e.g. 1111C) corresponds to a greater degree of skewness (per degree parameter 665 in FIG. 3). In some examples in which the amplitude of the initial peak (e.g. 1132) is greater than the general amplitude of the valley portion 1134B and / or the lowest amplitude (e.g. 1135B) of the top portion (e.g. 1131 B in FIG. 7) of the inspiratory phase (e.g. 111 1 C in FIG. 7), at least the top portion (1131 B) of the inspiratory phase 1111 C may be deemed as having a negative skew (e.g. being skewed with a negative slope. In this way, a greater degree of skewness (e.g. negative skew) corresponds to a relatively higher degree (e.g. proportion) of inspiratory flow occurring earlier in the same inspiratory phase of a given breath than would otherwise occur during normal, stable breathing.

[0101] In some examples, with further reference to the respiratory waveform portion 650 in FIG. 3, the peak parameter 670 may identify or track the presence (672 in FIG. 2) of at least an initial peak (e.g. 1132 in FIG. 7) of a top portion of an inspiratory phase versus a top portion of an inspiratory phase which is1618.322 11127 relatively flat or has other shapes relatively devoid of a peak. Meanwhile, the peak parameter 670 also may identify or track a location (674 in FIG. 3) of any such peak. As shown later in at least FIGS. 5, 7, a top portion of an inspiratory phase which has an initial peak (e.g. 1132 in FIG. 7) early in the inspiratory phase (e.g. 11 11 B) corresponds to a pattern associated with a complete concentric collapse and may be highly correlated with a higher degree of scoopiness (e.g. NED 678 in FIG. 3) and / or skewness (660 in FIG. 3).

[0102] It will be further understood that additional parameters identified from a sensed respiratory waveform per portion 650 (of specific patient information portion 500 in FIG. 3) may be used to determine a likelihood of CCC pattern or non-CCC pattern such as (but not limited to) the parameters of early inspiratory volume (e.g. inspiratory volume at some percentage (e.g. 30%) of a duration of the inspiratory phase), rise time (e.g. time of some percentage (e.g. elapsed time to 50% of peak inspiratory flow), early inspiratory peak flow (e.g. amplitude of an initial peak), and / or inspiratory peak time (relative to expiratory peak time).

[0103] At least some of these parameters of respiratory waveform portion 650 are further described and illustrated in association with at least FIGS. 5-8.

[0104] In some examples, sensing portion 600 comprises a drug induced sleep endoscopy (DISE) parameter 688 which corresponds to information obtained via performing a DISE procedure, such as determining a type, a degree, and / or a location of upper airway collapse observable during the DISE procedure.

[0105] In some examples, sensing portion 600 comprises an imaging portion 680, which may comprise patient information obtained via parameters regarding external and / or internal imaging 682, 684. In some such examples, the internal imaging 682 may comprise information regarding internal anatomy (e.g. not visible outside the patient’s body), whether the anatomy comprises portions of a body cavity or other internal tissue. This information may also comprise physiologic functions, internal physical state, etc. The internal imaging 682 may comprise the imaging previously described regarding a DISE procedure 688 (e.g. also 294 in FIG. 2; 1786 in FIG. 9B). In some examples, the external imaging 684 may comprise any form of imaging which captures aspects of a patient’s1618.322 11128 anatomy, physical state, physiologic functions, etc. which are visible from outside the patient’s body.

[0106] As noted above, in some examples, sensing portion 600 may comprise parameters regarding sensing impedance 690, pressure 692, vibration 694, and / or other sensing modalities 695 from which at least respiratory information and / or other patient information 500 (e.g. upper airway size, upper airway obstructions, respiratory effort, etc.) which can be sensed via such modalities.

[0107] It will be understood that at any given time, just some of the types, modalities, etc. delineated in specific patient information portion 500 may be available and that implementation method 205 and / or other example methods of the present disclosure may be implemented via such available information, i.e. without all of the information possibly available from the specific patient information portion 500. Moreover, it will be understood that in some examples the pertinence and / or applicability of any given parameter within the specific patient information portion 500 is not strictly limited to being represented within a particular portion (e.g. demographic 520, behavioral 535, EHR 540, survey 550, sensing portion 600). For instance, in some such examples, a particular parameter of specific patient information portion 500 may be represented in more than one of the respective portions (e.g. 520, 535, 540, 550, 600) or a different one of the respective portions (e.g. 520, 535, 540, 550, 600) than currently depicted in FIG. 3.

[0108] FIG. 4 is a block diagram representing example general patient information portion 800, which generally corresponds to at least some of the types of patient information in specific patient information portion 500 of FIG. 3, except for the information being obtained from data sets regarding multiple patients. For example, the general patient information of portion 800 may comprise PSG information 810 (e.g. formal sleep study data) from a large volume of sleep studies performed on a wide range of patients, which may comprise solely “SDB care” patients or may comprise patients not specifically seeking SDB care. In some examples, general patient information may be obtained from home sleep study (HST) (e.g. 614 in FIG. 3) instead of, or in addition to, the general patient information obtained from PSG studies.1618.322 11129

[0109] In some examples, the general patient information may comprise sleep care patient information 812, which may comprise patient information from a wide range (and large volume of) patients seeking any type of sleep care, regardless of whether those patients are “SDB care” patients.

[0110] In some examples, the general patient information may comprise patient information (as in FIG. 3) regarding patients which use a “SDB care” therapeutic device 830, which may comprise a device for stimulation 832 (e.g. external or implantable neurostimulator) or a forced air device 834 (CPAP, BiPAP, etc.). In some examples, this set of general population information may include candidates who did not receive a therapeutic device but were part of a patient screening process for a therapeutic device.

[0111] In some examples, this general patient information per parameter 830 may overlap with the scope of the PSG information 810 because most patients in the therapeutic device (830) category will have participated in a formal sleep study prior to using a therapeutic device. In this latter example, the general patient information (from parameter 830) may comprise a patient population specific to a particular therapeutic device manufacturer (e.g. Inspire Medical Systems, Inc. and / or others) and / or specific to research studies such as FDA research trials used to obtain FDA approval for a particular therapeutic device.

[0112] Among other aspects, at least some of these types of general patient information associated with general patient information 800 in FIG. 4 may be significantly different from other more general patient information data repositories which do not focus on patients seeking sleep care, patients seeking SDB care, patients screened for a SDB therapeutic device, and / or patients obtaining a SDB therapeutic device.

[0113] FIG. 5 is a diagram including graph 1100 representing a respiratory waveform 1102 including at least some respiratory cycles (e.g. 1110B) corresponding to a concentric collapse pattern of the upper airway. As shown in FIG. 5, the respiratory waveform 1102 comprises a series of respiratory cycles (e.g. 1110A) including an inspiration phase 1111A and an expiratory phase 1112Awith each inspiration phase (e.g. 1111 A, 1111 B, 1111 C, 1121 A) extending above dotted line 1114 and each expiration phase (e.g. 1112A, 1 112B, 11 12C1618.322 111301122A, 1122B) extending below dotted line 1114. The waveform is plotted relative to time (e.g. seconds) as represented via time increments T1 , T2, T3, T4, -T1 , -T2, -T3, -T4, etc. relative to an “end event” represented by / at vertical line C in the general center of the timeline. This same convention applies generally to FIGS. 6-8.

[0114] As further shown in FIG. 5 via dotted circle A1 , the inspiration phase 1111 A of example respiratory cycle 1110A exhibits a well-defined peak (e.g. a single peak). The respiratory cycle 1110A has a first amplitude (e.g. nominal amplitude), which spans the distance between line 1116A and 1116B. Due to some form of sleep disordered breathing, as additional respiratory cycles (e.g. breaths) occur, the amplitude of successive respiratory cycles begins to decrease while the peak of the inspiration phase of further respiratory cycles exhibit shapes other than the well-defined peak of inspiration phase 1110A at A1 . For instance, as shown at dashed circle B1 , some inspiration phases exhibit top portions including irregular shapes including double peaks. Eventually, the amplitude of each respiratory cycle becomes substantially less than the amplitude of first / nominal respiratory cycle 1110A with the amplitude expressed by respiratory cycle 1110B extending the span between dashed lines 1118A, 1118B and the highest amplitude portion of the inspiration phases 1111 B exhibiting an irregularshaped top portion (at dashed circles B2, B3) instead of a well-defined peak (e.g. single, relatively sharp peak) as at dashed circle A1. Dashed box D1 identifies several of these respiratory cycles which differ substantially from the nominal respiratory cycle(s) 1110A at least until indicator C (e.g. vertical line) which represents an end of an obstruction.

[0115] The pattern of respiratory cycles (at least until line C) exhibited in dashed box D1 represents a sleep disordered breathing event (e.g. obstructive sleep apnea (OSA)) generally corresponding to an obstruction in the upper airway caused by a complete concentric collapse (CCC) pattern in the upper airway. Further details regarding these respiratory cycles (e.g. 1 110B) are further described later in association with at least FIG. 7.

[0116] As further shown in FIG. 5, after the obstruction has ended at least temporarily (as generally represented via line C), the respiratory waveform 11021618.322 11131 exhibits respiratory cycles (e.g. 1120A) in which the inspiration phase 1121 A has a higher-than-nominal amplitude, as represented by the peak (at dashed circle E1 ), extending beyond line 1116A, and expiratory phase 1122A extending beyond line 1116B. As shown in dashed circle E1 , the peak of the inspiration phase 1121A again exhibits a well-defined peak (e.g. single, relatively sharp peak) which corresponds to a generally normal inspiration phase (albeit at a higher amplitude) in which the upper airway is no longer obstructed. However, as further respiratory cycles (e.g. breaths) occur, the waveform eventually again comprises a series of respiratory cycles (e.g. 1120B) exhibiting a substantially reduced amplitude and distorted peaks (e.g. within dashed circle F1 ) of the inspiratory phase 1121 B.

[0117] In sharp contrast to FIG. 5, FIG. 6 is a diagram including a graph 1150 representing a respiratory waveform 1152, which includes representative inspiratory phases 1160A, 1161 B, 1171 A, 1171 B, and representative expiratory phases 1162A, 1162B, 1162C, 1172B.

[0118] The respiratory waveform 1152 includes at least some respiratory cycles corresponding to a non-com plete concentric collapse (non-CCC) pattern of the upper airway in which, the inspiration phases of respiratory cycles (e.g. 1160B) corresponding to an obstructed airway exhibit a generally flat shaped top portion (e.g. dashed circles B5, B6) instead of a well-defined peak (e.g. A2 in FIG. 6) of the inspiration phase 1111 A of generally normal respiratory cycle 1160A. As further shown in FIG. 6, in a manner similar to FIG. 5, as represented via vertical line C (at which an obstruction clears), the respiratory cycles (e.g. 1170A including inspiration phase 1171 A, expiration phase 1172A) exhibit a higher-than-normal amplitude but which exhibit a well-defined peak (e.g. at dashed circle E2) before an at least partially obstructed upper airway again causes detioriation of the respiratory cycles. In particular, the respiratory cycles again trend toward substantially reduced amplitudes (e.g. extending between lines 1118A, 1118B instead of beyond 11 16A, 1116B) and exhibit a generally flat top portion 1170B (e.g. at circle F2) of an inspiration phase 1171 B.

[0119] FIG. 7 is a diagram including an enlarged portion of FIG. 5 (which generally corresponds to dashed box D1 in FIG. 5) including a graph 1200 further1618.322 11132 representing respiratory cycles corresponding to a complete concentric collapse (CCC) pattern.

[0120] As shown in FIG. 7, respiratory cycle 1110B comprises an inspiration phase 1111 B and expiratory phase 1 112B, wherein the inspiratory phase 11 11 B includes an irregularly shaped top portion 1131 A (at dashed circle B2), which comprises a first peak portion 1132 and a second peak portion 1136 with a valley portion 1134A interposed between the peak portions 1132, 1136. A successive respiratory cycle 1110C exhibits an irregularly shaped top portion 1131 B (at dashed circle B3) which includes a first peak portion 1 132, second peak portion 1136 and valley portion 1134B between peak portions 1132, 1136 with valley portion 1134B being even deeper (relative to peak portions 1132, 1134) than valley portion 1134A. In addition, the valley portion 1134B exhibits a generally negative slope and deeply concaved shape. In some examples, the combination of the first peak 1132, second peak 1136, and the negatively sloped, concave valley portions 1134A, 1134B of top portions 1131 A, 1131 B of the inspiration phase of these respiratory phases 11 10B, 1110C correspond to a complete concentric collapse (CCC) collapse pattern.

[0121] As previously described in association with at least respiratory waveform portion 650 of FIG. 3, FIG. 7 illustrates examples of applicability (to a specific sensed respiratory waveform) of various parameters such as (but not limited to) negative effort dependence (NED) parameter 678 (e.g. scoopiness), skewness parameter 660, peak parameter 670, and / or other respiratory waveform parameters, which may be used to determine whether a particular sensed respiratory waveform better corresponds to complete concentric collapse (CCC) pattern or better corresponds to a non-CCC pattern.

[0122] Accordingly, in at least some examples, the particular shapes of top portions of inspiratory phases of respiratory cycles of a respiratory waveform in FIG. 5 and 7 may comprise an attribute which may be a strong indicator of a CCC pattern, and therefore may serve as a favored attribute for determining a numerical score (at 1781 in FIG. 9B) regarding a probability of a CCC pattern.1618.322 11133

[0123] FIG. 8 is a diagram including an enlarged portion of FIG. 6 including a graph 1300 further representing respiratory cycles corresponding to a non-CCC pattern.

[0124] As shown in FIG. 8, respiratory cycle 1160B comprises an inspiration phase 1161 B including an irregularly shaped top portion 1331 A (at dashed circle B5), which comprises a first rounded corner portion 1333 and a second rounded corner portion 1337 with an intermediate portion 1335A interposed between the rounded corner portions 1333, 1337. A successive respiratory cycle 1160C exhibits an irregularly shaped top portion 1331 B (at dashed circle B6) which includes a first rounded corner portion 1333, second rounded corner portion 1337 and intermediate portion 1335B between the respective portions 1333, 1337. The intermediate portions 1335A, 1335B in FIG. 8 comprise a generally flat shape, at least in comparison to the deep valleyed shapes of the top portions 1131 A, 1131 B in FIG. 7. In one aspect, these top portions 1331 A, 1331 B along with the substantially reduced amplitude may be indicative of a non-CCC pattern, and in some instances, may be indicative of a tongue-based collapse pattern and / or epiglottis-based collapse pattern. Accordingly, in at least some examples, the particular shapes of top portions of inspiratory phases of respiratory cycles of a respiratory waveform in FIG. 6 and 8 may comprise an attribute which may be a strong indicator of a non-CCC pattern, and therefore may serve as a favored attribute for determining a numerical score (at 1781 in FIG. 9B) representative of a high probability of a non-CCC collapse pattern.

[0125] As previously described in association with at least respiratory waveform portion 650 of FIG. 3, FIG. 8 illustrates examples of applicability (to a specific sensed respiratory waveform) of various parameters such as (but not limited to) negative effort dependence (NED) parameter 678 (e.g. scoopiness), skewness parameter 660, peak parameter 670, and / or other respiratory waveform parameters (e.g. substantially reduced amplitude (e.g. at least 25%, 30%, 35%, 40%, 45%, 50% less, of the respiratory signal), which may be used to determine whether a particular sensed respiratory waveform better corresponds to complete concentric collapse (CCC) pattern or better corresponds to a non-CCC pattern. However, unlike the respiratory waveform in FIGS. 5, 7, the respiratory waveform1618.322 11134 in FIGS. 6, 8 represent an example in which a negative effort dependence (NED) parameter 678 (e.g. scoopiness), skewness parameter 660, peak parameter 670 and / or other parameters have a null value (or minimal value) or otherwise may be viewed as having little or no pertinence to the respiratory waveform for a non- CCC pattern as shown in FIGS. 6, 8. Among other observations in favor of determining a non-CCC pattern would be present, for example, the inspiratory phase 1331 A of respiratory cycle 1160B and inspiratory phase 1331 B of respiratory cycle 1160C lack a well-defined peak in the early portion of the respective top portion of the inspiratory phases (1331 A, 1331 B respectively). This pattern stands in sharp contrast to the inspiratory phase 1131A of respiratory cycle 1110B and inspiratory phase 1131 B of respiratory cycle 1110C in FIGS. 5, 7 which exhibit a well-defined peak (e.g. 1132) early in the early portion of the respective top portion of the inspiratory phases (1131 A, 1131 B respectively).

[0126] As further described later in association with at least FIGS. 9A-10B, any features of the respective respiratory waveforms of FIGS. 5-8 may be used to identify that the particular waveform (or portion thereof) corresponds to occurrence of a CCC pattern or of a non-CCC pattern. In some examples, the features of the respiratory waveforms which are indicative of a non-CCC pattern also may be further leveraged to distinguish between a tongue-based collapse pattern, epiglottis-based collapse pattern, a pharyngeal wall-based collapse pattern, and / or various combinations thereof.

[0127] FIGS. 9A-10B relate to example data models and tools used in identifying a probability of a collapse pattern (e.g. complete concentric collapse (CCC)) and / or implications on a patient SDB care pathway (e.g. 205 in FIG. 2; 1780 in FIG. 9B).

[0128] In general terms one or more types of data models may be applied to patient information (e.g. specific, general) to identify a probability of a collapse pattern for a specific patient. In some examples, the data model(s) may comprise one or more of the data model features further described later in association with at least FIG. 13.

[0129] As shown in FIG. 9A, in general terms example scoring tool 1750 uses patient information to determine a numerical score, which in some examples may1618.322 11135 comprise a probability of a collapse pattern (e.g. concentric collapse) of the upper airway for a particular patient. In some examples, the scoring tool portion 1750 comprises selectable physical attributes 1752 (e.g. gender, BMI, neck size) and / or selectable other attributes 1754 (e.g. disease burden (e.g. AHI, ESS), patient input (e.g. ESS), respiratory waveform features) of the patient. In row 1756, a value is assigned to each of the selected various attributes 1752, 1754 and in row 1758, a factor (e.g. weight) is applied to each of the various attributes. Row 1760 provides a subresult for each attribute, according to the value and factor. Column 1770 provides a score 1772 comprising a sum of the subresults or comprising the result of other quantitative operations (e.g. formula, equation) using the subresult (row 1760) of the various attributes. In some examples, the score 1772 may comprise a numerical score (e.g. a quantitative value), which may comprise a probability in some examples.

[0130] With further reference to FIG. 9A, in some examples, the respiratory waveform attribute (one of attributes 1754) may comprise a numerical score corresponding to an extent to which one or more portions of the respiratory waveform exhibit features indicative of a particular collapse pattern. For example, a relatively higher numerical value (row 1756) may be assigned to the respiratory waveform attribute (1754) when an irregularly shaped top portion (e.g. 1131 A — dashed circle B3 in FIG. 7) of an inspiration phase (1111 C) of a respiratory cycle 1110C has a high correspondence to a concentric collapse pattern (e.g. complete concentric collapse (CCC) pattern), whereas a relatively lower numerical value (row 1756) may be assigned to the respiratory waveform attribute (1754) when a generally flat-shaped top portion (e.g. 1 131 A - dashed circle B5 in FIG. 8) has a high correspondence to a non-CCC collapse pattern associated with tongue- dominated upper airway obstructions.

[0131] In some examples, via the factor parameter (1758) in FIG. 9A, the selected attributes (1752, 1754) can be assigned an appropriate weighting to reflect the relative importance (e.g. impact) which each particular attribute has toward a probability of a particular collapse pattern (e.g. complete concentric collapse (CCC)) occurring. With this in mind, it will be understood that the particular example attributes (e.g. gender, BMI, neck size (e.g. circumference),1618.322 11136AHI, etc.) shown in FIG. 9A are merely examples such that other attributes from patient information of FIGS. 3, 4 may be substituted for, or used in addition to, at least some of the attributes shown in FIG. 9A. As previously noted, the data model supporting the example sorting tool 1750 of FIG. 9A is scalable such that a fewer number or greater number of attributes 1752, 1754 (as selected from patient information in FIGS. 3-4 or other sources) may be used to determine the score 1772, where the number of attributes used may depend on availability of data values for a particular attribute, confidence in such values, etc.

[0132] In some examples, via the factor parameter (1758), the gender, BMI, and neck size attributes (1752) may be weighted greater than other attributes as being indicative of some collapse patterns (e.g. complete concentric collapse). Similarly, in some examples, via the factor parameter (1758), a respiratory waveform attribute (1754) may be given greater weight at least when the respiratory waveform attribute includes respiratory waveform shapes (and / or amplitudes) which are conspicuously associated with particular collapse patterns (e.g. concentric collapse), such as previously described in association with at least FIGS. 5-8.

[0133] FIG. 9B is a flow diagram representing an example method 1780 of a SDB patient care workflow (e.g. pathway) in relation to determining a probability of a (particular) collapse pattern. At 1781 in FIG. 9B, the example method comprises identifying (i.e. determining) a numerical score of a likelihood (e.g. probability) of a particular collapse pattern (e.g. complete concentric collapse (CCC)) and / or of a particular therapy outcome, which may be implemented via one of the example data models, methods, etc. of the present disclosure in association with at least FIGS. 1-19 and / or via other data models, methods, etc.

[0134] As further shown at 1782 in FIG. 9B, the method comprises performing a query whether the score (from 1781) meets a criteria (1783) (e.g. a first selectable criterion). In some examples, the criteria may comprise a selectable first threshold (e.g. 1794) having a numerical value, such as (but not limited to) 10%, 15%, 20%, etc. corresponding to a value expressing a probability of the particular collapse pattern (e.g. complete concentric collapse (CCC)) occurring. However, because about 20% of a pertinent population exhibits a complete1618.322 11137 concentric collapse (CCC) pattern according to some example data, some example methods may set 20% as the maximum selectable threshold value. Similarly, the incidence of a particular collapse pattern in a pertinent population data set may be used to set threshold values for use in query 1782 for comparison with an identified probability (at 1781 ) of each particular collapse pattern (or cause, such as tongue).

[0135] With further reference to query 1782 in FIG. 9B, if the score meets the criteria (1783), such as the score (e.g. 22) being equal to or greater than a threshold value of 20, then the method proceeds along the YES path 1784 via which the patient will participate in further evaluative procedures at 1786. In some examples, the further evaluative procedure may comprise the previously described DISE procedure (e.g. 294 in FIG. 3; 688 in FIG. 3). Stated differently, if at query 1782 the score (e.g. 22) exceeds the value (e.g. 20) of the first threshold (1794), the method draws the conclusion based on the objective patient information that the probability of a particular collapse pattern (e.g. complete concentric collapse (CCC)) is relatively high enough such that the further evaluative procedure (e.g. invasive surgical procedure) such as a DISE procedure is not necessary. Moreover, in view of the relatively high score and conclusion this is a CCC patient, the method 1780 would recommend not implementing a therapeutic device (e.g. certain types of electrical stimulation of a hypoglossal nerve) because poor outcomes may generally occur for a CCC patient receiving such therapy, in some examples.

[0136] On the other hand, if the query 1782 in FIG. 9B reveals that the score (e.g. 6%) does not meet the criteria, such as the score (e.g. 6%) being less than the selected threshold (e.g. 20%), then the NO path 1788 may be taken and the conclusion is drawn that this is a non-CCC patient. Via path 1788, the patient may proceed further toward implementation of a therapeutic device (e.g. implanting a nerve stimulation device) without a DISE procedure (e.g. no DISE occurs) as shown at 1790. In some examples, the value of the threshold (1794) may be selected to be a lower value such as 10%, which is significantly less (e.g. 50% less) than the incidence (e.g. 20%) at which a complete concentric collapse (CCC) pattern occurs in the pertinent population used for the comparison. In this1618.322 11138 way, this example arrangement (e.g. threshold value is 10%) may err on the side of sending more patients on the path 1784 toward further evaluative procedures (e.g. DISE), while also increasing the confidence for patients with scores (e.g. 6%) below the selected value (e.g. 10%) of the threshold (1794) that the YES path 1784 (having a DISE procedure) would be unlikely to demonstrate the patient having the particular collapse pattern (e.g. complete concentric collapse (CCC)).

[0137] Accordingly, in some examples, the example method 1780 may proceed on the path toward treatment with the therapeutic device 1790 without performing a DISE procedure or without performing an additional DISE procedure using patient information. In some examples, the patient information used in performing method 1780 may include the patient information received, obtained, and / or tracked as previously described in connection with the specific patient information portion 500 of FIG. 3 and / or general information 800 in FIG. 4. In some examples, such demographic, physical, and / or behavioral information may be used to proceed on the path toward treatment with a therapeutic device, at 1790 in FIG. 9B, where the information is obtained from sources including surveys and / or patient medical records (e.g., EHR portion 540 and survey portion 550 of FIG. 3). In some examples, at least portions of the demographic, physical, and / or behavioral information may be obtained using sensors but without a formal sleep study (e.g., home sleep study and via sensing portion 600 of FIG. 3).

[0138] As a specific example, the patient medical records in the EHR portion 540 of FIG. 3 may include information of reported snoring, tonsil grade below a threshold, and BMI below a threshold. Based on the patient information, at 1781 , the numerical score of probability of the collapse pattern is determined and queried, at 1782, for the score meeting the criteria or not. With the reported snoring, tonsil grade below the threshold, and BMI below the threshold, the query 1782 may reveal that the score (e.g. 6%) does not meet the criteria, such as the score (e.g. 6%) being less than the selected threshold (e.g. 20%), and the NO path 1788 may be taken and the conclusion is drawn that this is a non-CCC patient without performing a DISE procedure. In some examples, the patient information may additionally or alternatively include behavioral information, such1618.322 11139 as arousal thresholds, snoring profiles, and consumption habits. In some examples, the patient information may additionally or alternatively include sensed data obtained during a home sleep study. For example, the home sleep study may be used to obtain EEG and EMG information, respiratory information (e.g., waveform), cardiac information, and / or body temperature, among other information as previously described in connection with at least FIG. 3. As a specific example, the patient information may include signals used to derive airflow shape, which is used to determine the numerical score of probability of the collapse pattern, as previously described.

[0139] As made apparent via the below discussion of at least FIGS. 10A-10B, the determination of the score for identification of the probability of a particular collapse pattern (1781 in FIG. 9B) may be performed via data models other than, or in complementary relation to, the scoring tool portion 1750 of FIG. 9A.

[0140] FIG. 10A is a diagram including a graph 1900 which represents a probability of a collapse pattern (e.g. complete concentric collapse (CCC) pattern) according to an example data model. As shown in FIG. 10A, the graph 1900 includes an x-axis 1910 representing values of body mass index (B I), while a y- axis 1912 represents a probability of the collapse pattern (e.g. complete concentric collapse) expressed as a percentage. Graph 1900 includes three different plots 1920A, 1930A, 1940A based on a data model (e.g. implementing a regression analysis) using general patient information (e.g. FIG 5), with the data model of FIG. 10A also expressed in the chart 1950 of FIG. 10B. It will be understood that the general patient information in FIG. 4 may be based on at least some of the patient information in FIG. 3 which, as expressed in FIGS. 10A-10B, includes gender (522 in FIG. 3) and BMI (526 in FIG. 3).

[0141] As shown in FIGS. 10A-10B, a comparison of plot 1920A (male) and plot 1930A (female) reveals that, in general, males are more likely than females to exhibit a concentric collapse pattern (e.g. complete concentric collapse (CCC)) in view of the entire plot 1902A (male) having a position to the left of the entire plot 1930A (female) on graph 1900. This observation is confirmed via chart 1950 in FIG. 10B, which shows all of the percentage probabilities (e.g. 4%, 6%, 8%) in row 1920B (male) having a greater values than the percentage probabilities (e.g.1618.322 111401 %, 2%, 2%, etc.) in row 1930B (female) with a delta (i.e. difference) of 3%, 4%, 6%, etc. between such respective male and female probabilities for the different increasing BMI values expressed across row 1951. Row 1940B represents respective averages of the respective values within a respective column of the chart / table.

[0142] Moreover, a further comparison of males (plot 1920A in FIG. 10A, row 1920B in FIG. 10B) and females (plot 1930A in FIG. 10A, row 1930B in FIG. 10B) in the respective graph 1900 in FIG. 10A and chart 1950 in FIG. 10B also reveal that increases in BMI (1910 in FIG. 10A; 1952 in FIG. 10B) cause a substantially greater probability of concentric collapse for males than for females.

[0143] Accordingly, the graph 1900 in FIG. 10A and chart in FIG. 10B provide just one example of employing a data model using general patient information (e.g. not specific to the particular patient) to determine criteria (e.g. 1783 in FIG. 9B) by which a query (1782 in FIG. 9B) can made to evaluate a score of a particular patient regarding their likelihood of exhibiting a particular collapse pattern (e.g. complete concentric collapse (CCC)).

[0144] In further considering application of the data model of FIGS. 10A-10B to the method 1780 in FIG. 9B, in some examples the criteria 1783 (FIG. 9B) may comprise multiple thresholds (e.g. 1794, 1795) to which a score (from 1781 ) may be compared, with the different thresholds enabling different conclusions to be drawn regarding a patient pathway and / or whether additional attributes should be employed in satisfying the query at 1782. For example, forfemale patients having a score less than a first selectable threshold (1794) value of 5 (e.g. 5%), the example method 1780 may satisfy the query at 1782 with just the two attributes of gender and BMI that no evaluative procedure occur (i.e. path 1788 is taken). However, in some examples a second higher threshold (1795) value (e.g. 10) may be selected in addition to the first threshold 1794 (e.g. value 5) to define a threshold range of 5 to 10%, such that some example methods 1780 (FIG. 9B) may comprise that, for any patient score (e.g. 8) falling between (or within) the first threshold 1794 (e.g. value 5%) and second threshold 1795 (e.g. value 10%), the example method uses additional selectable attributes in query 1782 (FIG. 9B) such as (but not limited to) neck size, supine PW, respiratory waveform, etc. in1618.322 11141FIG. 9A and / or any attribute from general patient information in FIG. 4 (e.g. which may comprise at least some of the parameters in FIG. 3). Accordingly, in the example set forth above (e.g. threshold range of 5 to 10, and patient score of 8), the criteria 1783 may additionally include a respiratory waveform attribute (1754 in FIG. 9A; 650 in FIG. 3). For instance, as previously described in association with at least FIGS. 5-8, one respiratory waveform attribute may comprise a shape of a top portion of an inspiration phase of a respiratory cycle, wherein the shape of the top portion may be indicative of a particular collapse pattern. With this in mind, with gender and BMI producing a score of 8 for a specific patient, if the respiratory waveform attribute (e.g. generally flat-shaped, top portion 1331 A, 1331 B (FIG. 8) of inspiration phase) a low value representing a tongue-based, non-CCC pattern (e.g. for respiratory cycle 1160B in FIG. 8 at circle B5, a top portion 1331 A including a first rounded corner 1333, generally flat portion 1335A, second rounded corner 1337), then the particular patient’s score may remain at 8 or even be lowered, with value 8 remaining within the range of 5 to 10. Via this determination, in some examples the comparison at query 1782 would yield a NO, such that the method recommends path 1788 to proceed toward a therapeutic device without a further evaluative procedure (e.g. NO DISE procedure) at 1790. In other words, the additional respiratory waveform attribute (e.g. absence of CCC pattern) helped to confirm that the further evaluative procedure could be bypassed (e.g. omitted) in making a patient care determination in method 1780.

[0145] However, in some examples, a respiratory waveform attribute may be used to confirm a YES (path 1784) that a further evaluative procedure (1786 - DISE) is recommended or required before the patient can proceed toward a therapeutic device, with gender and BMI producing a score of 8 for a specific patient, if the respiratory waveform attribute (e.g. shape of top portion of inspiration phase) has a high value representing a complete concentric collapse pattern (e.g. for respiratory cycle 1110C in FIG. 7 at circle B3, a first sharp peak 1132, conspicuous valley 1134B, second sharp peak 1136), then the particular patient’s score may be increased to a much higher value such as 12, which exceeds the second threshold (e.g. 10) and is outside range of 5 to 10.1618.322 11142Accordingly, the comparison at query 1782 would yield a YES, such that the method recommends path 1784 in which a further evaluative procedure (e.g. DISE) be performed since the selectable threshold of 10 is still far below the general threshold of 20. However, if the value of the particular respiratory waveform attribute and / or its factor were higher because of a higher correspondence with the particular collapse pattern (e.g. CCC), then the score might be higher (e.g. 21 ) such that the method 1780 would determine that further evaluative procedure (e.g. DISE) is not warranted because the objective patient information presents a compelling conclusion for the relatively high probability of the particular collapse pattern (e.g. CCC).

[0146] In some examples, upon a patient undergoing a further evaluative procedure (e.g. DISE) at 1786 in FIG. 9B and the procedure revealing that complete concentric collapse (CCC) is not occurring or is minimal, then the patient care method 1780 may comprise the patient following a path (dotted line 1792) to proceed toward therapeutic device despite the patient having undergone a DISE procedure or other further evaluative procedure (at 1786).

[0147] It will be further understood that attributes (e.g. from FIGS. 3, 4) in addition to, or instead of, respiratory waveform attributes may be used in addition to selectable primary attributes (e.g. gender, BMI in some examples) to confirm or deny results of query 1782 in method 1780 of FIG. 9B.

[0148] It will be understood that in some examples, regardless of whether a single threshold (e.g. 1794 in FIG. 9B) or multiple thresholds (e.g. 1794, 1795, etc. in FIG. 9B) are being applied to perform the score evaluation at 1782 in FIG. 9B, the thresholds may be applied to an overall score (e.g. probability of a collapse pattern) and / or to specific attributes (e.g. BMI, neck size, supine pharyngeal width (SPW), etc.) used to determine the overall score. Moreover, as noted elsewhere, the single or multiple thresholds can be used at 1782 in FIG. 9B as criteria to evaluate a score for its positive predictive value or for its negative predictive value of a collapse pattern (and / or therapy outcome).

[0149] In some such examples, the positive predictive value and / or negative predictive value may be expressed or determined, at least in part, according to an error rate of a particular patient identified as having a high probability of a1618.322 11143 complete concentric collapse (CCC) pattern (based on a selectable threshold at 1782 in FIG. 9B) being implanted with a electrical stimulation therapeutic device (e.g. some examples of action 1790 in FIG. 9B) which does not efficaciously treat that collapse pattern. In some such examples, such a patient may sometimes be referred to as a CCC patient.

[0150] In some examples, the positive predictive value and / or negative predictive value may be expressed or determined, at least in part, according to an error rate of a particular patient identified as having a low probability of a CCC pattern (based on a selectable threshold(s) at 1782 in FIG. 9B) being sent to a further evaluative procedure 1786 in FIG. 9B (e.g. DISE), which would likely be unnecessary. In some such examples, such a patient may sometimes be referred to as a non-CCC patient.

[0151] In further considering the data model expressed in FIGS. 10A-10B (e.g. plot 1920A in graph 1900, relationships extrapolated from chart 1950) it may be observed that in the lower BMI values (e.g. 20 to 22, 22 to 24, 24 to 26), for males a 2% increase in probability of concentric collapse is present for each incremental step up the values of BMI, with even greater percentage probability value increases in the higher range of BMI values (e.g. 30 to 32, 32 to 34, 34 to 36). In some examples, this relationship may be represented via line 1928 in graph 1900 of FIG. 10A. Conversely, for females a much lower increase in probability (e.g. 0.7 %) occurs for each incremental step increase in BMI values with line 1938 in FIG. 10A expressing this relationship extrapolated from chart 1950 in FIG. 10B and the plot 1930A in FIG. 10A. In some examples, each of the relationships expressed via the respective lines 1928, 1938 may be used as a rule-of-thumb regarding a trend of increasing probability of a collapse pattern (e.g. concentric collapse) for a male and a female, respectively, for a given BMI value. In some examples, each rule-of-thumb may be implemented as a simple formula to help determine whether a particular patient may bypass participating in a further evaluative procedure 1786 (e.g. DISE) on their SDB patient care path. Similarly, an example method may extrapolate other rules-of-thumbs from relationships among attributes of patient information (FIGS. 3, 4) which contribute to identifying (i.e. making a determination) a probability of a collapse pattern (e.g. complete1618.322 11144 concentric collapse (CCC)) for a particular patient (e.g. 1781 in FIG. 9B or at 210 in FIG. 2) as part of guiding a patient along a SDB patient care path.

[0152] In addition, the regression data model expressed in FIGS. 10A, 10B (graph 1900, chart 1950) also may produce significantly different conclusions regarding a probability of a collapse pattern (e.g. concentric collapse) than conclusions drawn by using a simpler data model (e.g. FIG. 9A). In particular, using the same patient information, a simpler data model (FIG. 9A) may produce a probability of a concentric collapse pattern on the order of 10%, while the regression model (e.g. FIGS. 10A, 10B) which takes into greater account of relationships among the attributes (e.g. 1752, 1754) and their impact on a probability of concentric collapse, may produce a probability of concentric collapse which is 2x, 3x the probabilities produced via the linear model.

[0153] In some examples, implementations such as those embodied in FIGS. 10A-10B (graph 1900, chart 1950) may be expressed as, or utilize, a tree model in which patient attributes are used according to an “if, else” logic tree which addresses selectable attributes in sequence. For example, one example method may comprise using gender as a first attribute, and then select different second attributes based on the selected gender. For instance, in one example model, once the female gender is selected, then BMI becomes the second attribute to decide which patients are at low probability of a particular collapse pattern (e.g. complete concentric collapse (CCC)) such as at 1781 and query 1782 in FIG. 9B such that a further evaluative procedure 1786 (e.g. DISE) may be bypassed and the patient may proceed (e.g. path 1788) toward obtaining a therapeutic device (1790). On the other hand, once the male gender is selected, then one example method may use neck size (circumference) and BMI to determine which patients are at a low risk of a particular collapse pattern (e.g. complete concentric collapse (CCC)) such that a further evaluative procedure 1786 (e.g. DISE) may be bypassed and the patient may proceed (e.g. path 1788) toward obtaining a therapeutic device (1790). For neck size and / or BMI, a threshold may be selected by which the particular attribute may be evaluated as being indicative of a collapse pattern (e.g. complete concentric collapse) when the value of the attribute is above the threshold and being less indicative of the collapse pattern1618.322 11145 is below the threshold. In some such examples, multiple thresholds may be applied relative to an attribute.

[0154] In some examples, the method 1780 (FIG. 9B) also may be modified whereby when a patient score (from 1781 ) is sufficiently high, the method 1780 also may use the second threshold 1794 as a mechanism to bypass the further evaluative procedure (e.g. DISE) because the score (e.g. 30% of a probability of a collapse pattern (e.g. complete concentric collapse) is substantially greater (e.g. a selectable percentage such as 50%, 75%) than the selectable maximum threshold (e.g. 20%) by which high confidence exists of the occurrence of the particular collapse pattern (e.g. complete concentric collapse). Moreover, in some such examples, this additional bypass implementation (e.g. via second threshold 1795) may be used solely when particular selectable attributes form part of the score. For example, the additional bypass implementation may comprise use of the respiratory waveform attribute (obtained via PSG, home test, or other source) regarding a shape of a top portion of an inspiration phase of a respiratory cycle because this attribute has a high correlation to the particular collapse pattern. In this way, in some examples which employ the second threshold 1794, the method 1780 in FIG. 9B further leverages the patient score and second threshold 1794 to bypass the further evaluative procedure (1786, e.g. DISE), thereby further expediting a patient path toward a therapeutic device when appropriate or expediting the determination that a patient is ineligible for a therapeutic device when appropriate.

[0155] In some examples, BMI 526 and neck size (e.g. circumference) 527 may be used together as the sole set of attributes by which a score, e.g. probability of collapse pattern (e.g. complete concentric collapse) may be determined and by which a decision regarding to perform further evaluative procedure (e.g. DISE) or bypass such procedure may be executed.

[0156] While FIGS. 2, 5-8, 9A-10B have highlighted examples relating to complete concentric collapse (CCC) pattern or a tongue-based collapse pattern, it will be understood that at least some of the principles of sensing, analysis, prediction, decision-making may be equally applied for epiglottal-based collapse patterns and / or lateral wall and / or posterior wall (e.g. pharyngeal wall) based1618.322 11146 collapse patterns. Moreover, in some examples, a patient’s particular collapse pattern may involve some combination of concentric collapse, tongue-based collapse, lateral-posterior wall collapse, or epiglottal-based collapse. With this in mind, some therapeutic neurostimulation (and / or other forms of therapy) may be directed to stimulation of the muscles and / or nerves at least partially responsible for contraction, tone, etc. of the various portions of the upper airway (e.g. epiglottal portion, lateral and / or pharyngeal walls, etc.) such that the various sensing, analysis, prediction, and / or decision-making workflows may account for more than one type of collapse pattern.

[0157] FIG. 11 is a block diagram schematically representing an example stimulation portion 2200. The various functions and parameters of the stimulation portion 2200 may be implemented in a manner supportive of, and / or complementary with, the various parameters, tools, portions, data models, predictions, outcomes, etc. relating to SDB patient care (e.g. stimulation) throughout examples of the present disclosure.

[0158] In some examples, via target tissue parameter 2210, an example method may select or recommend which target tissue may best alleviate the particular collapse pattern (e.g. CCC, tongue-based non-CCC, etc.) identified via the example methods of the present disclosure, wherein the target tissue may comprise a single target (e.g. hypoglossal nerve and / or genioglossus muscle) or multiple targets such as any one or more of the below-identified nerves and / or muscles associated with promoting upper airway patency.

[0159] In some examples, via target tissue parameter 2210, stimulation may be delivered to selectable target tissues such as, but not limited to, upper airway patency-related (UAPR) tissues. In some examples, the upper airway patency- related tissue may comprise a hypoglossal nerve and / or muscle (e.g., genioglossus muscle) innervated by the hypoglossal nerve to cause contraction of at least the protrusor muscles to cause protrusion of the tongue to increase and / or maintain upper airway patency. In some examples, the upper airway patency-related tissue may comprise infrahyoid muscle-related nerves which includes nerve(s) innervating one or more infrahyoid strap muscles, such as thyrohyoid, omohyoid, sternohyoid, and / or sternothyroid. The infrahyoid muscle-1618.322 11147 related nerves may include an infrahyoid-muscle (IHM)-innervating nerve. In some examples, stimulation of the infrahyoid muscle (IHM)-innervating nerve and / or infrahyoid strap muscles may be used to alleviate pharyngeal wall- dominated collapse patterns, wherein the pharyngeal wall may comprise lateral walls and / or a posterior wall.

[0160] Moreover, via the stimulation portion 2200, a combination of target tissues may be selected such that stimulation of the multiple target tissues may work in a complementary manner to alleviate certain collapse patterns according to the level, type, and / or degree of obstruction.

[0161] In one example, target tissues may include any other muscles which affect and / or promote upper airway patency, and / or nerves which innervate such muscles. For example, at least some other target tissues may comprise muscles such as (but not limited to) the palatoglossus, palatopharnygeus, and / or levator veli palatini muscles, and nerves / nerve structures which innervate those muscles such as the pharyngeal plexus via the vagus nerve. In some examples, the “other” target tissues may comprise a tensor veli palatini muscle and / or mandibular division of the trigeminal nerve which innervate the tensor veli palatini muscle. In some examples, electrical stimulation of at least some of these “other” muscles and / or nerves may stiffen portions of the upper airway at least partially defining the upper airway at the level of the velum (soft palate) at which complete concentric collapse often occurs. Accordingly, upon an identification of a high enough probability of a concentric collapse pattern per the examples methods 205 (FIG. 2), 1780 (FIG. 9B), some examples may comprise implementing a therapeutic device (e.g. implanting a neurostimulator) to apply electrical stimulation at least some of these “other” muscles and / or nerves in order to alleviate complete concentric collapse (CCC)-patterned, obstructions to thereby promote upper airway patency. Such electrical stimulation may be applied independent of, or in complementary relation to, the various parameters of stimulation portion 2200 of FIG. 11 .

[0162] In some examples, target tissue includes a combination of nerves and / or muscles such as, but not limited to, terminal fiber ends of nerves where a nerve ending terminates into (or at) the muscle being innervated.1618.322 11148

[0163] In some examples, in addition to or instead of selecting different nerves for stimulation, the target tissue parameter 2210 also may comprise adjusting stimulation parameters via selecting between (or using a combination of) various locations along a nerve such as stimulating multiple different sites along a particular nerve, with some stimulation sites being more distal and some being more proximal.

[0164] Stimulation of target tissues implemented via stimulation portion 2200 may be controlled according to an amplitude, frequency, pulse width, duty cycle, duration, and the like to achieve desired therapeutic efficacy, which may depend on a region of the body, a type, size / shape, location of target tissue, number / location / size of stimulation electrodes and / or other stimulation elements, etc.

[0165] In some examples, in addition to or instead of selecting different nerves for stimulation, the target tissue parameter 2210 also may comprise adjusting stimulation parameters via selecting between (or using a combination of) different fascicles within a particular nerve in order to selectively stimulate target motor fibers while omitting (or minimally impacting) stimulation of other, non-target motor fibers and / or to selectively stimulate target motor fibers while omitting (or minimally impacting) stimulation of other, non-target sensory fibers.

[0166] In some examples, the stimulation portion 2200 may implement stimulation according to a bilateral parameter 2212 in which stimulation is applied to a target tissue on both sides (e.g., left and right) of the patient’s body. In some such examples, this bilateral stimulation may be delivered to the same nerve (e.g., hypoglossal nerve) on both sides of the body. However, in some examples, the bilateral stimulation may be delivered to different nerves (e.g., hypoglossal nerve and infrahyoid muscle-innervating nerve) such as stimulating one nerve (e.g., hypoglossal nerve) on a left side of the body while stimulating another nerve (e.g., infrahyoid muscle-innervating nerve) on a right side of the body, or vice versa.

[0167] In some examples, the bilateral parameter 2212 may be implemented in a manner complementary with the alternating parameter 2232, simultaneous parameter 2234, or demand parameter 2236 of multiple function 2230, as further described below.1618.322 11149

[0168] In some examples, the stimulation portion 2200 may comprise a multiple function 2230 by which various stimulation parameters may be implemented in dynamic arrangements. In some such examples, the stimulation portion 2200 may comprise an alternating parameter 2232 by which stimulation of one target tissue (e.g., hypoglossal nerve) may be alternated with stimulation of at least one other target tissue (e.g., infrahyoid muscle-innervating nerve). However, the alternating parameter 2232 also may be applied in combination with the bilateral parameter 2212 to apply stimulation to the same nerve (or different nerves) on opposite sides of the body in which stimulation may be applied on a left side of the body and then applied on the right side of the body in an alternating manner.

[0169] In some examples, the stimulation portion 2200 may comprise a simultaneous parameter 2234 by which stimulation may be applied simultaneously to at least two different target tissues. In some examples, the at least two different target tissues comprise two different nerves, such as the hypoglossal nerve and an infrahyoid muscle-innervating nerve. However, in some examples, the at least two different target tissues may comprise two different locations along the same nerve or two different fascicles of the same nerve. In some examples, the simultaneous parameter 2234 may apply stimulation per bilateral parameter 2212 simultaneously on opposite sides of the body to the same nerve (e.g., hypoglossal nerve) or different nerves.

[0170] In some examples, the stimulation portion 2200 may comprise a demand parameter 2236 by which stimulation may be applied to one or more nerves on a demand basis. For example, stimulation may be applied to one nerve (e.g., hypoglossal nerve) which may be sufficient to achieve the patient metric (e.g., therapy outcome and / or usage) for most nights, for most sleeping positions (e.g., left and right lateral decubitus, prone), etc. but may become insufficient for some nights (e.g., after consuming alcohol or certain drugs which relax upper airway muscles), some sleeping positions (e.g., supine). In the latter situation, in order to achieve the target patient metric, via the demand parameter 2236, stimulation of a different nerve (e.g., infrahyoid muscle-innervating nerve) may be implemented in addition to, or instead of, stimulation of the first nerve (e.g., hypoglossal nerve) which was previously being stimulated. In some examples,1618.322 11150 the first or primary nerve being stimulated may be a nerve other than the hypoglossal nerve such as, but not limited to, the infrahyoid muscle-innervating nerve.

[0171] In some examples, at least some aspects of the parameters (2232, 2234, 2236) of the multiple function 2230 may be implemented via selectively activating one or more therapeutic medical devices (TMD - e.g. 4285 in FIG. 17) placed into stimulating relation to the selected target tissues.

[0172] In some examples, the stimulation portion 2200 also may further implement at least some aspects of therapeutic medical device (TMD) (e.g. 4285 in FIG. 17), the parameters 2210, 2212, 2230 of stimulation portion 2200, and / or other stimulation examples according to one or more of a closed loop parameter 2220, open loop parameter 2222, titration parameter 2224, and relationship parameter 2238. Examples can include other parameters, as shown by other parameter 2227.

[0173] In some examples, the stimulation portion 2200 comprises a closed loop parameter 2220 to deliver stimulation therapy based on sensed patient physiologic information and / or other information (e.g., environmental, temporal, etc.). In some such examples, via the closed loop parameter 2220 the sensed information may be used to control the particular timing of the stimulation according to respiratory information, in which the stimulation pulses are triggered by or synchronized with specific portions (e.g., inspiratory phase) of the patient’s respiratory cycle(s). In some such examples and as previously described, this respiratory information and / or other information used with the closed loop parameter 2220 may be determined via the sensors, sensing elements, devices, sensing portions, as previously described.

[0174] In some examples, with or without timing stimulation relative to sensed respiratory information, the closed loop mode (2220) may comprise delivering stimulation therapy in response to sensed disease burden parameter, such as the average number of apnea events per a time period, such as an apnea-hypopnea index (AHI) of average number of apnea events per hour. Another disease burden parameter may comprise a degree of upper airway patency, among other physiologic indicators. For example, for some periods of time within a nightly1618.322 11151 treatment period or over the course of several days / weeks, a patient may experience few sleep disordered breathing events (e.g., number of apnea events (per hour) lower than a selectable threshold), such that stimulation therapy may not be delivered. However, upon the patient beginning to experience sleep disordered breathing at a level high enough to warrant stimulation therapy (e.g., number of apnea events (per hour) higher than a selectable threshold), then via the closed loop parameter 2220, stimulation therapy may be delivered until a disease burden parameter (determined according to sensed information) is lower than a threshold and / or maintained generally to precent re-occurrence.

[0175] In some examples the stimulation portion 2200 comprises an open loop parameter (e.g., 2222 in FIG. 11 ) by which stimulation therapy (e.g., “use”) is applied without a feedback loop of sensed physiologic information. In some such examples, in an open loop mode the stimulation therapy is applied during a treatment period without (e.g., independent of) information sensed regarding the patient’s sleep quality, sleep state, respiratory phase, AHI, etc. In some such examples, in an open loop mode the stimulation therapy is applied during a treatment period without (i.e. independent of) particular knowledge of the patient’s respiratory cycle information.

[0176] In some examples the stimulation portion 2200 comprises a titration parameter 2224 by which an intensity of stimulation therapy can be titrated (i.e. adjusted) to be more intense (e.g., higher amplitude, greater frequency, and / or greater pulse width) or to be less intense within a nightly treatment period. However, it will be understood that the previously described examples of stimulation may be performed without (e.g., independent of) a titration parameter 2224 and instead be based on titration according to a time period parameter of more than a day, such as supra-day time period.

[0177] In some examples, the titration parameter 2224 may be implemented as automatic titration while in some examples, the titration parameter 2224 may be implemented via manual titration by a patient (or clinician). In some examples, the titration parameter 2224 may be implemented via combination of patient / manual titration and automatic titration to guide the patient in a manner complementary with their manual titration.1618.322 11152

[0178] In some such examples and as previously described, such titration may be implemented at least partially based on sleep quality, which may be obtained via sensed physiologic information, in some examples. It will be understood that such examples may be employed with synchronizing stimulation to sensed respiratory information (i.e. closed loop stimulation) or may be employed without synchronizing stimulation to sensed respiratory information (i.e. open loop stimulation).

[0179] As further shown in FIG. 11 , in some examples, the titration parameter 2224 may be implemented via a cooperation parameter 2225 when multiple TMDs (e.g. neurostimulator) are implanted and configured to operate in cooperation with each other or configured to operate via an independent parameter 2226 wherein the multiple TMDs operate independently of each other (e.g., and may operate without communicating with each other).

[0180] In some such examples, the efficacy of operation comprises an efficacy of sensing and / or an efficacy of stimulation, either of which may be determined with respect to a disease burden parameter, in some examples.

[0181] In some examples per the titration cooperation parameter 2225, the adjusting of the selective operation comprises, via communication between the respective multiple TMDs, coordinating adjusting of the selective operation of the different TMDs relative to each other to enhance their efficacy of operation, which may comprise an efficacy of sensing and / or an efficacy of stimulation.

[0182] In some examples, at least some aspects of the titration parameter 2224 of the stimulation portion 2200 at least some of substantially the same features and attributes as described in US Patent No. 8,938,299 to Christopherson et al., issued January 20, 2015, entitled SYSTEM FOR TREATING SLEEP DISORDERED BREATHING, and which is hereby incorporated by reference in its entirety.

[0183] In some examples, via the relationship parameter 2238, stimulation may be implemented according to a relationship between the multiple IMDs (if present), a relationship between different target tissues, and / or other relationships relating to sensing and / or stimulation of target tissues.1618.322 11153

[0184] In some examples, stimulation portion may be implemented via FIG. 17 and / or comprises example implementation of FIG. 7

[0185] FIGS. 12-15 relate to training a constructed data model, and using the constructed data model, to identify a probability of a particular collapse pattern (e.g. a concentric collapse) and / or predicting a therapy outcome.

[0186] As shown at 2500 in FIG. 12, the method may include constructing a data model to identify a probability of a collapse pattern via known inputs corresponding to at least patient information relative to known outputs corresponding to the probability of the collapse pattern. In some such examples, the data model may be constructed via training the data model. In addition to constructing a data model to identify a probability of a collapse pattern, or instead of constructing a data model to identify a probability of a collapse pattern, some examples may comprise constructing a data model to predict a therapy outcome via known inputs corresponding to at least patient information relative to known outputs corresponding to the probability of the collapse pattern. In some such examples, the identification of the probability of a collapse pattern (e.g. such as via a data model) may be complimentary with, and / or may comprise at least part of an example implementation of, predicting a therapy outcome or determining patient eligibility for a particular type of sleep disordered breathing (SDB) therapy. In some of these examples, one type of SDB therapy may comprise electrical stimulation of an upper airway patency-related tissue. With these examples in mind, some further examples comprise predicting a successful therapy outcome of electrical stimulation of upper airway patency-related tissue (e.g. stimulating the hypoglossal nerve and / or other target tissues) directly from patient information (e.g. FIG. 3) such as (but not limited to) directly from PSG information (e.g. 620 in FIG. 3) sensed for / from a particular patient.

[0187] In some examples, various information relating to the example methods may be displayed via a user interface (e.g. 4540 in FIG. 18C), such as (but not limited to) displaying known outputs and / or known inputs of a constructable data model during training, displaying outputs and / or inputs of a constructed data model before, during, and / or after use of the data model. It will be understood that the displayed known inputs or current inputs may comprise patient1618.322 11154 information (e.g. FIGS. 3, 4). Moreover, the user interface may be used to display, and / or permit interaction with, the various tools, stimulation portions, SDB patient care paths, decisions, results, etc. of at least some of the various examples of the present disclosure. Display of the SDB patient care path and / or SDB patient care workflow (e.g. FIG. 2, 9C) may be used to facilitate triaging care, from the perspective of a clinician and / or from the perspective of a patient. Among other uses of these types of displayed information, in some example implementations, a method may include determining which patients in a SBD patient care pathway (e.g. pipeline) to treat first and / or which patients may skip certain steps (e.g. a DISE procedure) along the SDB patient care pathway.

[0188] In some examples, the user interface may be used to display information about how, in view of the identified probability of a collapse pattern (e.g. concentric collapse), a predicted therapy outcome for SDB stimulation therapies may compare to predicted outcomes of non-stimulation therapies such as (but not limited to) oral appliances, medications, surgical interventions, etc.

[0189] In some examples, in displaying patient information, the user interface may be used to display an electronic health record (EHR) (e.g. 540 in FIG. 3), which may comprise specific patient information which may be used as an input to a data model. However, in some examples, the electronic health record also may be used to receive, store, track and display outputs of a data model, as well as any of the various forms of information, patient care paths, workflows, etc. as described and illustrated in association with at least FIGS. 1-19.

[0190] In some examples, the data model may comprise at least one of the data model types 2530 shown in FIG. 13. In considering the various model types, it will be understood that at least some of the model types may operate in a complementary manner and that at least some of the model types may comprise machine learning models. As shown in FIG. 13, at least some example model types may be grouped (2540) according to their output such as models employing regression 2542, classification 2544, or clustering (unsupervised) 2546, and in some examples, these models may further comprise linear 2550 or non-linear 2552 aspects. For instance, a linear model may comprise a linear relationship between inputs and outputs of the model.1618.322 11155

[0191] In some examples, the regression model(s) 2542 may comprise the prediction of continuous values, and these regression models may be linear or non-linear (e.g. (e.g. at least some decision tree implementations have continuous output, neural networks, etc.). In some examples, some regression models may comprise deep learning models (e.g. neural networks), depending on the final layer in the network. In some examples, such a regression model may include example implementations expressed via a 0-1 continuous probability score of chance of collapse (of an upper airway).

[0192] In some examples, the classification model(s) 2544 may comprise predicting discrete values and may comprise models such as a logistic regression model or deep learning models, depending on the format of the final layer. In one context, the model may include example implementations which express a True / False value for the presence of complete concentric collapse (CCC). In some examples, such models also may, at a deeper layer, express a specific VOTE score. At least some of these classification models may be linear or nonlinear.

[0193] In some examples, the clustering (unsupervised) model 2546 may comprise automatic grouping and / or restructuring of data according to some statistics. In some example implementations, these models may be used in association with inputs to predictive models such as (but not limited to), performing clustering on isolated inspiration phases and using the clustering labels as one of many features as input to the predictive model.

[0194] As further non-limiting examples, some model types may comprise an artificial neural network 2560, support vector machine (SVM) 2562, deep learning 2564, and / or other models 2566.

[0195] Meanwhile, further example model types may comprise a correlation table 2570, a data structure 2722, among other models 2574, and which may include the above-described patterns and / or a probabilistic approach, which may be known inputs.

[0196] In some examples, per type 2546, the data model may comprise a clustering method(s), which may comprise hierarchical clustering, k-means clustering, density-based clustering, and the like. In some examples, the1618.322 11156 hierarchical clustering may be used to construct a hierarchy of clusters of sensed data. In some such examples, the hierarchical clustering utilizes a “bottom up” approach (e.g., agglomerative) wherein each data point starts in its own cluster, and pairs of clusters are merged at progressively higher levels of the hierarchy. However, in some examples, the hierarchical clustering utilizes a top-down approach in which all data points start in one cluster, and then clusters are split at progressively lower levels of the hierarchy.

[0197] In some examples, the k-means clustering implementation may comprise placing the sensed data into k clusters, where k is an integer equal or greater than two. Via such clustering, each data point belongs to a cluster having a mean that is closer to the data point than any means of the other clusters. However, in some examples, a machine learning model (MLM) may comprise density-based clustering, which may be used to group together data points that are close to one another, while identifying as outliers any data points that are far away from other data points.

[0198] In some examples, the artificial neural network 2560 may estimate a function(s) that depend on inputs. In some such examples, one or more layers of artificial neurons may receive input data and generate output data. Neural networks may comprise networks such as, but not limited to, learning networks (e.g., deep, deep structured, hierarchical, and the like), convolutional, auto-type networks (e.g., auto-encoder, auto-associator), Diablo networks, and neural network models (e.g., feedforward, recurrent).

[0199] In some examples, the SVM 2562 may utilize a linear classification. This classification may act to separate the data points into classes based on distance of the data points from a hyperplane. In some examples, the hyperplane is arranged to maximize the distances from the hyperplane to the nearest data points on either side of the hyperplane. This arrangement may group points located on opposite sides of the hyperplane into different classes. However, in some examples, the SVM 2562 may comprise a nonlinear classification that separates the data points with a hyperplane in a transformed feature space. The transformed feature space may be determined by one or more kernel functions, including nonlinear kernel functions. In some examples, the SVM 2562 is a1618.322 11157 multiclass SVM that separates data points into more than two classes, which may reduce a multiclass problem into multiple binary classification problems.

[0200] In some examples, the deep learning 2564 may comprise models such as, but not limited to, convolutional networks (e.g., deep belief, neural), belief networks, Boltzmann machines, deep coding networks, stacked auto-encoders, stacking networks (e.g., deep or tensor deep), hierarchical-deep models, deep kernel machines, and the like. It will be understood that such examples may comprise variants and / or combinations of the above-noted example networks.

[0201] In some examples, as represented per “other” type 2566 in FIG. 13, a MLM may comprise a mean-shift analysis that may be used to determine the maxima of a density function based on discrete data sampled from that function.

[0202] In some examples, as represented per “other” type 2566 in FIG. 13, a MLM may comprise structured prediction techniques and / or structured learning techniques. Such techniques may be used to predict structured objects and / or structured data. In some such examples, such structured prediction and / or structured learning techniques may comprise graphical models, probabilistic graphical models, sequence labeling, conditional random fields, parsing, collective classification, bipartite matching, Bayesian networks or models, and the like. It will be understood that such examples comprise variants and / or combinations of the above-noted example techniques.

[0203] In some examples, a MLM may comprise anomaly detection and / or outlier detection that may be used to identify data that does not conform to an expected pattern or are otherwise distinct from other data in a dataset.

[0204] In some examples, machine learning model may comprise learning methods that incorporate a plurality of the machine learning methods.

[0205] It will be understood that at least some example methods (and / or devices) of the present disclosure may receive and / or obtain at least patient information (e.g. 400 in FIG. 3, 500 in FIG. 4) for use in identifying a probability of a collapse pattern (e.g. concentric collapse), predicting a therapy outcome (e.g. response to SDB therapy), and / or determining patient eligibility - - all without use of a constructed data model and / or without use of trained data model, such as but not limited to, a machine learning model.1618.322 11158

[0206] FIG. 14 is a diagram schematically representing an example method 2700 of constructing a data model for use in later identifying a probability of a collapse pattern. As shown in FIG.14, the method 2700 comprises constructing a data model by providing known inputs 2702 and known outputs 2704 to the constructable data model 2710. The known inputs 2702 may be obtained and / or sensed via at least one implanted sensor of a particular IMD and / or via implanted sensors of a plurality of representative IMDs. In some examples, at least some of the known inputs 2702 and / or the known outputs 2704 may be obtained and / or sensed via at least one sensor located external to the patient’s body, herein sometimes referred to as “an external sensor”.

[0207] The known inputs 2702 may comprise patient information including specific patient information 500 (FIG. 3), general patient information 800 (FIG. 4), and / or other information. The known input sources 2702 may include various external and internal data sources, at least some of which are previously described in association with at least FIGS. 3-4 (400, 500). Additionally, the data model may be updated over time using feedback data (which may include updated patient information (FIG. 4, 5) regarding patient outcomes (e.g. patient response such as reduced disease burden or unchanged disease burden). In some such examples, the updating of the data model may sometimes comprise continual updating (e.g. re-training the data model) in the sense that each time new outcome data, patient information (e.g. specific and / or general), etc. becomes available, the data model is updated.

[0208] The known outputs 2704 may comprise indicators 2706 of a probability of a particular collapse pattern(s). In some examples, the output 2704 may comprise a probability of a therapy outcome.

[0209] As previously described, constructing the data model may comprise training a data model, such as one of the data models in data model types 2530 in FIG. 13 with one of the example data model types comprising a machine learning model. By providing such known inputs 2702 and known outputs 2704 to the constructable data model 2710 (FIG. 14), a constructed data model 2760 (FIG. 15) may be obtained. As noted elsewhere, the constructable data model 2710 (FIG. 14) may comprise a trainable MLM and the constructed data model1618.322 111592760 (FIG. 15) may comprise a trained MLM. In the particular example, the constructable data model 2710 (FIG. 14) is trained (forming the constructed data model 2760) using patient information (e.g. 500 in FIG. 3; 800 in FIG. 4). Once constructed, the data model 2760 as illustrated by FIG. 15, may be used in a method 2765 in which currently sensed inputs 2752 are fed into the constructed data model 2760, which produces an output 2754 as an indicator 2756 of probability of a particular collapse pattern for the specific patient.

[0210] FIG. 15 is a diagram schematically representing an example method 2765 of using a constructed data model 2760 for identifying a probability of a collapse pattern (e.g. complete concentric collapse (CCC)) using at least patient information as noted above. As shown in FIG. 15, inputs 2752 (e.g. specific patient information 500 in FIG. 3 and / or other patient information as described throughout various examples of the present disclosure)) are fed into the constructed data model 2760 (e.g., trained MLM), which then produces determinable outputs 2754, such as the indicator 2756 of the probability of a particular collapse pattern(s). In some examples, the output 2754 may comprise a therapy outcome.

[0211] In one aspect, the example data models are scalable, e.g. can be implemented with whatever subset of patient information is available at any particular timeframe the data model is constructed and used.

[0212] FIGS. 16A-16D are diagrams including front and side views schematically representing patient anatomy and example methods relating to collapse patterns associated with upper airway patency. More specifically, FIGS. 16A-16D are a series of diagrams schematically representing at least some different upper airway collapse patterns, including an anterior-posterior (AP) collapse pattern (FIG. 16A), a concentric collapse pattern (FIG. 16B), a lateral collapse pattern (FIG. 16C), and an anterior-posterior (AP) - lateral collapse pattern (FIG. 16D). In addition to observing such collapse patterns and / or other collapse patterns, at least some aspects of such collapse patterns may be measured, such as via impedance sensing using implanted electrodes (e.g., sensing elements and / or stimulation elements), using externally applied arrays of electrodes and / or other devices.1618.322 11160

[0213] By determining an upper airway collapse pattern, some example arrangements may determine whether to apply stimulation via a hypoglossal nerve (and / or genioglossus muscle), via an infrahyoid muscle (IHM)-innervating nerve (including which single or multiple portions thereof to stimulate), infrahyoid strap muscles, and / or via other nerves and muscles which may promote upper airway patency, and / or combinations of these nerves / muscles including unilateral and bilateral options. The “other” nerves / muscles include at least some of the nerves / muscles described in association with at least target tissue parameter 2210 of FIG. 1 1.

[0214] At least some more specific details regarding FIGS. 16A-16D are further described below in relation to at least FIGS. 16E-16G.

[0215] FIGS. 16E-16G are block diagrams schematically representing example devices and / or example methods relating to collapse patterns associated with upper airway patency.

[0216] FIG. 16E is a side view of example patient anatomy of a head-and-neck portion, which schematically represents an upper airway and related tissues. Meanwhile, FIG. 16F is a block diagram schematically representing an example sorting tool 3660 by which to sort and weigh a location, pattern, and degree of obstruction or patency. As shown in FIG. 16F, obstruction sorting tool 3660 includes functions for location detection 3662, pattern detection 3670, and degree detection 3680. In general terms, the location detection function 3662 operates to identify a site along the upper airway at which an obstruction occurs, and which is believed to cause sleep disordered breathing. In one example, the location detection function 3662 includes a velum (soft palate) parameter 3664, an oropharynx-tongue base parameter 3666, and an epiglottis / larynx parameter 3668. Each respective parameter denotes an obstruction identified in the respective physiologic territories of the velum (soft palate), oropharnyx-tongue base, and epiglottis which are generally illustrated for an example patient in FIG. 16E. In one aspect, these distinct physiologic territories define an array of vertical strata within the upper airway. Moreover, each separate physiologic territory (e.g., vertical portion along the upper airway) exhibits a distinct characteristic behavior regarding obstructions and associated impact on breathing during sleep.1618.322 11161Accordingly, each physiologic territory responds differently to activation of the different nerves and / or muscles which may alone, or in combination, promote upper airway patency.

[0217] With this in mind, the velum (soft palate parameter 3664 denotes obstructions taking place in the level of the region of the velum (soft palate) 3560, as illustrated in association with FIG. 16F. In some examples, obstructions taking place at this level (i.e. location) may typically comprise a concentric collapse pattern (e.g. diagram 3520 in FIG. 16B; parameter 3676 in FIG. 16F) including a complete concentric collapse (CCC), in some examples. However, in some examples, other levels along the airway may sometimes contribute to a CCC pattern.

[0218] FIG. 16E is a diagram including a side view schematically representing at least some anatomical features of the upper airway, as well as different sites or levels at which obstruction may occur. By determining a site or location of upper airway collapse, some example arrangements may determine whether to apply activation via a hypoglossal nerve, genioglossus muscle, an IHM- innervating nerve (including which portions thereof to stimulate), infrahyoid strap muscles, and / or via other nerves / muscles which contribute to upper airway patency, and / or combinations of these nerves / muscles including unilateral and bilateral options.

[0219] As shown in FIG. 16E, a diagram 3540 provides a side sectional view (cross hatching omitted for illustrative clarity) of a head-and-neck region 3542 of a patient. In particular, an upper airway portion 3550 extends from the mouth region 3544 to a neck portion 3553. The upper airway portion 3550 includes a velum (soft palate) region 3560, an oropharynx region 3562, and an epiglottis region 3564. The velum (soft palate) region 3560 includes an area extending below sinus 3561 , and including the soft palate 3560, approximately to the point at which tip 3548 of the soft palate 3546 meets a portion of tongue 3547 at the back of the mouth region 3544. The oropharynx region 3562 extends approximately from the tip of the soft palate 3546 (when in a closed position) along the base 3552 of the tongue 3547 until reaching approximately the tip region of the epiglottis 3554. The epiglottis-larynx region 3562 extends1618.322 11162 approximately from the tip of the epiglottis 3554 downwardly to a point above the esophagus 3557.

[0220] As will be understood from FIG. 16E, each of these respective regions 3560, 3562, 3564 within the upper airway correspond to the respective velum parameter 3664, oropharynx parameter 3666, and epiglottis parameter 3668, respectively of FIG. 16F.

[0221] FIG. 16E further illustrates relative location of the hyoid bone 3563 and thyroid cartilage 3565, as illustrated by dashed lines and with the arrows illustrating the direction of the movement of thyroid cartilage 3565, and optionally, the hyoid bone 3563, in response to activation of (e.g., electrical stimulation at) the target location of infrahyoid muscle (IHM)-related tissue (nerve and / or muscle), in accordance with some examples of the present disclosure.

[0222] The thyroid cartilage 3565 is connected to pharyngeal muscles connected to the pharyngeal walls (such as oropharynx walls) and pulling the thyroid cartilage 3565 down effectively causes the pharyngeal walls (e.g., oropharynx walls) to displace and / or redistribute tissue (e.g., at least adipose tissue) in at least the oropharynx portion 3562 to reduce extraluminal tissue pressure, which may increase and / or maintain patency of the at least the oropharynx portion of the upper airway 3550. For example, the thyroid cartilage 3565 may be connected to the inferior pharyngeal constrictor muscle, the stylopharyngeus muscle, and the thyrohyoid muscle.

[0223] As shown, the hyoid bone 3563 relates to the base 3552 of the tongue 3547 (e.g., genioglossus muscle). In some examples, pulling the hyoid bone 3563 inferiorly, as shown by the arrow, may pull on the middle pharyngeal constrictor muscle which effectively increases upper airway patency. In some examples, activating an infrahyoid strap muscle directly and / or via activating an infrahyoidmuscle (IHM)-innervating nerve may cause this action on the hyoid bone.

[0224] In some examples, moving the hyoid bone 3563 inferiorly may elongate (e.g., stretch, tug) at least one at least one pharyngeal constrictor muscle, such as the middle constrictor muscle(s). For example, the middle pharyngeal constrictor muscle may attach to the hyoid bone 3563 and depression of the hyoid bone 3563 may cause the middle pharyngeal constrictor muscle to elongate (e.g.,1618.322 11163 stretch) and increase airway patency in at least the oropharynx portion 3562. In some examples, elongating (e.g., stretching) the at least one pharyngeal constrictor muscle may stiffen the upper airway (e.g., increases pharyngeal muscle tone) and reduce collapsibility of the upper airway. In some examples, the hyoid bone 3563 may not move in a purely superior-inferior orientation. As such, as used herein, the hyoid bone 3563 being moved inferiorly may include moving generally inferiorly. For example, the patency of upper airway 3550 may increase wall stiffness (at least partially defined by pharyngeal muscles) become stiffened / stretched and / or to move in an orientation (e.g., superior-inferior, anterior-posterior, and / or medial-lateral), with such stiffening and / or movement acting to increase patency of the oropharynx portion.

[0225] In some examples, and as described above, activating the at least one IHM-innervating nerve (e.g., 826 in FIG. 7) or at least one IHM (e.g., 828 in FIG. 7) at or near a target location may cause a physiological response due to activation (e.g., contraction) of at least one IHM (e.g., infrahyoid strap muscle). The physiological response may include at least one of the thyroid cartilage 3565 moving inferiorly and the hyoid bone 3563 moving inferiorly (as described above), and which causes a physiological effect for treating SDB that occurs remotely from the activation and / or remotely from the physiological response, e.g., movement of the thyroid cartilage 3565 and / or movement of the thyroid cartilage 3565 and hyoid bone 3563 as described above. In some examples, the physiological effect comprises opening at least the oropharynx portion and / or stiffening of a pharyngeal wall of the patient (which at least partially forms a lumen of the oropharynx portion), which occurs remotely from the physiological response to the activation of moving at least the thyroid cartilage inferiorly.

[0226] Accordingly, in some examples, the physiological effect occurs a distance away from the activation applied at the target location and / or from the physiological response caused by the activation. For example, the thyroid cartilage 3565 moving inferiorly (and, optionally, the hyoid bone 3563 moving inferiorly) in response to activation of the IHM-innervating nerve and / orthe at least one IHM may occur a distance away from the physiological effect for treating the SDB (which occurs in or near the oropharynx portion 3562). The distance may be1618.322 11164 a multiple of a diameter of the upper airway 3550 of the patient. For example, the physiological effect may comprise stiffening of a pharyngeal wall (e.g., at least in the oropharynx portion 3562) of the patient which occurs remotely from the thyroid cartilage movement action (e.g., near to reference numeral 3565).

[0227] With these aspects of stimulation in mind, stimulation of the above- named nerves and / or muscles may be employed to overcome at least some pharyngeal-based (e.g. lateral and / or posterior wall) collapse patterns.

[0228] With further reference to FIG. 16F, in general terms the pattern detection function 3670 enables detecting and determining a particular pattern of an obstruction of the upper airway. In one example, the pattern detection function 3670 includes an antero-posterior parameter 3672, a lateral parameter 3674, antero-posterior-lateral (AP-Lateral) parameter 3675, a concentric parameter 3676, and composite parameter 3678.

[0229] The antero-posterior parameter 3672 of pattern detection function 3670 (FIG. 16F) denotes a collapse of the upper airway that occurs in the anteroposterior orientation, as further illustrated in the diagram 3510 of FIG. 16A. In FIG. 16A, arrows 3511 and 3512 indicate one example direction in which the tissue of the upper airway collapses, resulting in the narrowed air passage 3514. FIG. 16A also may be illustrative of a collapse of the upper airway in the soft palate region 3560, whether or not the collapse occurs in an antero-posterior orientation. For example, in some instances, the velum (soft palate) region 3560 exhibits a concentric (e.g., circular) pattern of collapse, as shown in diagram 3520 of FIG. 16B. In some such examples, the concentric collapse pattern may comprise a complete concentric collapse (CCC) pattern.

[0230] The concentric parameter 3676 of pattern detection function 3670 (FIG. 16F) denotes a collapse of the upper airway that occurs in a concentric orientation, as further illustrated in the diagram 3520 of FIG. 16B. In FIG. 16B, arrows 3522 indicate the direction in which the tissue of the upper airway collapses, resulting in the narrowed (or completely obstructed) air passage 3524.

[0231] The lateral parameter 3674 of pattern detection function 3670 (FIG. 16F) denotes a collapse of the upper airway that occurs in a lateral orientation, as further illustrated in the diagram 3530 of FIG. 16C. In FIG. 16C, arrows 3532 and1618.322 111653533 indicate the direction in which the tissue of the upper airway collapses, resulting in the narrowed air passage 3535.

[0232] The AP lateral parameter 3675 of pattern detection function (FIG. 16F corresponds to the AP lateral collapse pattern schematically represented in the diagram 3536 of FIG. 16D, which comprises a combination of the anterior- posterior pattern (FIG. 16A) and the lateral pattern (FIG. 16C) with arrows 3537A, 3537B, 3537C in FIG. 16D indicating example directions in which the tissue of the upper airway collapses, resulting in the narrowed air passage 3538. The narrowed air passage 3538 may comprise a triangular shape in some examples. In some examples, the AP-lateral collapse pattern at a velum / soft palate (3560 in FIG. 16E, 3664 in FIGS. 23F-23G) may respond better (e.g., increase patency) to activation of an infrahyoid-based patency tissue than a concentric collapse pattern which has a severity / completeness similar to the AP-lateral collapse pattern at the soft palate.

[0233] The composite parameter 3678 of pattern detection function 3670 (FIG. 16F) denotes a collapse of the upper airway portion that occurs via a combination of the other mechanisms (lateral, concentric, antero-posterior) or that is otherwise ill-defined from a geometric viewpoint but that results in a functional obstruction of the upper airway portion.

[0234] With further reference to obstruction sorting tool 3660 of FIG. 16F, in general terms the degree detection function or module 3680 indicates a relative degree of collapse or obstruction of the upper airway portion. In some examples, the degree detection function 3680 includes a none parameter 3682 a partial collapse parameter 3684, and a complete collapse parameter 3685. In some examples, the none parameter 3682 may correspond to a collapse of 25 percent or less, while the partial collapse parameter 3684 may correspond to a collapse of between about 25 to 75%, and the complete collapse parameter 3685 may correspond to a collapse of greater than 75 percent. In some examples, the at least one respiration parameter sensed from the first target tissue may include respiratory obstruction information, such as neural activity which is indicative of a relative degree of collapse or obstruction of the upper airway.1618.322 11166

[0235] It will be understood that various patterns of collapse occur at different levels of the upper airway portion and that the level of the upper airway in which a particular pattern of collapse appears can vary from patient-to-patient.

[0236] In some examples, obstruction sorting tool 3660 comprises a weighting function 3686 and score function 3687. In general terms, the weighting function 3686 assigns a weight to each of the location, pattern, and / or degree parameters (FIG. 16F) as one or more those respective parameters can contribute more heavily to the patient exhibiting sleep disordered breathing or to being more responsive to implantable upper airway activation. More particularly, each respective parameter (e.g., antero-posterior 3672, lateral 3674, AP-lateral (combination of antero-posterior and lateral) 3675, concentric 3676, composite 3678) of each respective detection modules (e.g., pattern detection function 3670) is assigned a weight corresponding to whether or not the patient is eligible for receiving implantable upper airway activation. Accordingly, the presence of or lack of a particular pattern of obstruction (or location or degree) will be become part of an overall score (according to score parameter 3687) for an obstruction vector indicative how likely the patient will respond to therapy via an implantable upper airway activation system.

[0237] FIG. 16G is diagram (e.g., chart) 3690 schematically representing an index or scoring tool to sort and weigh a location, pattern, and degree of obstruction or patency for a particular patient. Chart 3690 combines information regarding location (3662 in FIG. 16F), pattern (3670 in FIG. 16F), and degree (3680 in FIG. 16F) into a single informational grid or tool by which the obstruction is documented for a particular patient and by which appropriate activation settings may be determined and applied according to the various examples of the present disclosure, such as but not limited to those in association with at least FIGS. 1- 16F and 17-19.

[0238] Accordingly, in some examples, the information sensed and collected via at least FIGS. 16F-16G may be used to determine whether to implement activation of a hypoglossal nerve, a genioglossus muscle, an IHM-innervating nerve (including which single portion or multiple portions thereof to stimulate), an infrahyoid strap muscle, and / or other nerves / muscles which may contribute to1618.322 11167 upper airway patency, and / or combinations of these nerves / muscles including unilateral and bilateral options.

[0239] In some examples, information sensed and collected via the framework of FIGS. 16A-16G may be incorporated into the instructions 451 1 and / or information 4512 of control portion 4500 in FIG. 18A. In some examples, information sensed and collected via at least some aspects of the framework of FIGS. 16A-16G may be implemented via, and / or comprise, at least some of the patient information 500 of FIG. 3 and / or other patient information within examples of the present disclosure. In some such examples, via the updating of such instructions / information example devices and methods may adapt sensing and / or stimulation of target tissues dynamically (e.g., auto-titration) in order to efficaciously treat sleep disordered breathing (SDB) according to: changes in collapse patterns and levels at which collapse occurs, etc. over time; patient movement; patient changes in body mass index (BMI); gender; age; neck size; supine PW; relative proportion of different types of apneas (e.g., obstructive, central, mixed); and / or other factors.

[0240] With regard to the example framework of FIGS. 16A-16G and / or other examples of the present disclosure, it will be understood that in some example methods (and / or devices), various combinations of sensing and / or stimulating different target tissues (including unilateral and bilateral options, different sensing / stimulation settings, different stimulation therapy protocols, using different sensing and / or stimulation modalities, etc.) may be employed to implement a method of treating sleep disordered breathing in view of: an identified probability of one or more of the different types / levels of collapse (e.g., tonguebased, soft palate, lateral wall, concentric) according to the examples of FIGS. 1- 17; different types of apnea (e.g., obstructive, central, mixed); disease burden (e.g. AHI); body type; body mass index (BMI); gender; age; supine pharyngeal width (SPW); neck size; pharyngeal oropharyngeal pressure (POP); other attributes of patient information 500 in FIG. 3; other patient information within examples of the present disclosure.

[0241] FIG. 17 is a block diagram schematically representing an example arrangement 4200, deployable in an example method of (or as an example1618.322 11168 system for) to facilitate SDB patient care such as within (but not limited to) a patient management arrangement.

[0242] In some examples, the example arrangement 4200 may comprise a resource 4230, which comprises a computing resource (including stored programming) provided via a third party (third party service provider) to help provide and support the at least some aspects of the example methods throughout the present disclosure such as (but not limited to) a SDB patient care path / workflow such as (but not limited to) the example methods of FIGS. 2, 9B, and / or various example methods throughout the present disclosure. In some such examples, the resource 4230 may sometimes be referred to as a service provider resource. In some examples, resource 4230 may comprise at least a portion or, and / or an example implementation of, the control portion 4500 (FIG. 18A) and user interface 4540 (FIG. 18C). In some examples, the third party providing resource 4230 may comprise a device manufacturer, device supplier, or third party contracted by a device manufacturer. The resource 4230 may be hosted via the internet, World Wide Web, and / or other network communication link. Via a wireless communication protocol as represented by directional arrows 4227 and / or wired communication protocol, the resource 4230 may communicate with, and / or support communication among various other parties 4240 such as (but not limited to) a medical clinic 4250A, a sleep center 4250B, clinician device(s) 4250C (e.g. programmer 4262 and / or portal 4260), therapeutic medical device (TMD) 4285, patient remote 4289, and / or patient devices 4210.

[0243] Each of the entities associated with the respective devices 4250A, 4250B, 4250C may work together in at least some aspects to help coordinate care for the patient(s). Each entity may provide a particular form of expertise in patient care, such examples in which one entity may comprise a medical clinic (4250A), while another entity (associated with a different device 4250B) may comprise a sleep center, and some other entities comprise a clinician (4250C). Yet other entities may comprise providers which support patient care in some manner. There may be greater or fewer than the various devices (e.g. care entities) 4250A, 4250B, 4250C shown in FIG. 17 which form at least part of a care team.1618.322 11169

[0244] Each of the devices 4250A, 4250C, 4250C comprise a computing resource, such as workstation or other computing device which may be stationary or mobile. In some examples, at least one of the respective devices comprises a user interface (GUI) to facilitate display a workflow, inputs, outputs, etc. relative to perform a method which implements a SDB patient care path such as (but not limited to) the examples of at least FIGS. 2, 9C as supported via the examples of FIGS. 1-19. The user interface may comprise one example implementation of, and / or comprise at least some of the features and attributes of, the user interface 4540 described later in association with at least FIGS. 18C. Accordingly, it will be further understood that the various devices 4250A, 4250B, 4250C may comprise a control portion, or comprise an example implementation of one part of a control portion, such as control portion 4500 (FIGS. 18A-19).

[0245] As part of this arrangement, the devices may display at least some patient information (500 in FIG. 3; 800 in FIG. 4) as part of preparing and / or executing such methods, as well as managing a patient along a SDB patient care path which includes identifying a probability of a collapse pattern (e.g. expressed via a score), evaluating a score, recommending or implementing further actions, etc. In some examples, these devices may participate in training a data model, updating a data model, selecting and / or entering inputs to a data model, selecting and / or reporting outputs of a data model, modifying inputs / outputs of a data model, etc. At least some of this actions may comprise receiving patient information (500 in FIG. 3; 800 in FIG. 4), inputting patient information, and / or sending patient information to the resource 4230 and / or to other devices in arrangement 4200 via resource 4230 or directly.

[0246] In some examples, among other functions and features, the portal 4260 of devices 4250A, 4250B, 4250C, etc. may comprise a patient management app 4270 by which the particular care provider (e.g. medical clinic, sleep center, clinician, etc.) may manage patient care among a group of patients, for an individual patient, etc. In some such examples, this patient care may comprise implementing the previously described example methods of FIGS. 2, 9C as supported by the examples of FIGS. 1-19. In some examples, when deployed in relation to the respective devices 4250A, 4250B, 4250C, the portal 4260 may1618.322 11170 sometimes be referred to as a medical clinic portal, sleep center portal, clinician portal, respectively.

[0247] The patient management app 4270 also may enable communicating with other entities (e.g. 4250A, 4250B, 4250C) regarding patient care of the patients associated with devices 4210. In some examples, the clinician devices 4250A, 4250B, etc. may communicate with each other via at least resource 4230 (e.g. network communication link, internet, web, etc.) as represented via indicators 4227.

[0248] In some examples a clinician device 4250C comprises a programmer 4262 for programming, etc. the TMD 4285. For example, the programmer 4262 may periodically communicate with the TMD 4285 (or external medical device) wirelessly (e.g. inductive telemetry) to initially configure and / or later modify the configured stimulation therapy settings, sensing settings, therapy management settings, etc. of the TMD 4285, as well as send / receive patient information to and from the TMD 4285. However, in some examples, the programmer 4262 which may perform tasks or operations (relating to SDB patient care, maintenance) etc. in addition to programming stimulation-related aspects and / or sensing-related aspects of the TMD 4285.

[0249] As shown in FIG. 17, example arrangement 4200 also may comprise an array 4202 of computing devices 4210, each of which host a patient app 4212 relating to patient care. The devices 4210 may sometimes be referred to as patient devices, patient computing devices, and the like. The patient app 4212 may provide patient education and / or enable communication with resource 4230, and / or communicate (via resource 4230 and / or directly) with any one of the devices 4250A, 4250B, 4250C. Via such communication, the patient app 4212 may receive, send, display, store, and / or communicate patient information, usage information, etc. In some examples, the patient app 4212 is one vehicle via which a patient may participate in a SDB patient care path according to the various examples of the present disclosure. In some such examples, at least some of the patient devices 4210 may comprise a mobile computing device, such as a mobile phone, tablet, smartwatch, etc. which has a user interface 4540 (FIG. 18C) to provide for operation of, and display of, the patient app 4212.1618.322 11171

[0250] As shown in FIG. 17, in some examples the example arrangement 4200 may comprise a therapeutic medical device (TMD) 4285, which include implantable components and / or external components. In some examples, the TMD 4285 may be adapted for treating sleep disordered breathing (SDB) and / or other patient conditions (e.g. cardiac, etc.). A patient remote control 4289 may communicate with the TMD 4285 via a wireless communication protocol either directly or indirectly via an intermediary communication element (e.g. antenna, other). In some examples, such wireless communication may take the form of inductive telemetry. In some examples, the TMD 4285 may comprise an implantable pulse generator (IPG) for generating stimulation therapy signals to be delivered to the patient via a stimulation element (e.g. electrode) within the patient.

[0251] In general terms, the patient remote control 4289 enables a patient to have limited control over their stimulation therapy, such as turning the stimulation therapy on / off, pause, and / or increasing or decreasing the amplitude of stimulation within a lower and upper limit set by a clinician (and / or device manufacturer, supplier, etc.). In some examples, the patient remote control 4289 also tracks patient usage of these controls to enable a clinician, the patient, and others to learn about the patient’s usage, therapy effectiveness, patient adherence, etc. In some examples, the patient remote control also may receive some information from the TMD 4285 regarding stimulation metrics, sensing metrics, therapy metrics, usage metrics, etc.

[0252] In some examples, the patient remote control 4289 is in communication with the patient app 4212 such that patient app 4212 on device 4210 may receive the patient usage information from the patient remote control 4289, as well as whatever therapy, sensing, etc. information was communicated from the TMD 4285 to the patient remote control 4289. In some examples, the communication between the patient remote control 4289 and the patient app 4212 (on patient device 4210) may occur wirelessly 4229 via a number of wireless communication protocols such as, but not limited to, a Bluetooth® wireless communication protocol. In some examples, the communication between patient remote control1618.322 111724289 and patient app 4212 (on patient device 4210) may occur via a wired connection.

[0253] As noted here and elsewhere, the patient app 4212 may communicate this information (received from the patient remote control 4289) to one or more of the devices 4250A, 4250B, 4250C via resource 4230 to facilitate patient management according to examples of the present disclosure. In some examples, the patient app 4212 also may obtain some patient information through the patient’s use of the app 4212 which also may be communicated to the devices 4250A, 4250B, 4250C separately from, or integrated with, the patient usage information and therapy information from the patient remote control 4289 and / or TMD 4285.

[0254] As further shown in FIG. 17, in some examples a TMD 4285 may comprise a stimulation component 4286 and / or sensing component 4287. In some examples, the stimulation component 4286 comprises a stimulation engine to generate a stimulation signal to be applied to a tissue (e.g., nerve, muscle, etc.). In the examples in which the TMD 4285 comprises an implantable pulse generator (IPG), the tissue to be stimulated may comprise tissue to maintain or restore upper airway patency, such as but not limited to a hypoglossal nerve, infrahyoid-muscle (IHM)-innervating nerve, and / or other nerves / muscles which may directly or indirectly promote upper airway patency. In some such examples, the stimulation component 4286 also may comprise circuitry for generating and delivering the stimulation signal. In some examples, the stimulation component 4286 of the TMD 4285 also may comprise a stimulation element, such as an electrode through which the stimulation signal may be applied to the target tissue.

[0255] In some examples, the sensing component 4287 comprises a sensing engine to receive a sensing signal obtained relative to a tissue (e.g., muscle, organ, etc.). In the examples in which the TMD 4285 comprises an IPG for treating sleep disordered breathing (SDB), the tissue to be sensed may be related to respiration, oxygenation, cardiac functions, upper airway patency, and the like. In some such examples, the sensing component 4287 also may comprise circuitry for receiving and processing the sensing signal. In some examples, the sensing component 4287 of the TMD 4285 also may comprise a sensing element, such1618.322 11173 as an electrode or other element through which the sensing signal is obtained. In some examples, the sensing element may comprise an accelerometer for determining sleep information, respiratory information, posture information, physical control information. The accelerometer may be implantable, and in some examples, may be incorporated within a device including a stimulation generating element, such as an implantable pulse generator.

[0256] In some examples, the stimulation component 4286 and / or sensing component 4287 may be on-board the TMD 4285, which in some examples may comprise a microstimulator which houses stimulation circuitry, power element, and / or communication element.

[0257] In some examples, at least a portion of the stimulation component 4286 and / or sensing component 4287 may be separate from, and independent of, a housing of the TMD 4285 with one or both components 4286, 4287 being in wired or wireless communication with the TMD 4285.

[0258] FIG. 18A is a block diagram schematically representing an example control portion 4500. In some examples, control portion 4500 provides one example implementation of a control portion forming a part of, implementing, and / or generally managing example methods (including patient care paths / workflows), data models (e.g. training, operation), patient information, implantable medical devices (IMDs), sensing (e.g. portions, element(s)), stimulation (e.g. portion, elements), power / control elements (e.g. pulse generators), communication elements / pathways, devices, user interfaces, instructions, information, engines, elements, functions, and / or actions, etc., as described throughout examples of the present disclosure.

[0259] In some examples, control portion 4500 includes a controller 4502 and a memory 4510. In general terms, controller 4502 of control portion 4500 comprises at least one processor 4504 and associated memories. The controller 4502 is electrically couplable to, and in communication with, memory 4510 to generate control signals to direct operation of at least some of the example methods (including patient care paths / workflows), data models (e.g. training, operation), patient information, implantable medical devices (IMDs), sensing (e.g. portions, element(s)), stimulation (e.g. portion, elements), power / control elements (e.g.,1618.322 11174 pulse generator), communication elements / pathways, devices, user interfaces, instructions, information, engines, elements, functions, and / or actions, etc., as described throughout examples of the present disclosure. In some examples, these generated control signals include, but are not limited to, employing instructions 4511 and / or information 4512 stored in memory 4510 to implement the example methods and / or devices according to the various examples of the present disclosure which generally relate to SDB patient care. In some instances, the controller 4502 or control portion 4500 may sometimes be referred to as being programmed to perform the above-identified actions, functions, etc. such that the controller 4502, control portion 4500 and any associated processors may sometimes be referred to as being a special purpose computer, control portion, controller, or processor. In some examples, at least some of the stored instructions 4511 and / or information 4512 may form at least part of, and / or may be referred to as an engine.

[0260] In response to or based upon commands received via a user interface (e.g., user interface 4540 in FIG. 18C) and / or via machine readable instructions, controller 4502 generates control signals as described above in accordance with at least some of the examples of the present disclosure. In some examples, controller 4502 is embodied in a general purpose computing device while in some examples, controller 4502 is incorporated into or associated with at least some of the example methods (including patient care paths / workflows), data models (e.g. training, operation), patient information, implantable medical devices (IMDs), sensing (e.g. portions, element(s)), stimulation (e.g. portion, elements), power / control elements (e.g., pulse generator), communication elements / pathways, devices, user interfaces, instructions, information, engines, elements, functions, and / or actions, etc., as described throughout examples of the present disclosure.

[0261] For purposes of this application, in reference to the controller 4502, the term “processor” shall mean a presently developed or future developed processor (or processing resources) that executes machine readable instructions contained in a memory. In some examples, execution of the machine readable instructions, such as those provided via memory 4510 of control portion 4500 cause the1618.322 11175 processor to perform the various IMD operations and / or related operations, as generally described in (or consistent with) at least some of the various examples of the present disclosure. The machine readable instructions may be loaded in a random access memory (RAM) for execution by the processor from their stored location in a read only memory (ROM), a mass storage device, or some other persistent storage (e.g., non-transitory tangible medium or non-volatile tangible medium), as represented by memory 4510. In some examples, the machine readable instructions may comprise a sequence of instructions, a processorexecutable machine learning model, or the like. In some examples, memory 4510 comprises a computer readable tangible medium providing non-volatile storage of the machine readable instructions executable by a process of controller 4502. In some examples, the computer readable tangible medium may sometimes be referred to as, and / or comprise at least a portion of, a computer program product. In other examples, hard wired circuitry may be used in place of or in combination with machine readable instructions to implement the functions described. For example, controller 4502 may be embodied as part of at least one applicationspecific integrated circuit (ASIC), at least one field-programmable gate array (FPGA), and / or the like. In at least some examples, the controller 4502 is not limited to any specific combination of hardware circuitry and machine readable instructions, nor limited to any particular source for the machine readable instructions executed by the controller 4502.

[0262] In some examples, control portion 4500 may be entirely implemented within or by a stand-alone device.

[0263] In some examples, the control portion 4500 may be partially implemented in or via one of the example methods (including patient care paths / workflows), data models (e.g. training, operation), patient information, implantable medical devices (IMDs), sensing (e.g. portions, element(s)), stimulation (e.g. portion, elements), power / control elements (e.g., pulse generator), communication elements / pathways, devices, user interfaces, instructions, information, engines, elements, functions, and / or actions, etc. and partially implemented in a computing resource (e.g., at least one external resource) separate from, and independent of, the treatment devices (or portions thereof) but in communication with the1618.322 11176 treatment devices (or portions thereof). For instance, in some examples control portion 4500 may be implemented via a server accessible via the cloud and / or other network pathways. In some examples, the control portion 4500 may be distributed or apportioned among multiple devices or resources such as among a server, a treatment device (or portion thereof), and / or a user interface.

[0264] In some examples, control portion 4500 includes, and / or is in communication with, a user interface 4540 as shown in FIG. 18C.

[0265] Figure 18B is a diagram schematically illustrating at least some example arrangements of a control portion 4520 by which the control portion 4500 (FIG. 18A) can be implemented, according to one example of the present disclosure. In some examples, control portion 4520 is entirely implemented within or by an IMD 4525, which has at least some of substantially the same features and attributes as an IMD (e.g., 260, 262) as previously described throughout the present disclosure. In some examples, control portion 4520 is entirely implemented within or by a remote control 4530 (e.g., a programmer) external to the patient’s body, such as a patient control 4532 and / or a physician control 4534. In some examples, the control portion 4500 is partially implemented in the IMD 4525 and partially implemented in the remote control 4530 (at least one of patient control 4532 and physician control 4534). In some examples, the control portion 4520 in FIG. 18B is at least partially implemented via a clinician portal (e.g., 4662 in FIG. 19), which may or may not be in complementary relation with elements 4525 and 4530.

[0266] FIG. 18C is a block diagram schematically representing user interface 4540, according to one example of the present disclosure. In some examples, user interface 4540 forms part of and / or is accessible via a device external to the patient and by which the therapy system may be at least partially controlled and / or monitored. The external device which hosts user interface 4540 may be a patient remote (e.g., 4532 in FIG. 18B), a physician remote (e.g., 4534 in FIG. 18B) and / or a clinician portal (e.g., 4662 in FIG. 19). In some examples, user interface 4540 comprises a user interface or other display that provides for the simultaneous display, activation, and / or operation of at least some of the example methods (including patient care paths / workflows), data models (e.g. training,1618.322 11177 operation), patient information, implantable medical devices (IMDs), sensing (e.g. portions, element(s)), stimulation (e.g. portion, elements), power / control elements (e.g., pulse generator), communication elements / pathways, devices, user interfaces, instructions, information, engines, elements, functions, and / or actions, etc. In some examples, at least some portions or aspects of the user interface 4540 are provided via a graphical user interface (GUI) and may comprise a display 4544 and input 4542.

[0267] FIG. 19 is a diagram 4600 which schematically represents some example implementations by which a medical device 4610 may communicate wirelessly with devices outside the patient. In some examples, the medical device 4610 may comprise at least some implantable components and / or at least some external components.

[0268] As shown in FIG. 19, in some examples, the medical device 4610 may communicate with at least one of patient app 4630 on a mobile device 4620, a patient remote control 4640, a clinician programmer 4650, and a patient management tool 4660. The patient management tool 4660 may be implemented via a cloud-based portal 4662, the patient app 4630, and / or the patient remote control 4640, each of which may comprise a user interface (e.g., having a display, input) such as user interface 4540 in FIG. 18C.

[0269] Among other types of data, these communication arrangements enable the medical device 461 O to communicate, display, adjust, manage, etc. any of the example methods (e.g. SDB patient care paths / workflows, decisions, actions, etc.), data models (e.g. training, operation, etc.) commands, etc. in association with any of the various examples of the present disclosure.

[0270] Although specific examples have been illustrated and described herein, a variety of alternate and / or equivalent implementations may be substituted for the specific examples shown and described without departing from the scope of the present disclosure. This application is intended to cover any adaptations or variations of the specific examples discussed herein.

Claims

1. 1618.322 11178CLAIMS1 . A method of sleep disordered breathing care comprising: receiving, at least one of a clinician portal and a service provider resource, specific patient information for a first patient; determining via the service provider resource which is in communication with the clinician portal, based on the specific patient information, a numerical score representing a probability of a complete concentric collapse (CCC) pattern for an upper airway and / or a therapy outcome; comparing, via the service provider resource, the numerical score to a selectable first criterion; and upon the service provider resource determining that the numerical score fails to meet the selectable first criterion, the service provider resource communicating to the clinician portal for display at the clinician portal: a first recommendation to forego a drug-induced sleep endoscopy procedure for the first patient; and a second recommendation to implement electrical stimulation of an upper airway patency-related tissue for a non-complete concentric collapse pattern of the upper airway.

2. The method comprising: upon the service provider resource determining that the numerical score meets the selectable first criterion, the service provider resource communicating to the clinician portal for display at the clinician portal a third recommendation to perform a drug- induced sleep endoscopy procedure on the first patient; the service provider resource receiving, from the clinician portal, a result of the DISE procedure confirming the first patient having a CCC pattern; and the service provider resource communicating to the clinician portal for display at the clinician portal a fourth recommendation to:1618.322 11179 forego implementing electrical stimulation of the upper airway patency-related tissue including the hypoglossal nerve as the sole target tissue; or implement electrical stimulation of the upper airway patency- related tissue, which includes both a hypoglossal nerve and a non- hypoglossal nerve.

3. The method of claim 1 or 2, wherein the non-hypoglossal nerve comprises an infrahyoid muscle-related nerve and / or an infrahyoid muscle.

4. The method of claim 1 , 2, or 3, wherein the infrahyoid muscle comprises a sternothyroid muscle and the infrahyoid muscle-related nerve innervates the sternothyroid muscle.

5. The method of claim 1 , 2, 3, or 4, wherein the determining the numerical score comprises: submitting the patient specific information as an input to a constructed data model to determine, as an output, the numerical score.

6. The method of claim 1 , 2, 3, 4, or 5, wherein the specific patient information comprises at least one of gender and BMI.

7. The method of claim 1 , 2, 3, 4, 5, or 6, wherein the specific patient information comprises: a sensed respiratory waveform of the patient including inspiratory phase information including at least one sensed inspiratory phase in which a negative effort dependence parameter meets a selectable first criterion, a negative skewness meets a selectable second criterion, and / or a substantially reduced amplitude meets a selectable third criterion.

8. The method of claim 1 , 2, 3, 4, 5, 6, or 7, wherein the specific patient information comprises behavioral information comprising at least one of:1618.322 11180 an arousal threshold parameter; a snoring profile parameter; a consumption habits parameter; a loop gain parameter; a neuromuscular collapse response parameter; or a cortical sensorimotor-muscle coherence parameter.

9. The method of claim 1 , 2, 3, 4, 5, 6, 7, or 8, wherein the specific patient information comprises sleep study information.

10. The method of claim 1 , 2, 3, 4, 5, 6, 7, 8, or 9 wherein the specific patient information comprises electronic health record information, patient survey information, demographic information, and / or behavioral information.

11. A device to implement the method according to claim 1 , 2, 3, 4, 5, 6, 7, 8, 9, or 10, wherein the device comprises: memory configured to store machine readable instructions; and at least one processor coupled with the memory and configured to execute the machine readable instructions to: receive, at at least one of the clinician portal and the service provider resource, specific patient information for the first patient; determine via the service provider resource which is in communication with the clinician portal, based on the specific patient information, a numerical score representing the probability of a complete concentric collapse (CCC) pattern for an upper airway and / or the therapy outcome; compare, via the service provider resource, the numerical score to the selectable first criterion; and upon the service provider resource determining that the numerical score fails to meet the selectable first criterion, the service provider resource to communicate to the clinician portal for display at the clinician portal:1618.322 111 81 the first recommendation to forego a drug-induced sleep endoscopy procedure for the first patient; and the second recommendation to implement electrical stimulation of an upper airway patency-related tissue for a noncomplete concentric collapse pattern of the upper airway.

Images