Non-invasive topographic localization of glymphatic flow

Abstract

A computer-implemented method and system includes accessing neurophysiological and neurovascular data recorded during sleep. A device worn on the subject's head during sleep can collect the neurovascular data by measuring transcranial impedance. The device can be in the form of a single band, multiple bands, or a head cap worn on the subject's head. Electrodes positioned along the inner surface of the wearable device contact the subject's head and measure the transcranial impedance. Transcranial impedance measurements are collected from multiple locations on the subject's head and across multiple frequencies to obtain a more complete assessment of glymphatic flow.

Description

[Technical Field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of U.S. Provisional Patent Application No. 63 / 380,726, filed October 24, 2022. This application is also related to U.S. Patent Application No. 17 / 937,952, filed October 4, 2022, which is a continuation of and claims priority to U.S. Patent Application No. 17 / 713,329, filed April 5, 2022, which claims the benefit of U.S. Provisional Patent Application No. 63 / 244,080, filed September 14, 2021. The entire contents of the above applications are incorporated herein by reference.

[0002]

[0002] The present invention relates generally to the collection of neurophysiological and neurovascular data from wearable devices for use in the identification, prediction and treatment of neurodegeneration. [Background technology]

[0003]

[0003] Interest in cerebral fluid transport systems has rapidly increased in recent years with the discovery of the glymphatic system and its role in the clearance of brain proteins involved in neurodegeneration, such as amyloid beta. Disruption of glymphatic flow accelerates protein accumulation and cognitive decline in animal models of Alzheimer's disease, traumatic brain injury, and Parkinson's disease. Glymphatic flow is primarily active during sleep and is driven by cerebrovascular arterial pulsation; therefore, sleep, cerebrovascular integrity, and neurovascular connectivity are necessary for the clearance of waste products that accumulate in the awake brain. Reduced glymphatic flow results in the accumulation of proteins in the brain (called proteinopathy), which leads to neurodegeneration and can be detected by neuroimaging or molecular analysis of cerebrospinal fluid or plasma. Therefore, noninvasive techniques for monitoring and analyzing information related to glymphatic flux and its protein clearance, focusing on monitoring the mechanisms of glymphatic flux that result in protein accumulation rather than simply reporting protein accumulation in the brain as done by neuroimaging or molecular analysis, may be useful in the diagnosis and treatment of neurodegeneration. Summary of the Invention [Problem to be solved by the invention]

[0004]

[0004] The present disclosure is generally directed to the collection of neurophysiological and neurovascular data for use in the identification, prediction and treatment of neurodegeneration. [Means for solving the problem]

[0005] In one exemplary embodiment, the present disclosure is directed to a method implemented by one or more computer processors, the method including: (a) accessing, by the one or more computer processors, neurophysiological and neurovascular data recorded during sleep; (b) performing, by the one or more computer processors, function mapping from the neurophysiological and neurovascular data to targets that are markers of glymphatic flow; and (c) outputting, by the one or more computer processors, a target prediction model based on the function mapping.

[0006]

[0005] In another exemplary embodiment, the present disclosure is directed to a system including (a) one or more computer processors; (b) a neurophysiological data acquisition module configured to measure neurophysiological data; (c) a neurovascular data acquisition module configured to measure neurovascular data; and (d) a transmission module configured to transmit the electroencephalogram data and the neurovascular data to a second computing device.

[0007]

[0006] The above embodiments are non-limiting examples, and other aspects and embodiments are also described herein. The above Summary is provided to introduce various concepts that are further described below in the Detailed Description. This Summary is not intended to identify required or essential features of the claimed subject matter, nor is it intended to limit the scope of the claimed subject matter.

[0008]

[0007] Specific features, aspects, and advantages of the present invention will be better understood with reference to the following description and accompanying drawings. [Brief explanation of the drawings]

[0009] [Figure 1A] FIG. 1A illustrates one embodiment of a system for non-invasive measurement of cerebral proteinopathy and neurodegeneration, corresponding to molecular analysis or neuroimaging, constructed in accordance with the present invention. [Figure 1B] FIG. 1B illustrates one embodiment of a system for non-invasive measurement of cerebral proteinopathy and neurodegeneration, corresponding to molecular analysis or neuroimaging, constructed in accordance with the present invention. [Figure 2A]

[0009] FIG. 1 shows an embodiment of an in-ear sensor of a system for non-invasive measurement of cerebral proteinopathy and neurodegeneration, corresponding to molecular analysis or neuroimaging, constructed in accordance with the present invention. [Figure 2B] FIG. 1 shows an embodiment of an in-ear sensor of a system for non-invasive measurement of cerebral proteinopathy and neurodegeneration corresponding to molecular analysis or neuroimaging configured according to the present invention. [Figure 2C]

[0010] 1A-1C illustrate an assembly process for an ear device including a flex circuit and a man-made soft silicone component according to the present invention. [Figure 2D] 1A-1C illustrate an assembly process for an ear device including a flex circuit and a man-made soft silicone component according to the present invention. [Figure 2E] 1A-1C illustrate an assembly process for an ear device including a flex circuit and a man-made soft silicone component according to the present invention. [Figure 2F] 1A-1C illustrate an assembly process for an ear device including a flex circuit and a man-made soft silicone component according to the present invention. [Figure 2G] 1A-1C illustrate an assembly process for an ear device including a flex circuit and a man-made soft silicone component according to the present invention. [Figure 3]

[0011] FIG. 1 illustrates an exemplary computing environment for a system for non-invasive measurement of glymphatic flow, cerebral proteinopathy, and neurodegeneration, corresponding to molecular analysis or neuroimaging, configured in accordance with the present invention. [Figure 4]

[0012] FIG. 2 illustrates a functional description of a system according to an embodiment of the present invention. [Figure 5A]

[0013] FIG. 5A illustrates an exemplary method related to the collection of neurophysiological and neurovascular data from a wearable device for use in the identification, prediction, and treatment of neurodegeneration, in accordance with one embodiment of the present invention. [Figure 5B] FIG. 5B illustrates an exemplary method related to the collection of neurophysiological and neurovascular data from a wearable device for use in identifying, predicting, and treating neurodegeneration, in accordance with one embodiment of the present invention. [Figure 6A]

[0014] FIG. 2 illustrates a neurophysiological data acquisition module and exemplary associated data, according to one embodiment of the present invention. [Figure 6B] FIG. 2 illustrates a neurophysiological data acquisition module and exemplary associated data, according to one embodiment of the present invention. [Figure 6C] FIG. 2 illustrates a neurophysiological data acquisition module and exemplary associated data, according to one embodiment of the present invention. [Figure 6D] FIG. 2 illustrates a neurophysiological data acquisition module and exemplary associated data, according to one embodiment of the present invention. [Figure 6E] FIG. 2 illustrates a neurophysiological data acquisition module and exemplary associated data, according to one embodiment of the present invention. [Figure 6F] FIG. 2 illustrates a neurophysiological data acquisition module and exemplary associated data, according to one embodiment of the present invention. [Figure 7A]

[0015] FIG. 2 illustrates a neurovascular data acquisition module according to one embodiment of the present invention. [Figure 7B-1] FIG. 10 illustrates exemplary associated data for a neurovascular data acquisition module according to one embodiment of the present invention. [Figure 7B-2] Same as above [Figure 7C] FIG. 10 illustrates exemplary associated data for a neurovascular data acquisition module according to one embodiment of the present invention. [Figure 7D] FIG. 10 illustrates exemplary associated data for a neurovascular data acquisition module according to one embodiment of the present invention. [Figure 7E] FIG. 10 illustrates exemplary associated data for a neurovascular data acquisition module according to one embodiment of the present invention. [Figure 8A]

[0016] FIG. 1 illustrates a data acquisition module for markers of glymphatic flow, molecular analysis markers of neurodegeneration or neuroimaging markers, according to one embodiment of the present invention. [Figure 8B] 1A-1C show relevant images of a data acquisition module for markers of glymphatic flow, molecular analysis markers of neurodegeneration or neuroimaging markers, according to one embodiment of the present invention. [Figure 9]

[0017] FIG. 1 illustrates a learning module for functional mapping from non-invasive neurophysiological and neurovascular input data to a target output measure of glymphatic flow according to one embodiment of the present invention, or from input data for markers of glymphatic flow to a target output measure of brain proteinopathy or neurodegeneration from molecular analysis or neuroimaging according to another embodiment of the present invention. [Figure 10]

[0018] FIG. 1 illustrates a module for predicting target markers of glymphatic flow from non-invasive neurophysiological and neurovascular input data according to one embodiment of the present invention, or a module for predicting target molecular analysis or neuroimaging of cerebral proteinopathy and neurodegeneration using input data that are markers of glymphatic flow according to another embodiment of the present invention. [Figure 11A]

[0019] FIG. 1 illustrates a target engagement module, according to one embodiment of the present invention. [Figure 11B-1] FIG. 10 illustrates relevant data for a target engagement module, according to one embodiment of the present invention. [Figure 11B-2] Same as above [Figure 11B-3] Same as above [Figure 11C] FIG. 10 illustrates relevant data for a target engagement module, according to one embodiment of the present invention. [Figure 12A]

[0020] FIG. 1 illustrates one embodiment of a system for non-invasive measurement of glymphatic flow and neurodegeneration in an overhead configuration that positions data acquisition sensors used to acquire transcranial impedance over the left and right temporal and parietal lobes, constructed in accordance with the present disclosure. [Figure 12B] FIG. 1 illustrates one embodiment of a system for non-invasive measurement of glymphatic flow and neurodegeneration in an overhead configuration that positions data acquisition sensors used to acquire transcranial impedance over the left and right temporal and parietal lobes, constructed in accordance with the present disclosure. [Figure 12C] FIG. 1 illustrates one embodiment of a system for non-invasive measurement of glymphatic flow and neurodegeneration in an overhead configuration that positions data acquisition sensors used to acquire transcranial impedance over the left and right temporal and parietal lobes, constructed in accordance with the present disclosure. [Figure 12D]

[0021] FIG. 10 illustrates another embodiment of a system for non-invasive measurement of glymphatic flow and neurodegeneration in an overhead configuration, configured in accordance with the present disclosure, positioning data acquisition sensors used to acquire transcranial impedance in the anterior and posterior bands over the left and right temporal, parietal, and frontal lobes. [Figure 12E] FIG. 10 illustrates another embodiment of a system for non-invasive measurement of glymphatic flow and neurodegeneration in an overhead configuration, configured in accordance with the present disclosure, positioning data acquisition sensors used to acquire transcranial impedance in the anterior and posterior bands over the left and right temporal, parietal, and frontal lobes. [Figure 13]

[0022] FIG. 1 illustrates one embodiment of a system for non-invasive measurement of glymphatic flow and neurodegeneration in an overhead configuration in which data acquisition sensors used to acquire transcranial impedance are positioned in a head cap extending over the left and right temporal, parietal, and frontal lobes, constructed in accordance with the present disclosure. [Figure 14A]

[0023] FIG. 1 illustrates one embodiment of a system for non-invasive measurement of glymphatic flow and neurodegeneration in an overhead configuration, configured in accordance with the present disclosure, positioning data acquisition sensors used to acquire transcranial impedance at multiple left and right locations around the brain. [Figure 14B] FIG. 1 illustrates one embodiment of a system for non-invasive measurement of glymphatic flow and neurodegeneration in an overhead configuration, configured in accordance with the present disclosure, positioning data acquisition sensors used to acquire transcranial impedance at multiple left and right locations around the brain. [Figure 15]

[0024] FIG. 1 illustrates components of a controller for a head cap embodiment of a wearable device according to the present disclosure. [Figure 16]

[0025] 10A-10C illustrate a method of using a head cap embodiment of a wearable device according to the present disclosure. [Figure 17] 10A-10C illustrate a method of using a head cap embodiment of a wearable device according to the present disclosure. [Figure 18] 10A-10C illustrate a method of using a head cap embodiment of a wearable device according to the present disclosure. [Figure 19] 10A-10C illustrate a method of using a head cap embodiment of a wearable device according to the present disclosure. [Figure 20]

[0026] FIG. 10 shows exemplary data from a head cap embodiment of a wearable device according to the present disclosure in a subject sleeping with the device on for 100 minutes. DETAILED DESCRIPTION OF THE INVENTION

[0010]

[0027] With reference to glymphatic flow, which is involved in the clearance of the aforementioned proteins, the cerebral perivascular space (PVS) forms the core of the glymphatic fluid transport system. Fluorescent dye-labeled particles, which enable visualization of the perivascular space, reveal that this system comprises the periarteriolar, pericapillary, and perivenular spaces, all interconnected as a single network. The glymphatic flow system drains brain interstitial fluid (ISF) waste products, such as proteinosis proteins, through the periarteriolar and perivenular spaces into the meningeal lymph. The PVS is a crucial site affected by disease processes such as cerebral amyloid angiopathy, which is characterized by perivascular amyloid beta deposits, and is present in over 90% of brains with Alzheimer's disease. PVS accumulation of p-tau has also been observed in Alzheimer's disease and has been demonstrated to be present in arteries, arterioles, and veins within the brain. In the glymphatic flow system, cerebrospinal fluid (CSF) enters the periarteriolar space and flows in the direction of blood flow, propelled by arterial wall pulsation. CSF mixes with ISF, facilitated by aquaporin 4 (AQP4) water channels present in the vascular astrocyte endfeet that form the outer walls of perivascular spaces. During sleep, arterial pulsations drive bulk CSF flow toward the brain. This process is known to be highly dependent on wakefulness and AQP4 expression. Modern neuroimaging techniques now allow visualization of CSF flow and its protein clearance in the intact CNS through the constant pulsatile movement of CSF through the ventricular system that occurs with each cardiac contraction. Respiration also contributes to this rhythm, adding low-frequency pulsations. These sharp pulsations are superimposed on a circadian or diurnal rhythm, with maximum CSF production occurring during the night and minimum CSF production during sleep in the afternoon. MRI analysis in animal models using CSF tracers reveals enhanced glymphatic flow from ISF to subarachnoid CSF during NREM sleep.

[0011]

[0028] Sleep has a profound effect on most aspects of brain physiology, including glymphatic flow and cerebral blood volume. In animal models, the magnitude of cerebral blood volume during non-rapid eye movement (NREM) sleep doubles compared to wakefulness, and total blood volume in the brain changes substantially during the transition from wakefulness to sleep and back to wakefulness, indicating that cerebral blood volume dynamics are coupled to wakefulness. These changes in blood volume during sleep are linked to fluctuations in neural activity, which drive variations in arteriolar diameter. The strength of this neurovascular coupling changes with wakefulness and is significantly increased during sleep compared to wakefulness, explaining the increased cerebral blood volume during sleep, when neural activity is lower than during wakefulness.

[0012]

[0029] Amyloid-β and tau concentrations in the interstitial space and CSF follow a diurnal pattern, with both protein concentrations peaking during wakefulness and nadir during sleep. A single night of sleep deprivation produces a significant increase in soluble amyloid-β in brain ISF, as demonstrated by positron emission tomography (PET) using an amyloid-binding radioactive tracer. The net concentrations of amyloid-β, tau, and other proteins in CSF and ISF reflect the combined effects of diurnal and state-dependent variations in CSF production rate, ISF volume, ISF turnover, and glymphatic flow. Sleep slow-wave delta waves are associated with glymphatic flow and slow delta wave amplitude, which are involved in effective waste clearance during sleep. Sleep deprivation results in stalled glymphatic flow and the accumulation of waste solutes, including amyloid-β solutes and other proteins, in the brain.

[0013]

[0030] During sleep, glymphatic flow increases through the action of astrocytes using AQP4 channels, which are involved in regulating phases between sleep and wakefulness. Using the membrane channel protein AQP4 and water flux through AQP4, astrocytes reduce cell volume in the brain parenchyma, increasing ISF space by 23% during sleep, thereby allowing water and solutes to diffuse into adjacent perivascular spaces. During slow-wave sleep, the increase in ISF space, opening of gap junctions in astrocytic endfeet, and increased cerebral hemodynamics create an environment for nutrient exchange and waste removal between the ISF and CSF.

[0014]

[0031] Damage to the molecular and cellular components of the glymphatic system, cerebral neurovascular integrity, and its neurovascular connections caused by aging, chronic comorbidities, neurodegenerative diseases, and physical brain injury is closely linked to the accumulation of tau and amyloid-β in the extracellular fluid. The spread of these condensed proteins in the brain parenchyma leads to clinical deterioration and cognitive decline. Measuring the functional integrity of this system and the pathological accumulation of proteins in brain ISF requires invasive procedures, such as PET scans using radioactive tracers that bind to the protein of interest or lumbar punctures for CSF extraction and analysis of protein concentrations. The cost and complexity of these procedures limit their use for screening for cerebral proteinopathies or monitoring their progression. Without access to these tests, physicians currently rely on clinical and cognitive assessments to screen, diagnose, and monitor neurodegenerative disorders such as Alzheimer's disease. Opportunities for early intervention during the 10- to 20-year preclinical period of these disorders are lost, as are the evaluation of promising new therapies that may slow or halt disease progression. Another drawback of existing invasive testing methods is that they must be performed in a hospital or clinic setting by a trained technician, which creates an additional patient burden that further reduces the effectiveness of these costly and time-consuming invasive methods for monitoring the progression of proteinopathies, adjusting treatment plans, and monitoring responses to new therapies.

[0015]

[0032] The glymphatic system of protein waste clearance operates during sleep and depends on features measurable by sleep electroencephalography (EEG) and neurovascular blood volume, pulsatility, and coupling. Short-term deprivation during sleep results in an increase in CSF and ISF waste proteins, while long-term deprivation leads to the formation of ISF protein aggregates accompanied by neurodegeneration. A wearable device capable of measuring relevant EEG features, cerebral blood volume changes, and arterial pulsatility during sleep and transmitting that data to methods and systems capable of predicting CSF and ISF waste protein levels and ISF protein aggregates, as can be measured using existing invasive clinical techniques, would provide significant benefits to medical screening and monitoring of neurodegenerative proteinopathies and help select appropriate interventions to slow or halt protein accumulation and neurodegeneration. As used herein and described in more detail below, a "wearable device" is a device worn by a subject and includes one or more sensors for collecting neurophysiological and / or neurovascular data from the subject, a processor, a memory, a storage device for storing the collected data, and at least one wireless transmission module that enables the wearable device to transmit the collected data to another computing device. The wireless communication capabilities of the wearable device can allow the wearable device to be self-contained so that it does not require external wiring to other equipment. The wearable device is configured to be worn by the subject during sleep.

[0016]

[0033] In accordance with embodiments of the present disclosure, various figures are shown in Figures 1A-11C, and like reference numerals are used consistently throughout to refer to like and corresponding parts of the illustrated embodiments in all of the various views and figures of the drawings. It should also be noted that the leading digits of the reference numerals in Figures 1 and 4 correspond to the figure number in which that item or part is described in more detail.

[0017]

[0034] The following detailed description includes many specific details for purposes of illustration. Those skilled in the art will appreciate that many variations and modifications of the following details are within the scope of the present disclosure. Thus, the following exemplary embodiments are described for the claimed invention without loss of generality and without imposing limitations.

[0018]

[0035] Exemplary embodiments of the present disclosure include systems and methods that enable noninvasive measurement of cerebral proteinopathy and neurodegeneration. Such measurements are typically performed in medical facilities using sophisticated equipment and involve molecular analysis of cerebrospinal fluid or neuroimaging, including assays of proteinopathy protein amyloid, tau, or alpha-synuclein levels in plasma or CSF, cerebral amyloid or tau burden on PET scans, brain atrophy on structural brain MRI, and repeated diurnal assessments for measurement of net changes in proteinopathy protein CSF assays or soluble proteinopathy protein burden from neuroimaging. Exemplary embodiments of the present disclosure teach novel systems and methods for acquiring sleep neurophysiological data synchronized with neurovascular data. Neurophysiological data includes electroencephalogram (EEG) data. Neurovascular data includes transcranial impedance plethysmography (IPG), carotid pulse transit time (PTT), heart rate variability (HRV), and resting heart rate (RHR). Exemplary embodiments may further include obtaining one of the molecular analysis or neuroimaging tests for neurodegeneration after sleep and learning a function mapping from the sleep neurophysiology and neurovascular data to molecular analysis or neuroimaging markers of neurodegeneration. Learning the function mapping includes using a loss function to determine relevant features in the sleep data, identifying an optimal set of weights that produces a local minimum of the loss function, and using the optimal weights to create a function mapping, also referred to as a cerebral proteinopathy and neurodegeneration prediction model. Exemplary embodiments may further include applying the learned function mapping to new sleep neurophysiology and neurovascular data to screen for or monitor the molecular analysis or neuroimaging markers of neurodegeneration. Exemplary embodiments may further include administering a therapeutic intervention targeting aspects of sleep neurophysiology or neurovascularity before or during sleep, measuring target involvement or effects on sleep neurophysiology or neurovascularity, and applying the learned function mapping to predict treatment effects on cerebral proteinopathy and neurodegeneration.

[0019]

[0036] Another embodiment of the present disclosure includes a system and method that allows a person to monitor the effect of cardiovascular intervention or activity on molecular or neuroimaging markers of neurodegeneration, which embodiment teaches a novel system and method for recording cardiovascular intervention or activity data before sleep, and further recording neurophysiological and neurovascular data during sleep and applying learned function mapping to predict the therapeutic effect of cardiovascular intervention or activity on cerebral proteinopathy and neurodegeneration.

[0020]

[0037] Another embodiment of the present disclosure includes a system and method that allows a person to monitor the effect of pharmaceutical or neuromodulatory intervention on molecular analysis or neuroimaging markers of neurodegeneration, which embodiment teaches novel systems and methods for recording pharmaceutical or neuromodulatory intervention before or during sleep, and further for recording neurophysiological and neurovascular data during sleep and applying learned function mapping to predict the therapeutic effect of pharmaceutical or neuromodulatory intervention on cerebral proteinopathy and neurodegeneration.

[0021]

[0038] Another embodiment of the present disclosure includes a system and method that allows a person to monitor the effect of dietary intervention on molecular or neuroimaging markers of neurodegeneration, which embodiment teaches a novel system and method for recording the dietary intervention before sleep and further recording neurophysiological and neurovascular data during sleep and applying learned function mapping to predict the therapeutic effect of the dietary intervention on cerebral proteinopathy and neurodegeneration.

[0022]

[0039] Another embodiment of the present disclosure may be a wearable system including a wearable computer having a sensor and a wireless communication interface with a mobile computer supporting a wireless network interface that communicates with a second computer including a network interface, each computer further including a processor, a memory unit operable to store a computer program, an input mechanism operable to input data into the computer system, an output mechanism for presenting information to a user, and a bus coupling the processor to the memory unit, the input mechanism, and the output mechanism, wherein the wearable system includes various executable program modules stored in the wearable system and operable when executed to perform functions. The wearable computer with sensors may include a neurophysiological data acquisition module with sensors capable of recording EEG and a neurovascular data acquisition module with sensors capable of recording neurovascular data, data stored in the wearable computer and, when executed, recording neurophysiological and neurovascular sensor-acquired data in the memory unit of the wearable computer. A transmitting module can include a wireless network interface (e.g., Bluetooth or WiFi wireless) and instructions stored on the wearable computer that, when executed, transmit the records stored in the memory unit to the mobile computer via the wireless network interface. A second transmitting module can also be stored on the wearable computer and can include a wired network interface and instructions that, when executed, transmit the records stored in the memory unit via a bus network interface to a local computer station, which transmits the data to a second computer via the wireless network interface.A learning module can be stored on the second computer and, when executed, learns a function mapping from the transmitted recordings to molecular analysis or neuroimaging markers of neurodegeneration stored on the second computer, uses a loss function to determine relevant features in the recordings, identifies an optimal set of weights that produces a local minimum of the loss function, and creates the function mapping using the optimal weights. A brain proteinopathy and neurodegeneration prediction module can also be stored on the second computer and, when executed, applies the learned function mapping, also referred to as a brain proteinopathy and neurodegeneration prediction model, to newly transmitted recordings of sleep neurophysiology and neurovascular sensor data from a wearable computer with sensors to calculate predicted values ​​of molecular analysis or neuroimaging markers of neurodegeneration from the newly transmitted recordings. Thus, the brain proteinopathy and neurodegeneration prediction model can determine predicted brain proteinopathy and neurodegeneration values ​​from new recordings of data without the need for invasive imaging or assay procedures as described above.

[0023]

[0040] Another embodiment of the present disclosure may be a wearable system including a wearable computer having a sensor and a wireless communication interface with a mobile computer supporting a wireless network interface for communicating with a second computer including a network interface, each computer further including a processor, a memory unit operable to store a computer program, an input mechanism operable to input data to the computer system, an output mechanism for presenting information to a user, and a bus coupling the processor to the memory unit, the input mechanism, and the output mechanism, and the wearable system includes various executable program modules stored in the wearable system and operable to perform functions when executed. In one embodiment, the wearable system may include a data acquisition module having a sensor capable of recording cardiovascular activity data, the data being stored in the wearable computer and, when executed, storing the acquired data in the computer's memory unit. In another embodiment, the wearable system may include a data entry module capable of recording dietary and nutritional data, the data being stored in the mobile computer and, when executed, recording the data in the computer's memory unit. In another embodiment, the wearable system can include a data entry module capable of recording drug and neurostimulation data, the data being stored in a mobile computer and, when executed, recording the data in a memory unit of the computer. In any of the above embodiments, the transmission module can include a communications interface and associated instructions stored on the wearable or mobile computer, the instructions, when executed, transmitting the records stored in the memory unit to the second computer via a wireless network interface.In any of the above embodiments, the second computer may store a module that, when executed, applies the learned function mapping to transmitted recordings of sleep neurophysiological and neurovascular sensor data from a wearable computer with sensors in order to calculate values ​​of molecular analysis or neuroimaging markers of neurodegeneration from the transmitted recordings.

[0024]

[0041] The details of the present invention and various embodiments can be better understood with reference to the figures of the drawings. Figures 1A and 1B illustrate one embodiment of a functional description of a system constructed in accordance with the present invention and are not intended to be limiting, as one skilled in the art will recognize, upon review of this application, that other configurations may be used without departing from the scope of the claimed invention. Referring to Figure 1, the system can include a wearable device that can be worn on a subject's head. In one example, the wearable device can be worn on the subject's ear. The wearable device can include the neurophysiological and neurovascular data acquisition sensor 200 described above for noninvasive measurements. As further shown in Figure 1, the wearable device can include a data storage, processing, and transmission module 300 that can communicate with a mobile computer and / or a local computer and / or a remote second computer to store acquired sensor data, convert the data into preliminary measurements of sensor impedance to determine whether the sensor is correctly positioned and the bladder is correctly pressurized, and transmit the data for further processing into measurements of brain proteinopathy and neurodegeneration corresponding to molecular analysis or neuroimaging markers. The noninvasive sensor 200 is connected to the module 300 via a form-fitting band 100 that houses electrical wiring conduits that electrically isolate the digital sensor signal from the analog sensor signal, air tubes that connect the air bladder in the sensor 200 to the piezoelectric pump in 300 and inflate the in-ear portion of the ear device 200 to ensure sufficient interfacial contact between the surface sensor and the ear and ear canal epithelium, and a nitinol band designed to ensure a secure fit of the wearable device. A diagram of the worn device is shown at 110. It should be understood that in alternative embodiments, the components of the wearable device can have other configurations and can be attached to the user by other mechanisms. For example, the air bladder can be replaced with another type of fluid-filled bladder.

[0025]

[0042] 2A-2B illustrate one embodiment of an ear sensor in a system constructed in accordance with the present invention and are not intended to be limiting, as those skilled in the art will recognize upon review of this application that other configurations may be used without departing from the scope of the claimed invention. Referring to FIG. 2(a), flex circuitry housed in ear device 200 is shown at 210. Component 212 illustrates the positioning of an inertial measurement unit (IMU) that measures milligravity, which can sense ballistic acceleration within the ear from blood being ejected into the aorta with each cardiac contraction. Component 214 illustrates the positioning of a photoplethysmography (PPG) sensor on the flex circuit to contact the ear canal wall. PPG can measure pulsatile changes in capillary blood volume in the ear with each heartbeat and arterial blood oxygen saturation by measuring the reflection of two wavelengths of infrared light through the ear canal skin. Flex circuit conductors 216 terminate on the surface of ear device 200, where they make contact with two conductive electrodes covering the ear device, which may be conductive ink, fabric, or other material. Referring to FIG. 2(b) showing diagram 220, one layer of the flex circuit is designed to flex naturally to accommodate differences in the anatomical structure of a user's ear, the second layer of the two-layer flex is a stiffening material applied to the underside of the flex circuit to keep the active and passive electronic components rigid so they do not become dislodged by bending, and the fourth layer has a ground plane between the analog and digital signals to reduce or eliminate electronic coupling that causes signal corruption.

[0026]

[0043] Referring to Figures 2C-2G, Figures 230, 240, 250, 260, and 270 illustrate the assembly process for an ear device including the flex circuit 210 and artificial soft silicone component shown in Figures 230 and 250. Figure 240 shows the flex circuit 242 attached to the silicone component 230. Figure 250 shows a pouch 252 in the second silicone component that is pressurized with air or fluid once inside the ear canal. The expansion forces the PPG sensor 264 into intimate contact with the ear canal skin for improved signal-to-noise measurements and further ensures good interfacial contact between the conductive electrodes 272 and 273, used to measure neurophysiological currents from the brain and transcranial bioimpedance, and the ear canal epithelium. Note the ergonomic bend in the second silicone component 252, which mimics the initial bend in the human ear canal. The positioning of the IMU 262 on the device is shown in Figure 2F.

[0027]

[0044] 3 illustrates one embodiment of such a wearable device, shown as a wearable computing device 305, including a data storage, processing, and transmission module 300 configured in accordance with the present invention; this is not intended to be limiting in scope, as those skilled in the art will recognize upon review of this application that other configurations may be used without departing from the scope of the claimed invention. Specifically, data storage, processing, and transmission module 300 includes one or more processors 308, memory 310, a storage device 314, and a transmission module 312. The components of data storage, processing, and transmission module 300 are responsible for storing ear device sensor data, processing that data to determine correct sensor placement and bladder pressurization into preliminary measurements of sensor impedance, and transmitting that data for further offline processing into molecular analysis or measurements of brain proteinopathy and neurodegeneration corresponding to neuroimaging markers. 3 also shows the aforementioned sensors 200, such as an IMU and / or PPG, that, when placed in the patient's ear, collect data and provide it for storage in a storage device 314, processing by a processor 308, and transmission by a transmission module 312. The input / output device 314 can be a user interface in the form of buttons or a display screen. In other embodiments of the wearable computing device, the input / output device may be omitted.

[0028]

[0045] The illustrative embodiment of FIG. 3 shows a wearable computing device 305 capable of wirelessly communicating with a mobile computing device 320, such as a smartphone or tablet, and a local computing device 340, such as a desktop computer, via a wireless link. It should be understood that having both a mobile computing device and a local computing device is not a requirement, and in alternative embodiments, only one of the mobile computing device 320 and the local computing device 340 may communicate with the wearable computing device 305. As shown in FIG. 3 , the mobile computing device 320 and the local computing device 340 include readily recognizable conventional components, including a processor 322 / 342, memory 324 / 344, a transmission module 326 / 346, an input / output interface 328 / 348, and a storage device 330 / 350. One or both of the mobile computing device 320 and the local computing device 340 can transmit data collected by the wearable computing device 305 to a second computing device 360. The second computing device is typically located remotely and accessed via a wide area network, such as the Internet. The second computing device may include computing components such as a processor 362, a memory 364, a transmission module 366, an input / output interface 368, and a storage device 370. The storage device 370 may store various program modules executable by the processor 362. Specifically, the storage device 370 may include a learning software module and a predictive software module. For example, the learning software module may include a machine learning or deep learning algorithm that determines a function mapping from collected neurophysiological and neurovascular data to target markers of glymphatic clearance or neurodegeneration.The prediction module can include a model based on function mapping that is used to predict glymphatic clearance or neurodegeneration based on newly collected sensor data from the patient by the wearable computing device.

[0029]

[0046] 4 illustrates one embodiment of a functional description of a system configured in accordance with the present invention and is not intended to be limiting in scope, as those skilled in the art will recognize upon review of this application that other configurations may be used without departing from the scope of the claimed invention. Referring to FIG. 4 , a user's sleep neurophysiological data is captured and recorded by a neurophysiological data acquisition module 600 via sensors positioned on the user's head, such as the aforementioned sensors positioned in the ear. A neurovascular data acquisition module 700 further captures the user's sleep neurovascular data via sensors positioned on the user's head. As previously discussed, the neurophysiological data acquisition module 600 and the neurovascular data acquisition module 700 may be implemented in a wearable device.

[0030]

[0047] In contrast, the Brain Proteinopathy and Neurodegeneration Target Data Acquisition Module 800 is an invasive module typically implemented using high-performance equipment in a medical facility and responsible for acquiring one or more molecular analyses or neuroimaging markers of neurodegeneration. The markers of interest can be CSF or plasma assays for one of several brain proteinopathy proteins, including amyloid beta 42, tau, alpha-synuclein, and neurofilament light, which are markers of neurodegenerative disorders; neuroimaging scans such as PET scans with radioactive tracers of the protein of interest; or MRI using contrast agents to measure brain atrophy. The CSF assay or PET scan can be performed after sleep to assess the CSF protein or protein levels accumulated in the brain, or both before and after sleep to assess net changes in protein levels that occurred during sleep. The marker can also be MRI using intrathecal contrast agents injected into the CSF before sleep and drained into the meningeal lymphatic and cavernous sinus systems by brain interstitial fluid transport during sleep, with uptake and clearance rates measurable using repeated MRI scans.

[0031]

[0048] The target marker data measured by module 800 is used by target learning module 900 to create target prediction module 1000, which maps new neurophysiological data from module 600 and new neurovascular data from module 700 to predictive molecular or neuroimaging markers of neurodegeneration for longitudinal monitoring, without the need to repeat the invasive measurements of module 800. Target learning module 900 and target prediction module 1000 can be implemented as software using machine learning or deep learning algorithms to provide output. Target learning module 900 and target prediction module 1000 can also be implemented as software running on one or more remote computers, such as the second computer described above. Target prediction module 1000 can also map sleep neurophysiological data from module 600 and sleep neurovascular data from module 700 taken from a new patient to molecular or neuroimaging markers of neurodegeneration for noninvasive screening that can be used in medical decision-making to assess the need for invasive measurements of module 800. The target engagement module 1100 measures the effect, or degree of target engagement, of a putative therapeutic intervention on sleep neurophysiological or neurovascular data and how that effect translates into changes in predictive molecular or neuroimaging markers of neurodegeneration. The intervention can be one of a cardiovascular intervention, a dietary intervention, a sleep intervention, a pharmacological intervention, or a neurostimulation intervention.

[0032]

[0049] 5A and 5B illustrate two method embodiments of the present invention and are not intended to be limiting in scope, as those skilled in the art will recognize upon reviewing this application that other methods may be used without departing from the scope of the claimed invention. Referring to FIG. 5A, a method for creating and applying a target prediction model is shown. Beginning at operation 505, a processor of a computing device accesses neurophysiological and neurovascular data recorded during sleep of one or more patients. The data can be recorded using the wearable device described above or using other devices. At operation 510, the processor can use a learning software module to determine a functional mapping between the recorded data accessed at operation 505 and target marker data. The learning software module can use a machine learning algorithm to determine the functional mapping. The target marker data can be data indicative of glymphatic flow in the brain. Alternatively, the target marker data can be one of molecular analysis markers of neurodegeneration or neuroimaging markers of neurodegeneration, such as those described in connection with module 800 of FIG. 4. After completing the functional mapping, at operation 515, the learning software module outputs the target prediction model. Because the target prediction module can be used in a variety of ways, the example method of FIG. 5A can end after operation 515. However, to illustrate one application of the target prediction module, the example method of FIG. 5A can continue with operations 520-530. Specifically, at operation 520, newly collected neurophysiological and neurovascular data from the patient using the wearable device can be input into the target prediction model to generate markers of glymphatic clearance or markers of neurodegeneration. At operation 525, the output markers can be used to determine and administer a therapeutic intervention for the patient. In certain embodiments, information regarding the therapeutic intervention can be input into the wearable device. Finally, after the therapeutic intervention is administered, the wearable device can be used again at operation 530 to collect new data from the patient to determine the effectiveness of the therapeutic intervention.

[0033]

[0050] 5B , an exemplary method for operating a wearable device according to an exemplary embodiment of the present disclosure is shown. Beginning at operation 550, neurophysiological and neurovascular data sensors of the wearable device may be activated to collect data while a patient is sleeping with the wearable device attached to the patient's ear. For example, a processor of the wearable device may execute a data acquisition module stored in a storage module of the wearable device, which controls the operation of the sensors. At operation 555, sensors of the wearable device, such as an IMU and / or PPG, collect neurophysiological and neurovascular data while the patient is sleeping. At operation 560, the wearable device may store the collected data in a storage device of the wearable device. In some cases, as shown at operation 565, the processor of the wearable device may perform certain initial signal processing on the data collected by the sensors and store such processed data in a storage device of the wearable device. Finally, at operation 570, a transmission module of the wearable device may manage transmission of the collected and / or processed data from the wearable device to one or both of a mobile computing device and a local computing device via a wired or wireless communication link. The mobile computing device and / or the local computing device may relay the data from the wearable device to a second computing device that may execute the learning software module and the predictive software module.

[0034]

[0051] 6A-6F illustrate one embodiment of a neurophysiological data acquisition module 600 of a system constructed in accordance with the present invention and are not intended to be limiting in scope, as those skilled in the art will recognize upon review of this application that other configurations may be used without departing from the scope of the claimed invention. Referring to FIG. 6A, raw EEG data is acquired at 610 from a sleeping user. The EEG data can be acquired using a conventional EEG head cap sensor with a bedside amplifier for recording the signal, or using a wearable device with integrated electronics for amplifying and recording the signal. Various derived values ​​can be analyzed from the raw EEG signal to generate a hypnogram 620 marking the four stages of sleep: N1, N2, N3, and rapid eye movement (REM). Additional sleep performance parameters 680 can be derived from the raw EEG signal, including total sleep time, sleep onset latency, wakefulness after sleep onset, sleep efficiency, and number of wakefulnesses. At 630, slow wave oscillations are detected in the raw EEG signal, and their density (oscillations per minute), amplitude, and duration are reported for 30- or 60-second epochs during sleep. At 640, slow wave activity, defined as the relative or absolute power in the EEG signal in the frequency band between 0.5 Hz and 4.5 Hz, is calculated and reported for each 30- or 60-second epoch. Options for calculating the power spectrum include fast Fourier transform, Welch's periodogram, or multitaper method. At 650, another sleep EEG element, sleep spindles, is detected in one or more EEG derivatives. Sleep spindles are of two types: fast waves and slow waves, depending on spindle frequency. Spindle density, frequency, duration, and amplitude for each type are reported for each 30- or 60-second epoch. Both slow wave oscillations or activity 630 and 640 and sleep spindles 650 are sleep microstructural markers of glymphatic flow. Finally, at 670 the total power spectral density in the EEG signal is calculated and reported for 30 or 60 second epochs.

[0035]

[0052] One or more of the EEG data elements shown in Figure 6A can be used in connection with a target prediction model. For example, the EEG data elements shown in Figure 6A can be used to train the target prediction model. As another example, if the EEG data elements are newly collected data after the target prediction model has been trained, the EEG data elements can be input into the target prediction model to predict new target markers of neurodegeneration.

[0036]

[0053] Referring to Figure 6B, 622 shows an example sleep hypnogram from an individual during an overnight EEG recording using a standard 10-20 scalp wet sensor montage, illustrating the results of C3-M1 derivation. This sleep hypnogram shows the individual's wakefulness (W) and N1, N2, N3, or REM sleep stages in 30-second epochs. 642 shows slow-wave activity in bandwidths ranging from 0.5 Hz to 4.5 Hz using a Welch periodogram in 30-second epochs. Slow-wave activity peaks during N3 sleep. Important features of slow-wave activity include duration, power, and frequency, which can be used in conjunction with a target prediction model. For example, the features of the slow-wave sleep element shown in Figure 6B can be used to train a target prediction model. If the EEG slow-wave sleep element is newly collected data after the target prediction model has been trained, the EEG slow-wave sleep data element can be input into the target prediction model to predict new target markers of neurodegeneration.

[0037]

[0054] Simultaneous recordings were made on the same night and on the same individual using ear devices containing modules 100, 200, and 300 as shown in Figure 1A. In Figure 6C, 624 shows the sleep hypnogram recorded by the ear device, and 644 shows the Welch periodogram and slow wave activity with a bandwidth of 0.5 Hz to 4.5 Hz using 30-second epochs.

[0038]

[0055] Referring to Figure 6D, for the same individual from the same night of EEG recording, sleep microstructure spindle activity from a standard 10-20 scalp wet sensor montage using C3-M1 derivation, spanning 10 Hz to 16 Hz bandwidths using a Welch periodogram, is shown at 652. The same analysis is shown at 654 using sensors from the ear device consisting of modules 100, 200, and 300 shown in Figure 1A. In both figures, peak spindle activity is seen to be consistent with N2 stage sleep. Key features of spindle activity include duration, power, and frequency. For example, the features of the sleep spindle elements shown in Figure 6D can be used to train a target prediction model. If the EEG sleep spindle elements are newly collected data after the target prediction model has been trained, the EEG sleep spindle data elements can be input into the target prediction model to predict new target markers of neurodegeneration.

[0039]

[0056] For the same individual on the same night of EEG recording, Figure 6E shows the sensitivity agreement of hypnogram stages and sleep performance parameters between different EEG electrode locations. Figure 6F (682) shows the sensitivity agreement of hypnogram stages between the C3-M1 derivations for scoring the hypnogram and derivations between the left and right ear electrodes shown in Figure 1A (200). Figure 6F (684) shows the agreement of sleep performance parameters derived from the sleep hypnogram between the C3-M1 derivations for measuring the hypnogram and derivations between the left and right ear electrodes shown in Figure 1A (200). The differences in sensitivity agreement and performance parameters are due to anatomical differences in sensor locations for measurement, highlighting the importance of consistent anatomical sensor placement across uses and between users to extract sleep EEG features for training target prediction models and as input to trained target prediction models for predicting new target markers of neurodegeneration. In-ear lead placement, as performed by the wearable ear device shown in FIG. 1A, using the anatomical ear fitting shown in FIGS. 2C-2G, ensures consistent anatomical placement by the user.

[0040]

[0057] 7A-7E illustrate one embodiment of a neurovascular data acquisition module 700 of a system constructed in accordance with the present invention and are not intended to be limiting, as those skilled in the art will recognize upon review of this application that other configurations may be used without departing from the scope of the claimed invention. Referring to FIG. 7A, the same sensors 272 and 273 shown in FIG. 2G used to record in-ear EEG are multiplexed to record transcranial impedance from left to right ear during sleep at 80 kHz, well above the EEG bandwidth. Transcranial impedance tracks fluid changes in the brain that occur during sleep, for example, from sleep-related changes in glymphatic flow and cerebrospinal fluid production, shown in module 710, and from cardiac pulsations, leading to beat-to-beat measurable cerebral perfusion pulsations, shown in module 730. With each heartbeat, a significant portion of cardiac output is ejected into the internal carotid and vertebral arteries and delivered to the brain parenchyma. This change in cerebral blood volume reduces transcranial impedance, which can be measured by the sensors. The amplitude of the impedance waveform and the numerical integral of the impedance waveform pulse are proportional to the volume of blood ejected into the brain during the cardiac cycle. The moving average of transcranial impedance drifts downward during sleep. This drift corresponds to diurnal variations in cerebral fluid volume, for example, fluid volume increasing during sleep from increased glymphatic flow and cerebrospinal fluid production, and can be measured by module 710. A larger drop in transcranial impedance between sleep onset and sleep nadir correlates with greater glymphatic flow and a predicted greater increase in total cerebral fluid composition. The pulsations recorded by module 730 facilitate fluid flow through the brain interstitial space, which promotes glymphatic flow and protein clearance during sleep. A larger magnitude of these pulsations, measured as amplitude or the numerically integrated area that measures ejection volume, is expected to drive greater glymphatic flow, leading to improved clearance. A photoplethysmography sensor placed in the ear of the device shown in FIG. 1A and shown at 264 in FIG. 2F records heart rate variability during sleep in module 750, resting heart rate by module 770, and sinus arrhythmia which can be used to estimate respiration rate, also by module 770.Heart rate variability during sleep provides a measure of sympathetic-parasympathetic tone in the autonomic nervous system, with lower heart rate variability indicating lower parasympathetic tone and higher sympathetic tone. During sleep, higher sympathetic tone, activated by the locus coeruleus in the brain, reduces cerebral interstitial fluid flow and protein biomarker clearance. Higher sleep heart rate variability is expected to result in greater glymphatic flow and protein clearance. Lower sleep resting heart rates, as recorded by module 770, are compensated for by higher stroke volumes to maintain tightly controlled cerebral blood flow. Higher stroke volumes result in larger cerebral perfusion pulsations, measurable by module 730, and enhance the driving force of glymphatic flow. Conversely, diseases that harden arteries, specifically arterioles that enter the brain parenchyma alongside which perivascular spaces containing cerebrospinal fluid propelled into the brain parenchyma to drive glymphatic flow, reduce the driving force. The reduction in propulsive force is measured by a reduction in the magnitude of the cerebral perfusion pulse in module 730. Changes in respiratory rate, recorded in module 770, also affect venous return to the heart, thereby altering stroke volume and thereby altering blood volume pulses in 730.

[0041]

[0058] One or more of the neurovascular data elements shown in Figure 7A can be used in connection with a target prediction model. For example, the neurovascular data elements shown in Figure 7A can be used to train a target prediction model. As another example, if the neurovascular data elements are newly collected data after the target prediction model has been trained, the neurovascular data elements can be input into the target prediction model to predict new target markers of neurodegeneration.

[0042]

[0059] Referring to Figure 7B, 714 shows a sleep recording of cerebrospinal fluid volume from the transcranial impedance sensors shown by 272 and 273 in Figure 2G of the left in-ear device. The sleep hypnogram shown at 712 corresponds to the recorded impedance 714 and shows a drop in impedance as the patient transitions into deeper sleep in stages N2 and N3 (slow wave sleep). The drop in impedance corresponds to an increase in fluid (or water) composition in the brain, which corresponds to an increase in glymphatic flow through brain tissue and a resulting increase in cerebrospinal fluid production.

[0043]

[0060] The intervals between cerebral perfusion pulses during this patient's sleep are shown in 732 and the corresponding ECG signal at 734. Cerebral perfusion output is the impedance difference between the peak and trough of the pulse. A larger difference indicates a larger change in fluid or water composition as a result of the heart beating, which directly correlates with cerebral output.

[0044]

[0061] Referring to FIG. 7C, for each 500 second epoch during sleep, at 752 a sleep hypnogram is shown, at 754 a resting heart rate recording by module 770 is shown, at 756 a resting heart rate variability recording by module 750 is shown, and at both 745 and 756 recordings from sensors by the wearable device shown in FIG. 1A against a baseline electrocardiogram (ECG) recording are shown.

[0045]

[0062] Referring to FIG. 7D, pulse transit time module 790 records a ballistocardiogram (BCG) 791 from the in-ear inertial measurement unit (IMU) shown at 262 in FIG. 2F. The BCG I, J, and K waveforms are highlighted in recording 791. Using wavelet analysis of the BCG recording, the power of the J waveform is shown at 792, whose peak coincides with the BCG J waveform peak. A PPG waveform recording from the in-ear sensor 264 shown in FIG. 2F is shown at 793, and an ECG recording is shown at 794. The IMU, PPG, and ECG recordings are time synchronized by module 790. A close-up of a 2.5 second period of the data shown at 791-794 is shown at 795 in FIG. 7E. A calculation of the cardiac ejection pulse transit time is shown at 795. The maximum blood acceleration is marked by the BCG J wave 796, and the trough of the PPG waveform 797 corresponds to the timing of capillary blood expansion within the ear, which results in a reduction in infrared scattering to the PPG collector. The timing difference between 797 and 796 for that single cardiac cycle is the pulse transit time from the aortic valve to the in-ear sensor. Pulse wave velocity can be calculated as the vascular distance traveled by blood from the aortic valve to the in-ear sensor divided by the pulse transit time. In adults, this distance remains fixed, and changes in pulse transit time are inversely proportional to changes in pulse wave velocity.

[0046]

[0063] Figures 8A and 8B illustrate one embodiment of a glymphatic flow, cerebral proteinopathy, and neurodegeneration target data acquisition module 800 of a system constructed in accordance with the present invention and are not intended to be limiting, as those skilled in the art will recognize upon review of this application that other configurations may be used without departing from the scope of the claimed invention. Referring to Figure 8A, in one embodiment 810, glymphatic flow imaging can be used in conjunction with diffusion tensor MRI to measure glymphatic flow of CSF within the perivascular space, or with a CSF tracer such as gadolinium, which, when injected into the subarachnoid space, is excreted through extravascular glymphatic flow transport in the brain. By timing the injection to just before sleep onset, MRI scans can be performed before sleep and again at the end of sleep. Differences in contrast intensity in key regions of interest provide neuroimaging targets for glymphatic system function occurring during sleep.

[0047]

[0064] In another embodiment 830, a PET scan can be used with a radioactive tracer specific for a particular protein of interest and the standard uptake value ratio (SUVr) to quantify proteinopathy accumulation in the brain. In another embodiment, the radioactive tracer binds to the soluble form of the proteinopathy protein, and the PET scan SUVr quantifies the change in concentration of the soluble protein after sleep.

[0048]

[0065] In another embodiment, MRI 850 can be used to measure changes in brain volume or the degree of neurodegeneration in anatomical regions of interest, such as the hippocampus, entorhinal cortex, thalamus, orbitofrontal cortex, parietal, temporal, anterior and posterior cingulate gyrus, and precuneus region, over two time points separated by several months to provide a target rate of neurodegeneration in an individual.

[0049]

[0066] In another embodiment 870, CSF concentrations of proteinopathy proteins are assayed in the morning or both before and after sleep. The absolute concentration normalized to the population or the diurnal difference between concentrations before and after sleep measures glymphatic flux as the net clearance of the protein and provides a target molecular marker for that protein that also serves as a marker of cerebral proteinopathy and neurodegeneration.

[0050]

[0067] In another embodiment 890, plasma concentrations of proteinopathy proteins are assayed in the morning or both before and after sleep. The absolute concentration normalized to the population or the diurnal difference between concentrations before and after sleep measures glymphatic clearance function as the net glymphatic clearance of the protein and provides a target molecular marker for that protein that also serves as a marker for cerebral proteinopathy and neurodegeneration.

[0051]

[0068] The target marker data acquisition module 800 may be implemented using embodiments 810, 830, 850, 870, or 890. Alternatively, the target marker data acquisition module 800 may be implemented using a combination of the embodiments shown in FIG.

[0052]

[0069] Referring to FIG. 8B, 812 and 814 show PET scans and SUVr markers for amyloid protein in two different individuals. 816 shows the SUVr scale, with values ​​near 0 representing low tracer concentrations, which occur when the tracer binds low or no amyloid protein. Values ​​above 1.5 indicate high tracer concentrations, which occur with high amyloid accumulation. The PET scan shown in 812 shows no amyloid accumulation in any of the brain regions of interest within this coronal slice. In contrast, the scan shown in 814 shows significant amyloid accumulation in many regions of interest (indicated by arrows), including the orbitofrontal cortex and temporal region. Key markers from the PET scan include the SUVr values ​​for each region of interest for proteins of interest that can be used in conjunction with a target prediction model. For example, the markers in the PET scan shown in FIG. 8B can be used as targets in training a target prediction model. After the target prediction model is trained, new EEG and neurovascular features can be input into the target prediction model to predict new target markers of neurodegeneration. The target marker data acquisition module 800 may be implemented using one of the embodiments 810, 830, 850, 870, or 890. Alternatively, the target marker data acquisition module 800 may be implemented using a combination of the embodiments shown in FIG.

[0053]

[0070] FIG. 9 illustrates one embodiment of a biomarker learning module 900 of a system constructed in accordance with the present invention and is not intended to be limiting in scope, as those skilled in the art will recognize upon review of this application that other configurations may be used without departing from the scope of the claimed invention. Referring to FIG. 9 , in one embodiment, input data for 901 can be neurophysiological feature data acquired by module 600, neurovascular feature data acquired by module 700, or a combination thereof, or glymphatic flow target data acquired by 810 of module 800. In another embodiment, input data for 901 can be glymphatic flow markers acquired by 810, and molecular or neuroimaging target output data acquired by 800 can be any one or combination of the embodiments described by 830, 850, 870, or 890 of FIG. 8A. Target marker data can be continuously assessed, e.g., mm markers using diffusion tensor MRI. 2The target neuroimaging marker data can be a pseudodiffusion coefficient for measuring glymphatic flow in units per second, an SUVr in a PET scan, or a picomolar concentration in a CSF molecular assay of a protein, or a categorical value such as high or low based on thresholding a continuous assessment measurement using standardized levels of disease stage. The target neuroimaging marker data can be whole-brain values ​​or values ​​for a specific region of interest known to be primarily affected by a specific neurodegenerative disease process. The input training data 901 should be carefully collected for a population of 12 or more patients at different disease process levels, ranging from healthy to severely affected. Having data representative of patients at different extremes of the disease process provides a range of target marker data values, such as the amyloid PET scan SUVr values ​​shown in Figure 8B for a healthy patient 812 and a patient 814 with advanced amyloid deposition in the brain. These extreme values ​​allow the predictive model and fitting method 903 to learn a combination of feature patterns in the neurophysiological data 600 and neurovascular data 700 corresponding to healthy patients and those corresponding to diseased patients. Each patient's data must be collected simultaneously with the target glymphatic flow, molecular, or neuroimaging marker data, collected together on one or more nights. The device shown in Figure 1 is designed for simultaneous, time-synchronized neurophysiological and neurovascular data capture from patients and to ensure reproducibility of data acquisition by ensuring consistent positioning of all sensors using the anatomical form-fitting ear device shown in Figure 2G. Repeated acquisition of neurophysiological and neurovascular data over several nights provides additional features for training the predictive model in 903, improving the tradeoff between bias and variance in model fitting and improving predictions. The neurophysiological and neurovascular data are preprocessed by feature extraction, as shown in Figures 6A-6F and 7A-7E.

[0054]

[0071] At 903, a model and fitting method are selected. The model can be one of many known machine learning and deep learning models, including those used in commercially available software packages and services, and the fitting method is selected based on the selected model. Random forest, a general-purpose machine learning model, is highly effective at identifying patterns and ranges of neurophysiological and neurovascular feature values ​​that best separate or predict target glymphatic flow, molecular, or neuroimaging markers. Acquiring patient training data is expensive and requires institutional review board-approved clinical trials. Random forest has an advantage over deep learning models in that it provides good results with a relatively small number of patient data examples. Once the model is trained on a patient population representative of the target application, predictions of cerebral glymphatic flow, proteinopathy, or neurodegeneration can be made in new, unknown patients from neurophysiological and neurovascular sleep data acquired by the device shown in FIG. 1A without requiring the patient to undergo expensive and invasive clinical diagnostic procedures such as brain MRI, cerebrospinal fluid molecular analysis, or PET neuroimaging. This allows the device shown in Figure 1A to screen for markers indicative of outlying values ​​of glymphatic flow, cerebral proteinopathy, or neurodegeneration, and to monitor the progression of those markers over time in individuals who screen positive.

[0055]

[0072] After a model, such as a random forest, is selected, neurophysiological and neurovascular data are input features, and target marker data is the target output in model cross-validation 905. Patient data in 901 are first randomly divided into several folds, such as five non-overlapping sets. A model in 903 is fitted to the data in the first four folds, and the fitted model is used to predict the target marker data for patients in the fifth fold. This procedure is repeated four more times, each time excluding a new fold of patients and their data, resulting in a model prediction of the complete target marker data from the input neurophysiological and neurovascular feature data. In 907, the model is evaluated using a goodness-of-fit, which may be the root mean square of the prediction in 903 to the target marker data. Operations 901-907 can be repeated using several different models to determine which model provides the most accurate prediction of the target marker data. The best target prediction model is advanced to 909, where holdout patients and their data, including target marker data never seen by the model, are tested, and the prediction model and its performance against the data are output at 911.

[0056]

[0073] FIG. 10 illustrates one embodiment of a target prediction module 1000 of a system constructed in accordance with the present invention; it is not intended to be limiting in scope, as those skilled in the art will recognize upon review of this application that other configurations may be used without departing from the scope of the claimed invention. Referring to FIG. 10 , in one embodiment, 1001 represents new neurophysiological and neurovascular data acquired from an individual by modules 600 and 700, which is input to the target prediction model output at 911. Features are extracted from the data as shown in FIGS. 6A-6F and 7A-7E. This new neurophysiological and neurovascular data may be acquired from a new, unknown patient, or from a patient who was part of the original training data but for whom new data is acquired months or years later to assess disease progression. In operation 1003, the module runs the target prediction model on the new input data of operation 1001 and outputs 1005 a marker of glymphatic flow. In another embodiment, 1001 is the glymphatic flow output of a target prediction model 1005, and in operation 1003, the module runs the target prediction model on this new input data of operation 1001 and outputs 1005 markers of cerebral proteinopathy or neurodegeneration.

[0057]

[0074] The target prediction can be a predictive value of glymphatic flux, levels of protein accumulation such as those described above, or the extent of neurodegeneration measured as changes in whole brain volume or regional volume using structural MRI. The predictive value can be a continuous value of the marker value used to monitor disease progression or response to intervention, or a categorical value to determine the presence or absence of a disease stage. The target prediction value for disease staging can be submitted for regulatory approval with appropriate regulatory validation studies and for reporting of sensitivity, specificity, positive and negative predictive values ​​compared to a reference standard target such as a molecular analysis or neurodegeneration test.

[0058]

[0075] 11A-11C illustrate one embodiment of a target engagement module 1100 of a system constructed in accordance with the present invention and are not intended to be limiting in scope, as those skilled in the art will recognize upon review of this application that other configurations may be used without departing from the scope of the claimed invention. Referring to FIG. 11A, in one embodiment, target engagement module 1110 records data from an intervention before or during sleep, and module 1130 records neurophysiological and neurovascular data during sleep and input data from intervention performance 1110. In module 1150, the engagement of the intervention with one of the neurophysiological targets or features shown in FIG. 6A or one of the neurovascular targets or features shown in FIG. 7A, or a combination of both, is calculated as the change in the neurophysiological or neurovascular feature value from baseline. Processes 1110, 1130, and 1150 can be repeated multiple times with different levels or dosage intensities of therapeutic intervention to establish target engagement response curves.

[0059]

[0076] The predictive model 911 can be run using data inputs 1110 and 1130 to output a prediction of the expected impact of the therapeutic intervention on glymphatic flow, cerebral proteinopathy, and neurodegeneration, either in the short term or, if sustained, in the long term. The output glymphatic flow, cerebral proteinopathy, or neurodegeneration prediction can be used to characterize the effect of the intervention on glymphatic flow, molecular, or neuroimaging markers of neurodegeneration. In other embodiments, operation 1110 can record a cardiovascular, dietary, pharmaceutical, or neuromodulation intervention, and this data is then input into 1130 for run-through calculation of target engagement in operation 1150. Thus, the effect of the intervention on neurophysiological and neurovascular data can be elucidated, and the predictive model 911 can be used to determine the effect of the intervention on markers of neurodegeneration.

[0060]

[0077] 11B, in one embodiment, the input intervention can raise (reverse Trendelenburg) or lower (Trendelenburg) the head relative to the feet to a flat supine position, as shown at 1112. Trendelenburg increases intracranial perfusion pressure. Monitoring transcranial impedance at 1134 during the intervention at 1112 shows that Trendelenburg causes a decrease in transcranial impedance, returning to normal once the patient is returned to a flat supine position, demonstrating the intervention's targeted involvement of cerebrospinal fluid volume. Monitoring the amplitude envelope of the in-ear PPG tracing at 1132 shows that the opposite occurs, with Trendelenburg causing a decrease in amplitude or blood volume, progressing to the ears and scalp, returning to normal once the patient is returned to a flat supine position.

[0061]

[0078] In another embodiment, the input intervention can be a cardiopulmonary intervention, such as a Valsalva maneuver or hyperventilation. Transcranial impedance tracings after these two cardiopulmonary interventions are shown at 1136 and 1138. At 1136, the patient performed a 45-second Valsalva maneuver beginning at the 1-minute mark. The Valsalva maneuver increases intrathoracic pressure, thereby decreasing venous return from the head. The decrease in venous return results in venous congestion and CSF volume expansion. This is demonstrated by the decrease in transcranial impedance during the Valsalva in tracing 1136. At 1138, the same patient began mild hyperventilation at 1 minute and progressed to full hyperventilation at 1 minute 15 seconds. Each inspiration decreases intrathoracic pressure and increases venous return, resulting in a decrease in CSF volume and an increase in transcranial impedance, as shown in tracing 1138. In both of these interventions, monitoring of neurovascular tracings demonstrated the intervention's targeted engagement with neurovascular measures, in this case CSF volume.

[0062]

[0079] Referring to FIG. 11C, in another embodiment, once the intervention's target contribution to one or more of the neurophysiological and neurovascular data has been determined, a predictive model 911 is used to determine the effect of the intervention on glymphatic flow, cerebral proteinopathy, or markers of neurodegeneration. FIG. 1152 shows the rate of intrathecal gadolinium clearance from the brain measured by serial MRI under three different interventions. The first intervention is a control, i.e., normal sleep without any intervention; the second intervention is sleep deprivation; and the third intervention is dexmedetomidine, which increases the duration of slow-wave sleep. These three interventions contribute to sleep-EEG measures, specifically slow-wave sleep duration. FIG. 1152 shows the corresponding slower clearance of contrast agent with sleep deprivation (no slow-wave sleep) compared to the control, with the control exhibiting slower clearance than dexmedetomidine (increased slow-wave sleep), as expected from a predictive model 911 trained on sleep-EEG data, neurovascular data, and target cerebral proteinopathy and neurodegeneration data. Embodiments described in the section

[0080] In addition to the above, various embodiments of the present disclosure include, but are not limited to, the embodiments described in the following sections:

[0063]

[0081] Item 1. A method implemented by one or more computer processors, comprising: a. accessing, by one or more computer processors, neurophysiological and neurovascular data recorded during sleep; b. performing, by one or more computer processors, a function mapping from said neurophysiological and neurovascular data to targets that are markers of glymphatic flow; c. outputting, by one or more computer processors, a target prediction model based on the function mapping.

[0064]

[0082] Item 2. The method according to Item 1, a. performing, by one or more computer processors, a second function mapping from the marker of glymphatic flow to a target that is one of a molecular analysis marker of neurodegeneration or a neuroimaging marker of neurodegeneration; b. outputting, by one or more computer processors, a target prediction model based on the second function mapping.

[0065]

[0083] Item 3. The method of item 1, wherein the neurophysiological data is an electroencephalogram recording, and sleep macrostructure and sleep microstructure features are extracted from the electroencephalogram recording.

[0084] Item 4. The method of item 1, wherein the neurovascular data includes one or more of electrical impedance recordings taken from the patient's head, photoplethysmography measured at the patient's head, and inertial measurement unit acceleration measured in the patient's ear, and the neurovascular data is used to calculate sleep cerebrospinal fluid volume changes, cerebral perfusion pulsatility, heart rate variability, resting heart rate, pulse transit time, pulse wave velocity, and respiratory rate.

[0066]

[0085] Item 5. The method of item 1, wherein the neurophysiological and neurovascular data are acquired from a wearable device.

[0086] Item 6. The method of item 5, wherein the sensor of the wearable device is inserted into the patient's ear and measurements are taken from the ear or ear canal.

[0067]

[0087] Item 7. The method of item 6, wherein the wearable device has a pouch that is pressurized within the ear canal to increase interface contact of the sensor with the ear canal wall.

[0068]

[0088] Item 8. The method according to item 2, wherein the molecular analytical marker of neurodegeneration is a CSF or plasma assay of one of beta-amyloid, tau, p-tau, alpha-synuclein, and neurofilament light.

[0069]

[0089] Item 9. The method according to item 2, wherein the neuroimaging marker of neurodegeneration is a PET scan with a radioactive tracer that binds to one of beta amyloid, tau, or glucose.

[0070]

[0090] Item 10. The method according to item 2, wherein the neuroimaging marker of neurodegeneration is an MRI scan.

[0091] Item 11. The method according to Item 1, The method further includes inputting the new neurophysiological data and the new neurovascular data into a target prediction model and outputting a predictive marker of glymphatic flow.

[0071]

[0092] Item 12. The method according to Item 1, inputting the intervention data into a target prediction model; inputting new neurophysiological data and new neurovascular data into a target prediction model; outputting a predicted target marker of glymphatic flow; determining the effect of the intervention data on predictive target markers of glymphatic flow.

[0072]

[0093] Item 13. One or more computer processors; a neurophysiological data acquisition module configured to measure neurophysiological data; a neurovascular data acquisition module configured to measure neurovascular data; a transmission module configured to transmit the electroencephalogram data and the neurovascular data to a second computing device.

[0073]

[0094] Item 14. The system of item 13, wherein the one or more computer processors, the neurophysiological data acquisition module, the neurovascular data acquisition module, and the transmission module are disposed in a wearable device.

[0074]

[0095] Item 15. The system of item 14, wherein the wearable device is configured to be worn on the patient's ear.

[0096] Item 16. The system of item 15, wherein the sensor of the wearable device is inserted into the ear and measurements are taken from the ear or ear canal.

[0075]

[0097] Item 17. The system of item 16, wherein the wearable device includes a pouch that is pressurized within the ear canal to increase interfacial contact of the sensor with the walls of the ear canal.

[0076]

[0098] Item 18. The system according to item 13, wherein the neurophysiological data is an electroencephalogram recording, and wherein sleep macrostructure and sleep microstructure features are extracted from the electroencephalogram recording.

[0077]

[0099] Item 19. The system of item 13, wherein the neurovascular data includes one or more of electrical impedance recordings taken from the patient's head, photoplethysmography measured at the patient's head, and inertial measurement unit acceleration measured in the patient's ear, and the neurovascular data is used to calculate sleep cerebrospinal fluid changes, cerebral perfusion pulsations, heart rate variability, resting heart rate, pulse transit time, pulse wave velocity, and respiratory rate.

[0078]

[0100] Item 20. The system of item 13, wherein the second computing device is one of a mobile phone, a local computing device, and a remote computing device.

[0079]

[0101] Item 21. The system of item 13, wherein the second computing device includes a machine learning algorithm configured to input neurophysiological data and neurovascular data and configured to output a target prediction model.

[0080]

[0102] Item 22. The system of Item 21, wherein the target prediction model is configured to receive new neurophysiological data and new neurovascular data as inputs and to output predictive markers of neurodegeneration.

[0081]

[0103] Item 23. The system of item 21, wherein the target prediction model is configured to receive as inputs intervention data, new neurophysiological data, and new neurovascular data, and to output predictive markers of cerebral proteinopathy or neurodegeneration and a determination of the effect of the intervention data on the predictive markers of cerebral proteinopathy or neurodegeneration.

[0082]

[0104] Item 24. A method implemented by one or more computer processors, comprising: a. accessing, by one or more computer processors, neurophysiological and neurovascular data recorded during sleep; b. performing, with one or more computer processors, function mapping from the neurophysiological and neurovascular data to a target, wherein the target is one or more of a marker of glymphatic flow, a molecular analysis marker of neurodegeneration, or a neuroimaging marker of neurodegeneration; c. outputting, by one or more computer processors, a target prediction model based on the function mapping; and d. receiving as input into a target prediction model the newly collected neurophysiological and neurovascular data collected during sleep; e. receiving as input data associated with a therapeutic intervention; f. Measuring the effectiveness of the therapeutic intervention.

[0083]

[0105] Item 25. A method implemented by one or more computer processors, comprising: a. activating a neurophysiological data sensor and a neurovascular data sensor of a wearable device; b. collecting neurophysiological and neurovascular data during sleep with sensors; c. storing the neurophysiological data and the neurovascular data in a storage device of the wearable device; d. processing the neurophysiological and neurovascular data using a processor of the wearable device; e. transmitting, by a transmission module of the wearable device, the neurophysiological data and the neurovascular data to a second computing device.

[0084]

[0106] Item 26. A non-transitory computer-readable medium containing computer-executable instructions for performing the operations recited in item 1.

[0107] Item 27. A non-transitory computer-readable medium containing computer-executable instructions for performing the operations recited in Item 24.

[0085]

[0108] Item 28. A non-transitory computer-readable medium containing computer-executable instructions for performing the operations recited in Item 25. Additional Examples of Wearable Devices for Measuring Transcranial Impedance

[0109] The following description provides additional examples of head-mounted wearable devices that can be used to measure glymphatic flow for the purpose of assessing neurodegeneration. Similar to the previous example, the following example of a head-mounted wearable device uses electrodes to measure transcranial impedance. The electrodes can include input electrodes that introduce a small input current at the scalp of a subject wearing the device while sleeping. The wearable device includes additional electrodes that can measure electrical potentials at various locations on the head in response to the input current introduced by the input electrodes. The electrical potential measurements are used to identify equipotential surfaces. The equipotential measurements are collected over time while the subject is sleeping, and the time-dependent change in the equipotential provides a measure of glymphatic flow. The frequency of the input current introduced by the input electrodes can also be varied to obtain a more complete set of impedance data.

[0086]

[0110] The present invention and details of various embodiments can be better understood with reference to the drawings. Figures 12A, 12B, and 12C illustrate one embodiment of a functional description of a system configured in accordance with the present invention and are not intended to be limiting in scope, as one of ordinary skill in the art will recognize, upon review of this application, that other configurations can be used without departing from the scope of the present disclosure. Referring to Figure 12A, transcranial electrodes for acquiring continuous electrical impedance spectroscopy data can be positioned over different anatomical cerebral lobes, such as the left frontal lobe 1210 and the corresponding right frontal lobe (not visible in Figure 12A), the left parietal lobe 1220 and the corresponding right parietal lobe (not visible in Figure 12A), and the left temporal lobe 1230 and the corresponding right temporal lobe (not visible in Figure 12A). The system can include a wearable device that can be worn on the subject's head and is designed to position the transcranial electrodes over the different transcranial cerebral lobes. 12B and 12C , wearable device 1250 can be worn over a user's head similar to headphones and includes left and right bands 1251 and 1256 joined at their proximal ends by a central band 1255. The distal end of left band 1251 includes a left earpiece 1253 that can encircle at least a portion of a subject's left ear, and the distal end of right band 1256 includes a right earpiece 1259 that can encircle at least a portion of the subject's right ear. Although not required, left earpiece 1253 can include an optional left earbud 1254 that fits within the subject's left ear canal, and right earpiece 1258 can include an optional right earbud 1259 that fits within the subject's right ear canal. If earbuds are included, the earbuds can be used for correct positioning of the wearable device and can include optional sensors. The central band 1255 can house a controller for controlling the operation of the wearable device. The controller can include a power source (e.g., a battery) for providing current to the input electrodes. The controller can also include a processor and memory for controlling the power source and for processing electrical potential data measured by the electrodes.The controller may further include a transmission module for wired or wireless communication of data to and from a nearby computing device, such as a mobile phone, tablet, or PC. The inner surfaces of the left band 1251 and right band 1256, which contact the subject's scalp, may include multiple impedance spectroscopy electrodes 1275. In some embodiments, the impedance spectroscopy electrodes may be positioned along the inner surface of the central band 1255. A power source may provide an input current to one of the electrodes 1275, while the other electrodes measure electrical potentials in response to the input current. While collecting impedance measurements during sleep, the controller may change which of the multiple electrodes of the wearable device 1250 provides the input current to collect a broader set of electrical potential measurements. The shape of the wearable device 1250 allows for noninvasive collection of electrical impedance spectroscopy data across the left and right sides of the head and the temporal or parietal lobes.

[0087]

[0111] 12D and 12E , a wearable device 1260 includes similar components to the previously described wearable device 1250. Briefly, the wearable device 1260 includes a left front band 1261, a left rear band 1262, a right front band 1266, and a right rear band 1267 joined at their respective proximal ends by a central band 1265. The distal ends of the left front band 1261 and the left rear band 1262 include a left earpiece 1263 that can encircle at least a portion of a subject's left ear, and the distal ends of the right front band 1266 and the right rear band 1267 include a right earpiece 1269 that can encircle at least a portion of a subject's right ear. Although not required, the left earpiece 1263 can include an optional left earbud 1264 that fits into the subject's left ear canal, and the right earpiece 1268 can include an optional right earbud 1269 that fits into the subject's right ear canal. If earbuds are included, they can be used to properly position the wearable device and can include optional sensors. The center band 1265 can incorporate a controller for controlling the operation of the wearable device. The controller can include a power source (e.g., a battery) for providing current to the input electrodes. The controller can also include a processor and memory for controlling the power source and processing potential data measured by the electrodes. The controller can further include a transmission module for wired or wireless communication of data to and from a nearby computing device, such as a mobile phone, tablet, or PC. The inner surfaces of the left anterior and posterior bands 1261 and 1262 and the right anterior and posterior bands 1266 and 1267 can include a plurality of electrodes 1276 that contact the subject's scalp and are used to make impedance measurements along the subject's head. The controller can select one of the electrodes as an input electrode that provides an input current, and the other electrodes can measure electrical potentials in response to the input current.

[0088]

[0112] Wearable device 1260 differs from wearable device 1250 in that it includes multiple bands on the left and right sides that drape over the user's head to spread the distribution of electrodes on the user's head. In the examples shown in FIGS. 12D and 12E, two bands are shown on the left and two bands are shown on the right. However, other embodiments may include additional bands on each side to further distribute the electrodes over the subject's head. In the case of impedance spectroscopy electrodes 1276 located along the inward-facing surfaces of left anterior and posterior bands 1261 and 1262 and right anterior and posterior bands 1266 and 1267 of the wearable device, the electrodes contact the subject's head and can acquire impedance spectroscopy data across the temporal, parietal, and frontal lobes for noninvasive measurement of glymphatic flow.

[0089]

[0113] FIG. 13 illustrates an embodiment of an electrical impedance spectroscopy sensor in a system constructed in accordance with the present invention in a head cap used for clinical electroencephalography (EEG) data acquisition. The illustration is not intended to be limiting, as those skilled in the art will recognize, upon review of this application, that other configurations may be used without departing from the scope of the present disclosure. Referring to FIG. 13 , electrodes 1310 are placed on the left frontal lobe, 1320 on the right frontal lobe, 1330 on the left temporal lobe, 1340 on the right temporal lobe, 1350 on the left parietal lobe, and 1360 on the right parietal lobe. EEG signals have unique frequencies and are passively measured by electrodes in contact with the subject's head. However, electrical impedance spectroscopy systems multiplex data acquisition through these electrodes during clinical EEG data acquisition. As previously mentioned, electrical impedance measurement is an active, not passive, technique in that input currents are provided at input electrodes and response potentials are measured at other electrodes.

[0090]

[0114] Continuing with the description of the head cap embodiment of the wearable device shown in FIG. 13, FIGS. 14A and 14B illustrate one embodiment of a technique for measuring electrical impedance spectroscopy with a head cap used for clinical electroencephalography data acquisition. While not shown in FIGS. 13, 14A, and 14B, a controller similar to the controller described above in FIGS. 12B-12E can be attached to the top or side of the head cap and electrically coupled to the electrodes. Alternatively, the controller can be separate from the head cap but electrically coupled to the head cap electrodes by one or more cables that transmit power and data signals between the controller and the electrodes. The illustrations are not intended to be limiting, as those skilled in the art will recognize, upon review of this application, that other configurations can be used without departing from the scope of this disclosure. Referring to FIG. 14A, an electrical impedance spectroscopy system inputs current across multiple frequencies into a left temporal electrode 1410 using a sinking right temporal electrode 1420. During current input, the system measures the potential at the remaining electrodes shown in FIG. 14A. This process is repeated throughout the night while the user is asleep, across all selected frequencies, to obtain a time series of impedance spectroscopy data. A finite element model is used to derive equipotential surfaces, indicated by 1430 and 1450 overlying the left hemisphere and 1440 and 1460 overlying the right hemisphere, from the potentials recorded at the sensing electrodes. Other equipotential surfaces can be derived from the electrodes, and increasing the number of electrodes on the brain allows for greater spatial resolution of the equipotential surfaces. The time dependence of these equipotential surfaces during sleep is used to measure glymphatic flow and flow shunting across different anatomical regions of the brain.

[0091]

[0115] As shown in Figure 14B, the electrodes for the current input and sink can be rotated to new electrodes 1470 on the left frontal lobe and 1480 on the right posterior temporal lobe. Changing the orientation of the input and sink electrodes determines the value of a new equipotential surface 1490. The input and sink electrodes can be rotated through all diametrically opposed pairs of electrodes, with each new orientation determining a new equipotential surface. The time series of changes recorded during a sleep episode in the equipotential surface from a particular pair of input-sink electrodes are measures of brain compartment volume changes and glymphatic flow. Multiplexing electrical impedance spectroscopy with input-sink electrode rotation during a sleep episode and taking the union of the resolved equipotential surfaces at each different orientation provides topographic measures of brain compartment volume changes, glymphatic flow, and flow shunts in the sleeping brain.

[0092]

[0116] 15-19, further details are shown of an exemplary embodiment of a wearable device in the form of a head cap. As with the previous example, the electrodes of the head cap described in connection with FIGS. 15-19 can be multiplexed to passively measure EEG signals and actively measure electrical impedance to assess glymphatic flow. The illustrations of FIGS. 15-19 are not intended to be limiting, as those skilled in the art will recognize upon review of this application that other configurations can be used without departing from the scope of the present disclosure.

[0093]

[0117] FIG. 15 shows further details of a controller 1500 for use with the head cap form of the wearable device of FIGS. 15-19. The controller 1500 can also be used with the previously described wearable device for measuring transcranial electrical impedance. As previously described, the controller 1500 can be attached directly to the head cap or can be a separate component coupled to the electrodes of the head cap by electrical wiring. The controller 1500 includes a main module 1501 and a multiplexer module 1502. The main module 1501 includes a power source (battery) 1503, a microcontroller 1506, a bioimpedance circuit 1504, and a memory 1505. The battery 1503 provides power to the components of the controller 1500, including the input current used for impedance measurements. The microcontroller 1506 executes instructions to control the operation of the controller 1500, and the memory 1505 can store executable instructions and impedance data collected from the electrodes. The bioimpedance circuit 1504 is based on a four-point or Kelvin configuration, where two electrodes / leads input current (I+, I-) and the other two electrodes / leads measure the resulting potential difference (V+, V-). This measurement occurs at one or more frequencies between 100 Hz and 1 GHz. To provide tomography capabilities, a multiplexer module 1502 expands the connectivity of the bioimpedance circuit 1504 by providing four multiplexers 1508 (8:1, although 16:1 is also possible) connected to each of the I+, I-, V+, and V- leads, allowing for dynamic reconfiguration of the drive / sense configuration. These multiplexers are programmed on the fly by the microcontroller 1506. For parallel programming multiplexers, an input / output expander 1507 is used, controlled by an I2C or SPI bus 1509.

[0094]

[0118] The addition of these multiplexers allows for sensing of scalp potentials across a large portion of the scalp surface relative to a common reference (see Figure 17) while simultaneously changing the current input point (see Figure 16). Together with measurements at multiple frequencies, such a system provides sequential multi-frequency electrical impedance tomography (MFEIT) capability.

[0095]

[0119] Referring to FIG. 16, a diagram for collecting impedance measurements using the wearable device of FIG. 15 is shown. FIG. 16 illustrates a method for changing input-sink electrodes. Two multiplexers redirect input current from the current source leads (I+, I−) to two cap electrodes (1610) selected from any of the cap electrodes to sequentially input excitation in the subject's scalp current along different vectors. An example of the controller 1500 using an 8:1 multiplexer allows five different vectors to be selected as shown in FIG. 16. However, in other embodiments, fewer or more input-sink electrode pairs can be implemented.

[0096]

[0120] FIG. 17 shows a diagram of a method for measuring potentials from which impedance data can be derived. Two other multiplexers in the controller 1500 sequentially connect one of the two sensing inputs (V+, V−) of the bioimpedance circuit 1504 to one of the cap electrodes (1711) not used for the current input, while the other input is designated as a common reference (1712). The common reference electrode can be a cap electrode or a forehead electrode. Note that a multi-channel sensing system can also be used to sense all electrodes simultaneously without the need to multiplex pairs of sensing electrodes. This eliminates the need for a sensing multiplexer in the multiplexer module 1502 coupled to the sensing inputs (V+, V−).

[0097]

[0121] The current implementation using an 8:1 multiplexer allows 12 electrodes (other than the current input-sink electrodes) across the cap to be measured against a reference electrode, thereby achieving uniform potential mapping across the head.

[0098]

[0122] Combined with the five current input-sink configurations, this results in a total of 60 different electrical measurements for a given excitation frequency. These 60 different configurations are represented in the MUX state machine shown in Figure 19 using the logic diagram shown in Figure 18. Note that different mux state machines and logic diagrams can be used depending on the underlying multi-channel analog front end used.

[0099]

[0123] Referring to FIG. 20 , exemplary data from a head cap embodiment of a wearable device according to the present disclosure is shown, obtained from a subject wearing and sleeping for 100 minutes using the head cap shown in FIGS. 14A and 14B . At 2010 and 2050, two directions of input current are shown through two pairs of electrodes on the head cap, marked T3 and T4 in 2010 and F7 and T6 in 2050. The input current directions at 2010 correspond to MUX states #25-#36 of the MUX state machine of FIG. 19 . The sensing electrodes for these states are Fp1, F7, T5, O1, F3, P3, Ep2, F8, T6, O2, F4, and P4, whose locations are shown at 2010. At 2020, 2030, and 2040 in FIG. 20, the measured potentials at the sensing nodes are shown from wakefulness to sleep for 10, 45, and 90 minutes of sleep, respectively. Input current direction 2010 reveals a progression of increasing potentials at 2020, 2030, and 2040, starting anteriorly in the left frontal lobe and spreading posterolaterally, corresponding to increased fluid movement into the extracellular compartment from 10 to 90 minutes of sleep. The input current direction at 2050 corresponds to MUX states #13-#24 in the MUX state machine of FIG. 19 . Input current direction 2050 reveals a second progression of increasing potentials at 2060 (10 minutes), 2070 (45 minutes), and 2080 (90 minutes), starting on the left temporal lobe and spreading posteriorly, corresponding to increased fluid movement into the extracellular compartment from 10 to 90 minutes of sleep. Fluid movement from intracellular to extracellular compartments of brain parenchyma during sleep occurs due to glymphatic flow. FIG. 20 shows that this embodiment measures both the magnitude of fluid movement, or glymphatic flow, and the topographical dynamics of flow through brain regions. A higher density of electrodes in the head cap used would result in a higher topographic resolution of flow during sleep, consistent with the present embodiment.

[0100]

[0124] Additional exemplary embodiments described in the section

[0125] In addition to the above, various embodiments of the present disclosure include, but are not limited to, the embodiments described in the following sections:

[0101]

[0126] Item 1. A device for topographic localization of glymphatic flow, comprising: a left band including a left electrode along an inner surface of the left band; a right band including a right electrode along an interior surface of the right band; a connecting band coupling the left band and the right band, the connecting band including a controller, the controller including a processor and a power source electrically coupled to the left electrode and the right electrode for collecting potential data for determining impedance; Including, A device in which at least one of the left and right electrodes is an input electrode that supplies an input current over a range of frequencies generated from a power source, and the other of the left and right electrodes measures an electrical potential in response to the input current.

[0102]

[0127] Item 2. The device of item 1, wherein the left and right impedance data derived from the potential data provide an indication of changes in glymphatic flow localized to the left and right electrodes.

[0103]

[0128] Item 3. The device according to item 1, which generates a topographic image showing changes in glymphatic flow based on left and right impedance data derived from potential data.

[0104]

[0129] Item 4. The device of item 1, wherein the controller changes the input electrode from at least one electrode to a different electrode among the left electrode and the right electrode.

[0130] Item 5. The device of item 4, wherein potential data collected from multiple input electrode configurations is combined into a topographic representation of changes in glymphatic flow.

[0105]

[0131] Item 6. The device of item 4, wherein the processor applies a predictive model to potential data collected from multiple input electrode configurations as a topographical representation of changes in glymphatic flow.

[0106]

[0132] Item 7. A device for topographic localization of glymphatic flow, comprising: left anterior and left posterior bands, each including a left electrode along an inner surface of the left anterior and left posterior bands; right anterior and right posterior bands, each including a right electrode along an inner surface of the right anterior and right posterior bands; connecting bands coupling the left anterior and left posterior bands to the right anterior and right posterior bands, the connecting bands including a controller, the controller including a processor and a power source electrically coupled to the left and right electrodes for collecting electrical potential data for determining impedance; Including, A device in which at least one of the left and right electrodes is an input electrode that supplies an input current over a range of frequencies generated from a power source, and the other of the left and right electrodes measures an electrical potential in response to the input current.

[0107]

[0133] Item 8. The device of item 7, wherein the left and right impedance data derived from the potential data provide an indication of changes in glymphatic flow localized to the left and right electrodes.

[0108]

[0134] Item 9. The device according to item 7, which generates a topographic image showing changes in glymphatic flow based on left and right impedance data derived from the potential data.

[0109]

[0135] Item 10. The device according to item 7, wherein the controller changes the input electrode from at least one electrode to a different electrode among the left electrode and the right electrode.

[0110]

[0136] Item 11. The device of item 10, which combines potential data collected from multiple input electrode configurations into a topographic representation of changes in glymphatic flow.

[0111]

[0137] Item 12. The device of item 10, wherein the processor applies a predictive model to potential data collected from multiple input electrode configurations as a topographical representation of changes in glymphatic flow.

[0112]

[0138] Item 13. A device for topographic localization of glymphatic flow, comprising: a head cap configured to be worn on the subject's head, the head cap including left and right electrodes, the left and right electrodes multiplexed to collect EEG signals and transcranial impedance signals; a controller coupled to the head cap and including a processor and a power supply electrically coupled to the left and right electrodes for collecting potential data for determining impedance; Including, A device wherein at least one of the left and right electrodes is an input electrode that supplies an input current over a range of frequencies generated by a power source, and the other of the left and right electrodes measures an electrical potential in response to the input current.

[0113]

[0139] Item 14. The device of item 13, wherein the left and right impedance data derived from the potential data provide an indication of changes in glymphatic flow localized to the left and right electrodes.

[0114]

[0140] Item 15. The device according to item 13, which generates a topographic image showing changes in glymphatic flow based on left and right impedance data derived from potential data.

[0115]

[0141] Item 16. The device according to item 13, wherein the controller changes the input electrode from at least one electrode to a different electrode among the left electrode and the right electrode.

[0116]

[0142] Item 17. The device of item 16, which combines potential data collected from multiple input electrode configurations into a topographic representation of changes in glymphatic flow.

[0117]

[0143] Item 18. The device of item 16, wherein the processor applies a predictive model to potential data collected from multiple input electrode configurations as a topographic representation of changes in glymphatic flow.

Claims

1. 1. A device for topographic localization of glymphatic flow, comprising: a left band including a left electrode along an inner surface of the left band; a right band including a right electrode along an inner surface of the right band; a connecting band coupling the left band and the right band, the connecting band including a controller, the controller including a processor and a power source electrically coupled to the left electrode and the right electrode for collecting potential data for determining impedance; Including, a device wherein at least one of the left electrode and the right electrode is an input electrode that supplies an input current over a range of frequencies generated from the power source, and the other of the left electrode and the right electrode measures an electrical potential in response to the input current.

2. 2. The device of claim 1, wherein left and right impedance data derived from the potential data provide an indication of changes in the glymphatic flow localized to the left and right electrodes.

3. The device of claim 1 , wherein the device generates a topographic image showing changes in glymphatic flow based on left and right impedance data derived from the potential data.

4. The device of claim 1 , wherein the controller changes the input electrode from the at least one electrode to a different one of the left electrode and the right electrode.

5. The device of claim 4 , wherein the device combines the potential data collected from multiple input electrode configurations into a topographical representation of changes in the glymphatic flow.

6. The device of claim 4 , wherein a processor applies a predictive model to the potential data collected from multiple input electrode configurations as a topographical representation of changes in the glymphatic flow.

7. 1. A device for topographic localization of glymphatic flow, comprising: a left anterior band and a left posterior band, each including a left electrode along an inner surface of the left anterior band and the left posterior band; a right anterior band and a right posterior band, each including a right electrode along an inner surface of the right anterior band; connecting bands coupling the left anterior and left posterior bands to the right anterior and right posterior bands, the connecting bands including a controller, the controller including a processor and a power source electrically coupled to the left and right electrodes for collecting electrical potential data for determining impedance; Including, a device wherein at least one of the left electrode and the right electrode is an input electrode that supplies an input current over a range of frequencies generated from the power source, and the other of the left electrode and the right electrode measures an electrical potential in response to the input current.

8. 8. The device of claim 7, wherein left and right impedance data derived from the potential data provide an indication of changes in the glymphatic flow localized to the left and right electrodes.

9. The device of claim 7 , wherein the device generates a topographic image showing changes in glymphatic flow based on left and right impedance data derived from the potential data.

10. The device of claim 7 , wherein the controller changes the input electrode from the at least one electrode to a different one of the left electrode and the right electrode.

11. The device of claim 10 , wherein the device combines the potential data collected from multiple input electrode configurations into a topographical representation of changes in the glymphatic flow.

12. The device of claim 10 , wherein a processor applies a predictive model to the potential data collected from multiple input electrode configurations as a topographical representation of changes in the glymphatic flow.

13. 1. A device for topographic localization of glymphatic flow, comprising: a head cap configured to be worn on the subject's head, the head cap including left and right electrodes, the left and right electrodes being multiplexed to collect EEG signals and transcranial impedance signals; a controller coupled to the head cap, the controller including a processor and a power source electrically coupled to the left electrode and the right electrode for collecting potential data for determining impedance; Including, a device wherein at least one of the left electrode and the right electrode is an input electrode that supplies an input current over a range of frequencies generated by the power source, and the other of the left electrode and the right electrode measures an electrical potential in response to the input current.

14. 14. The device of claim 13, wherein left and right impedance data derived from the potential data provide an indication of changes in the glymphatic flow localized to the left and right electrodes.

15. The device of claim 13 , wherein the device generates a topographic image showing changes in glymphatic flow based on left and right impedance data derived from the potential data.

16. The device of claim 13 , wherein the controller changes the input electrode from the at least one electrode to a different one of the left electrode and the right electrode.

17. 17. The device of claim 16, wherein the device combines the potential data collected from multiple input electrode configurations into a topographical representation of changes in the glymphatic flow.

18. 17. The device of claim 16, wherein a processor applies a predictive model to the potential data collected from multiple input electrode configurations as a topographical representation of changes in the glymphatic flow.

Images