Ultrasonic transducer array and system for high-resolution intracranial characterization through an intact skull

Abstract

A system and method for high-resolution, non-invasive characterization of a cranial cavity and its contents through an intact skull.

Description

[0001] ULTRASONIC TRANSDUCER ARRAY AND SYSTEM FOR HIGH-RESOLUTION INTRACRANIAL CHARACTERIZATION THROUGH AN INTACT SKULL

[0002] FIELD OF THE INVENTION

[0003] [1] The present invention relates generally to systems and methods for non-invasive acoustic interrogation of the cranial cavity. In particular, the invention relates to a high-bandwidth ultrasonic transducer array and a physics-informed computational framework for high-fidelity characterization of intracranial structures and physiological parameters through an intact skull.

[0004] BACKGROUND

[0005] [2] Stroke and traumatic brain injury (TBI) are among the leading causes of global mortality and long-term disability. Annually, approximately 15 million people suffer a stroke, resulting in 5 million deaths and another 5 million left with permanent disabilities. Similarly, TBI affects an estimated 69 million people each year. Beyond the human toll, the financial burden is staggering, with stroke and TBI contributing over $1 trillion in combined annual costs. Rapid diagnosis and continuous monitoring are essential for effective intervention; however, current imaging modalities face significant limitations in providing timely, accessible, and high-fidelity brain assessment.

[0006] [3] Stroke and TBI diagnosis and treatment depend significantly on brain imaging, which is essential for ensuring accurate assessment and enabling timely and appropriate care. Beyond these urgent applications, brain imaging also serves as a vital tool for detecting and monitoring a wide range of conditions, including tumors, multiple sclerosis, neurodegenerative diseases, and other brain disorders and pathologies. However, the effective delivery of such care is currently hindered by the limitations of existing imaging infrastructures.

[0007] [4] Technologies such as magnetic resonance imaging (MRI), computed tomography (CT), functional MRI (fMRI), positron emission tomography (PET), electroencephalography (EEG), functional near-infrared spectroscopy (fNTRS), and transcranial Doppler ultrasound (TCD), have all been instrumental in advancing our understanding of brain structure and function. While these technologies enable medical interventions that can save lives, current imaging technologies also face significant limitations. While MRI, fMRI, CT, and PET are invaluable for diagnosing brain conditions, they are costly, complex, bulky, and require specialized infrastructure, which limits their portability and accessibility, especially during emergencies orin resource-limited settings. Transporting patients from the ER / ICU to these scanners can lead to medical complications, such as disconnecting life-support devices or events requiring resuscitation, especially for critically ill patients. Additionally, CT and PET expose patients to ionizing radiation, while MRI and fMRI require patients to remain completely still for long periods, which is challenging for children or individuals with movement disorders. The confined spaces of these scanners can also cause significant discomfort, and patients with metallic implants are often unable to undergo MRI due to safety risks.

[0008] [5] Portable modalities like TCD and EEG, while more affordable, have limited utility due to relatively poor spatial resolution. EEG does not have the ability to localize deeper brain structures or create detailed brain images. TCD is highly operator-dependent and is restricted to specific temporal windows where the skull does not obstruct ultrasound transmission, limiting its use. Similarly, fNTRS faces spatial resolution constraints and is limited to cortical activity, making it less versatile than MRI or PET for comprehensive imaging. Furthermore, portable versions of CT / MRI often fail to deliver satisfactory image quality and still require specialized setups for use.

[0009] [6] The aforementioned limitations result in delayed diagnoses and suboptimal care, especially for life-threatening conditions like stroke and TBI, where every moment is critical. Such delays can have devastating consequences, leading to increased complications, irreversible brain damage, and a higher risk of mortality. These limitations, combined with the substantial financial and mortality burden associated with these conditions, underscore the urgent need for more timely and accessible brain imaging solutions that can be deployed at the point of care.

[0010] [7] However, achieving true portability and accessibility involves fundamental physical trade-offs across different imaging modalities. For electromagnetic and radiation-based technologies like MRI and CT, the necessity for high-strength magnetic fields or high-energy radiation sources inherently requires large, heavy, and power-intensive hardware to maintain diagnostic signal-to-noise ratios; as these systems are miniaturized for portability, they often suffer from a significant loss in spatial resolution and image quality. In contrast, while ultrasound technology is inherently portable and cost-effective, its application to brain imaging is fundamentally obstructed by the human skull.SUMMARY OF THE INVENTION

[0011] [8] The present invention provides a system and method for non-invasive intracranial characterization through an intact skull using low-frequency acoustic waves. The invention addresses the limitations of conventional transcranial imaging by integrating high-bandwidth acoustic sensing with a physics-informed computational inversion framework.

[0012] [9] In one aspect, the invention relates to an ultrasonic transducer array comprising a plurality of transducer elements. Each element is configured for high-fidelity, low-frequency trans-skeletal propagation, exhibiting a fractional bandwidth of at least 80% at center frequencies below 1 MHz. In some embodiments, the transducer elements comprise a layered acoustic structure including a transduction layer and at least one matching layer. The array is configured to provide angularly diverse acoustic propagation paths, capturing waveform-level information while maintaining phase coherence across the cranial interface.

[0013]

[0010] In another aspect, the system is configured to implement a synthetic aperture framework. By utilizing diverse transmission-reception patterns and spatial diversity, the system synthesizes a large effective aperture to enhance spatial resolution and depth sensitivity within the intracranial cavity. This synthetic aperture approach, combined with the wide-bandwidth interrogation, enables the system to overcome traditional resolution limits of low-frequency transcranial ultrasound.

[0014]

[0011] In a further aspect, the invention provides a processing unit configured to perform physics-informed computational inversion. The processing unit extracts waveform-level acoustic information - including phase, temporal, and multi-path characteristics - and transforms these into a physically consistent, spatially resolved representation of the intracranial environment. The inversion framework accounts for inhomogeneous cranial properties and non-linear propagation effects, resolving structural and functional heterogeneities on the order of the wavelength by mapping measured wavefield data to the internal properties of the cranial cavity. The system optionally utilizes a distributed computing architecture, comprising local and cloud-based high-performance computing (HPC) resources, to enable real-time longitudinal monitoring of physiological changes at the point of care.BRIEF DESCRIPTION OF THE DRAWINGS

[0015]

[0012] Fig. 1A is a side view of a high-density cranial transducer array, illustrating a configuration of approximately 1000 transducer elements arranged in a conformal, helmet-like geometry optimized for trans-skeletal interrogation.

[0016]

[0013] Fig. IB is a top view of the high-density cranial transducer array of Fig. 1A, illustrating the spatially diverse and dense distribution of the elements across the surface sampling manifold.

[0017]

[0014] Fig. 2 is a high-level system diagram illustrating the distributed computing architecture, including the interface between the transducer array, local processing resources, and a cloudbased inversion engine.

[0018]

[0015] Fig. 3A is a detailed cross-sectional view of an individual ultrasonic transducer element, illustrating its layered acoustic matching and backing structure.

[0019]

[0016] Fig. 3B is a schematic perspective view of a portion of the transducer array, illustrating the physical integration of the layered elements.

[0020]

[0017] Fig. 4A is a top-view schematic representation illustrating the interaction of an initial acoustic wavefront with the heterogeneous skull and intracranial brain tissues, showing the onset of phase aberrations.

[0021]

[0018] Fig. 4B is a side-view wave evolution sequence, illustrating the complex multi-path scattering, distortions, and wavefield variations that occur within the cranial cavity over time during trans-skeletal propagation.

[0022]

[0019] Fig. 5 is a detailed block diagram of the ultrasound system architecture, illustrating the signal acquisition chain including amplifiers, gain control, and high-speed digitization components.

[0023]

[0020] Fig. 6 is a flowchart illustrating an operational method for data acquisition, synchronization, and physics-informed computational inversion.

[0024]

[0021] Fig. 7A is a simulated comparative analysis featuring a plurality of 2D slices all extracted from a single reconstructed 3D volume, showing a ground-truth (GT) representation derived from a realistic MRI-based anatomical model with an inserted hemorrhage (left), and a corresponding physical-domain reconstruction generated by the system (right) illustrating the successful recovery of the hemorrhage.

[0025]

[0022] Fig. 7B is a simulated comparative analysis featuring 2D slices extracted from a reconstructed 3D output of a healthy, non-pathological cranial environment, illustrating the system's ability to resolve consistent volumetric structures across different spatial planes.DETAILED DESCRIPTION OF THE INVENTION

[0026]

[0023] The human skull presents a highly heterogeneous, anisotropic, and frequencydependent acoustic barrier that induces severe distortion, scattering, refraction, and attenuation of acoustic waves. These effects disrupt conventional assumptions of straight-line propagation and stable phase relationships, significantly degrading signal fidelity and spatial interpretability.

[0027]

[0024] As a result, most existing ultrasound-based approaches to brain assessment rely on simplified measurement strategies, restricted acoustic windows, or indirect indicators, which limit spatial resolution, depth sensitivity, or robustness across patients. Many techniques further discard complex propagation effects, such as phase distortion, multi-path interference, and mode conversion, as noise or artifacts, rather than treating them as informative components of the measurement process.

[0028]

[0025] Consequently, a critical gap exists in neurological care: there remains an urgent unmet need for a brain imaging solution that combines the safety and portability of ultrasound with the high-fidelity spatial resolution typically reserved for stationary scanners. Because current ultrasound systems fail to resolve intracranial heterogeneities through the intact skull, clinicians are still forced to choose between high-risk patient transport to MRI / CT or relying on insufficient qualitative bedside data. There is, therefore, a necessity for a system that can bridge this gap by fundamentally overcoming the acoustic barriers of the skull to provide diagnostic-grade imaging at the point of care.

[0029]

[0026] To address this need and resolve intracranial structures on the order of the wavelength without sacrificing safety or portability, a fundamental shift is required- one that integrates specialized hardware engineering with complex wave-field physics. This necessitates a synergistic approach combining high-sensitivity, wide-bandwidth transducer architectures, specifically optimized for low-frequency trans-skeletal propagation, with computational frameworks that preserve and exploit waveform -level information. By moving beyond simplified ray-tracing and accounting for phase coherence, multi-path interference, and mode conversion, the present invention transforms these complex wave effects from a source of image degradation into informative signal components. This integrated integration of dedicated sensor hardware and physics-informed inversion enables high-resolution intracranial characterization through an intact skull, providing a reliable diagnostic tool at the point of care.

[0027] The present invention relates generally to systems and methods for non-invasive acoustic interrogation of the cranial cavity. In particular, the invention relates to the use of low-frequency acoustic signals and high-performance computing (HPC) for high-fidelity waveform analysis to characterize the cranial environment, including the skull and internal structures, and to enable continuous monitoring of physiological variations through an intact skull across various clinical and point-of-care settings.

[0030]

[0028] Accordingly, in a first aspect, the present invention provides an ultrasonic transducer array configured for non-invasive low-frequency acoustic interrogation of tissue within a cranial cavity through an intact skull, comprising a plurality of ultrasonic transducer elements, wherein: (a) each ultrasonic transducer element comprises a layered acoustic structure including at least one electro-acoustic transduction layer and at least one acoustic matching layer, the layered acoustic structure being configured to maintain phase coherence and temporal fidelity of acoustic waves at a center frequency below 1 MHz during propagation through the intact skull; and (b) the plurality of ultrasonic transducer elements is arranged: (i) in a spatial geometry providing angularly diverse acoustic propagation paths through the skull; and (ii) with an inter-element spacing configured to provide spatial sampling of propagated acoustic wavefields sufficient to resolve spatial variations in acoustic-wave propagation caused by the skull.

[0031]

[0029] In specific embodiments of the above ultrasonic transducer array, each ultrasonic transducer element of the plurality of ultrasonic transducer elements serves both as a transducer and a receiver, and is configured to both emit an acoustic wave at a center frequency below 1 MHz as well as receive such a wave transduced by another ultrasonic transducer element.

[0032]

[0030] In specific embodiments of the ultrasonic transducer array according to any of the embodiments above, the effective center-to-center spacing between adjacent sampling locations is configured to be less than about 15 mm, less than about 14 mm, 13 mm, 12 mm, 11 mm, 10 mm or less, or between about 6 mm and about 9 mm, or less than 6 mm.

[0033]

[0031] In further or alternative specific embodiments, the above ultrasonic transducer array comprises more than 50 ultrasonic transducer elements, such as more than 100, 250, 500, 750, 1000 ultrasonic transducer elements, or more.

[0034]

[0032] In certain embodiments of the ultrasonic transducer array according to any of the embodiments above, the ultrasonic transducer elements have a miniaturized physical footprint designed to maintain a sampling density within the constrained surface area of a humancranium, thereby enabling the resolution of spatial variations in acoustic-wave propagation caused by the skull.

[0035]

[0033] The term “miniaturized physical footprint” refers to a size that is less than 15 mm, such as less than 10 mm, 9 mm, 8 mm, 7 mm, 6 mm, 5 mm, 4 mm, 3 mm, or less.

[0036]

[0034] In certain embodiments of the ultrasonic transducer array according to any of the embodiments above, the electro-acoustic transduction layer comprises a piezoelectric composite material.

[0037]

[0035] In certain embodiments of the ultrasonic transducer array according to any of the embodiments above, at least one acoustic matching layer comprises a composite acoustic structure configured to provide a graded or specific acoustic impedance matching between the transduction layer and the skull. Optionally, the composite acoustic structure is a 1-3 composite structure.

[0038]

[0036] In certain embodiments of the ultrasonic transducer array according to any of the embodiments above, each ultrasonic transducer element further comprises a backing layer having an acoustic impedance configured to substantially match the impedance of the transduction layer to reduce transducer ring-down, and wherein the backing layer possesses high internal acoustic attenuation configured to suppress rear-face reflections.

[0039]

[0037] In some embodiments, at least a subset of the transducer elements further comprises an acoustic lens configured to broaden the angular response of the element. The acoustic lens is optionally disposed on the distal emitting surface of the element. By broadening the angular response and increasing the effective acceptance angle, the acoustic lens facilitates the capture of multi-path acoustic signals from a broader range of incident angles within the anatomical region, thereby enhancing the data-driven reconstruction process.

[0040]

[0038] In certain embodiments, the ultrasonic transducer array according to any of the embodiments above, is configured to preferentially capture transmitted and multi-path acoustic wave components propagating through the skull, including acoustic wave components received by transducer elements positioned on opposing sides of the skull relative to transmitting elements, rather than only reflected acoustic wave components.

[0041]

[0039] In certain embodiments of the ultrasonic transducer array according to any of the embodiments above, the transducer elements are distributed circumferentially around at least a portion of the head to provide angularly diverse acoustic propagation paths through the skull.

[0042]

[0040] In certain embodiments of the ultrasonic transducer array according to any of the embodiments above, the acoustic waves transmitted by a subset of transducer elements arereceived by a different or partially overlapping subset of transducer elements than those used for transmission after propagating through the cranial cavity and skull.

[0043]

[0041] In certain embodiments of the ultrasonic transducer array according to any of the embodiments above, the center frequency is below 1 MHz, such as 0.9, 0.8, 0.7, 0.6, 0.5, 0.4, 0.3, 0.25, 0.2 MHz, or below.

[0044]

[0042] In certain embodiments of the ultrasonic transducer array according to any of the embodiments above, each ultrasonic transducer element is configured to: (i) operate at said low center frequency while maintaining a miniaturized physical footprint and exhibiting a broadband fractional bandwidth of at least 80% measured at -6dB, thereby enabling high-fidelity capture of phase-coherent, waveform-level information during transmission through the skull; and / or (ii) transmit and receive acoustic energy across the entire effective acceptance angle permitted by the skull geometry, such that the angular coverage of the transducer element is optimized to match the maximum acoustic transmission window of the intact skull without total internal reflection.

[0045]

[0043] In certain embodiments of the ultrasonic transducer array according to any of the embodiments above, the plurality of transducer elements is arranged in a high-density configuration to achieve said spatial sampling at or above the Nyquist rate, utilizing the miniaturized physical footprint of each element to maintain said sampling density within the constrained surface area of a human cranium, thereby enabling the resolution of spatial variations in acoustic-wave propagation caused by the skull.

[0046]

[0044] In certain embodiments of the ultrasonic transducer array according to any of the embodiments above, the spatial sampling is achieved by at least one of: (i) a simultaneous acquisition from the plurality of transducer elements, or (ii) a sequential acquisition from a subset of transducer elements translated across a plurality of spatial positions relative to the skull.

[0047]

[0045] In a second aspect, the present invention provides a system for non-invasive low-frequency ultrasound interrogation of tissue within a cranial cavity through an intact skull, comprising: (i) an ultrasonic transducer array configured to generate and receive low-frequency acoustic waves that propagate through the skull; (ii) a data acquisition unit configured to selectively excite one or more subsets of transducer elements and to receive acoustic signals from one or more subsets of transducer elements that are different from or overlapping with the excited subsets; and (iii) a processing unit configured to process received acoustic signalsbased on waveform-level acoustic information preserved during propagation through the skull, and to reconstruct a physical-domain representation via physics-informed computational inversion of acoustic-wave propagation patterns within the cranial cavity, thereby resolving skull-induced distortions.

[0048]

[0046] In some embodiments, the data acquisition unit optionally comprises multiplexing circuitry

[0222] configured to selectively connect the plurality of transducer elements to a smaller number of system channels, such that the number of transducer elements in the array exceeds the number of active acquisition channels. For example, the system may employ a multiplexing ratio of 2:1, 4:1, 8:1, or higher, enabling a high-density spatial sampling of the wavefield while reducing overall system complexity, power consumption, and the number of physical interconnects, thereby enhancing the portability of the system.

[0049]

[0047] In specific embodiments of the above system, the ultrasonic transducer array is the ultrasonic transducer array according to any of the embodiments above.

[0050]

[0048] In certain embodiments of the system according to any of the embodiments above, the processing unit is configured to optionally augment the physics-informed computational inversion using one or more enhancement factors to further enhance spatial resolution; wherein the enhancement factors include, but are not limited to: prior anatomical data, structural estimates, extrinsic contrast agents, and super-resolution methodologies.

[0051]

[0049] In certain embodiments of the system according to any of the embodiments above, the data communication between the transducer array and the processing unit is optionally performed via a high -bandwidth communication interface, which may include a wireless communication interface providing a cable-free configuration, and / or a high-speed wired interface (e.g., USB, Thunderbolt, Ethernet, or optical fiber).

[0052]

[0050] In certain embodiments of the system according to any of the embodiments above, the processing unit is configured to: (i) process broadband acoustic signals exhibiting a fractional bandwidth of at least 80% to resolve skull-induced distortions and distinguish between direct-path and multi-path acoustic components; (ii) operate on broadband waveform-level acoustic information preserved during propagation through the skull, including at least one of phase information, temporal information, spectral content, or multi-path characteristics of the received acoustic signals; (iii) generate one or more outputs indicative of spatial variations in acoustic-wave propagation within the cranial cavity, the outputs optionally comprising a visual representation or image derived from the determined acoustic-wave propagation pattern; (iv) support repeated or continuous acquisition and processing of acoustic signals over time, and toanalyze temporal variations, high-precision phase shifts, and / or frequency shifts in the received acoustic signals, such that the generated outputs are indicative of dynamic changes within the cranial cavity. In certain embodiments, the analysis of these temporal variations further comprises extracting Doppler shift information from the received acoustic signals to monitor hemodynamic changes or blood flow velocity within the cranial cavity; (v) derive one or more physiological metrics based on variations in acoustic-wave propagation within the cranial cavity, the physiological metrics optionally including indicators related to intracranial pressure, compliance, or changes in intracranial fluid or tissue properties; or any combination thereof; and / or (vi) perform the physics-informed computational inversion in real-time or near realtime, such that the spatial representation is generated during the acoustic interrogation procedure.

[0053]

[0051] In certain embodiments of the system according to any of the embodiments above, the ultrasonic transducer array is integrated into a conformable support structure configured to be worn on a patient’s head.

[0054]

[0052] In certain embodiments of the system according to any of the embodiments above, the conforming article and the plurality of ultrasonic transducers are configured to maintain a stable acoustic coupling with the patient’s head during repeated or prolonged measurements, such that waveform-level acoustic information is preserved over time.

[0055]

[0053] In certain embodiments of the system according to any of the embodiments above, the processing unit comprises a distributed computing architecture including at least one high-performance computing (HPC) resource, wherein said HPC resource is disposed within the local processing unit, within a remote cloud-based computing environment, or distributed between both. In specific embodiments thereof, the HPC resource comprises a Graphics Processing Unit (GPU) array configured to provide an aggregate computing capacity dedicated to at least one of: signal conditioning, real-time mapping, or physical-domain reconstruction, and / or the remote cloud-based computing resource comprises a scalable engine configured to provide a computing capacity on the order of at least 1 PFLOP / s, or a sufficient computational throughput to enable high-resolution three-dimensional (3D) physical-domain reconstruction of the intracranial environment.

[0056]

[0054] In a third aspect, the present invention provides a method for performing non-invasive low-frequency ultrasound interrogation of tissue within a cranial cavity through an intact skull, in which skull-induced distortion significantly affects acoustic-wave propagation, the methodcomprising: (i) transmitting excitation pulses from one or more selected transmitting elements of a multi-element ultrasonic transducer array having a plurality of elements spatially distributed around at least a portion of the patient’s skull, the excitation pulses being transmitted through an acoustic coupling medium and configured to generate low-frequency acoustic waves that propagate through the skull along multiple distinct propagation paths; (ii) receiving the resulting propagated acoustic waves at a subset of spatially distributed transducer elements of the multi-element ultrasonic transducer array, the subset including one or more elements that are different from, overlapping with, or identical to the selected transmitting elements, to capture spatially and temporally diverse waveforms associated with the multiple propagation paths; (iii) extracting, from the received propagated acoustic waves, one or more wave-propagation characteristics, including temporal, spectral, phase, and / or amplitude-related information that preserves propagation effects induced by the skull, the extracted characteristics being provided as input to a physics-informed computational inversion process; (iv) determining, based on the extracted wave-propagation characteristics, a physically consistent representation of acoustic wave propagation through heterogeneous cranial structures by applying said physics-informed computational inversion that transforms the extracted wave-propagation characteristics into a spatially resolved physical-domain representation accounting for skull-induced distortion effects; and (v) generating one or more outputs based on the determined physically consistent representation of acoustic wave propagation, the outputs comprising a spatial representation and / or one or more derived physiological metrics indicative of structural or functional heterogeneity within the cranial region, and further indicative of temporal changes in such heterogeneity based on repeated or comparative determinations over time.

[0057]

[0055] In certain embodiments of the above method, the transmitting the excitation pulses and receiving the propagated acoustic waves, are performed using a synthetic aperture configuration in which subsets of transducer elements of the multi-element ultrasonic transducer array are selectively activated across multiple transmission events to sample multiple distinct acoustic-wave propagation paths through the skull.

[0058]

[0056] In certain embodiments of the method according to any of the embodiments above, the extracted wave-propagation characteristics: (i) preserve phase-coherent, time-resolved acoustic waveform information associated with the propagated acoustic waves, rather than envelope-only or intensity-only measurements; and / or (ii) include effects arising from multiple propagation paths and wave interference produced as the acoustic waves traverseheterogeneous cranial structures, and wherein such multi-path and interference effects are utilized as informative components in determining the physically consistent representation of acoustic-wave propagation.

[0059]

[0057] In certain embodiments of the method according to any of the embodiments above, the low-frequency acoustic waves have a center frequency below 1 MHz, such as 0.9 MHz, 0.8 MHz, 0.7 MHz, 0.6 MHz, 0.5 MHz, 0.4 MHz, 0.3 MHz, 0.25 MHz, 0.2 MHz, or below.

[0060]

[0058] In certain embodiments of the method according to any of the embodiments above, the one or more outputs are generated based on repeated determinations of the physically consistent representation of acoustic-wave propagation over time, such that the outputs are indicative of temporal changes within the cranial region associated with clinical conditions, including, but not limited to, stroke (including ischemic and hemorrhagic subtypes), traumatic brain injury (TBI), or cerebral edema (vasogenic and cytotoxic), intracranial pressure (ICP) variations, midline shift, cerebral herniation, hydrocephalus (communicating and non-communicating), intracranial hemorrhage (including subarachnoid, subdural, epidural, and intraventricular hematomas), cerebral vasospasm, hypoxic-ischemic encephalopathy (HIE), neuroinflammation, encephalitis, meningitis, cerebral abscesses, intracranial cysts, cerebral perfusion fluctuations, arteriovenous malformations (AVM), cerebral venous sinus thrombosis (CVST), hypoxic insult, anoxic brain injury, brain atrophy associated with neurodegenerative progression, intracranial hypertension, intracranial hypotension, or the presence, localization, and progression of space-occupying lesions such as primary or metastatic tumors

[0061]

[0059] In certain embodiments of the method according to any of the embodiments above, the spatial representation is rendered as a visual image representing the determined physically consistent representation of acoustic-wave propagation, and wherein the visual image is optionally updated over time to reflect temporal changes within the cranial region.

[0062]

[0060] In certain embodiments of the method according to any of the embodiments above, the acoustic coupling medium is provided as part of a wearable interface configured to maintain consistent acoustic coupling during repeated measurements.

[0063]

[0061] In certain embodiments of the method according to any of the embodiments above, the transmitting and receiving are performed through an acoustic coupling interface between the transducer elements and the patient’s head.

[0064]

[0062] In certain embodiments of the method according to any of the embodiments above, the extracted wave-propagation characteristics include contributions from one or more acousticwave modes propagating within the skull, including longitudinal waves, shear waves, and surface acoustic waves.

[0065]

[0063] In certain embodiments, the method according to any of the embodiments above, further comprises a step of performing one or more pre-processing operations on the received acoustic signals prior to extracting waveform-level acoustic information.

[0066]

[0064] In certain embodiments of the method according to any of the embodiments above, the physical-domain representation corresponds to at least one physical parameter of the cranial region, including acoustic impedance, sound velocity, attenuation, or a related physical property.

[0067]

[0065] In certain embodiments of the method according to any of the embodiments above, the physically consistent representation: (i) comprises applying a data-driven or machine-learned computational inversion model trained on waveform-level acoustic data to infer spatial variations within the cranial region corresponding to physical effects of acoustic-wave propagation; and / or (ii) enables resolving spatial variations on the order of the wavelength associated with the center frequency, by exploiting phase-coherent multi-path propagation effects captured in the waveform -level acoustic information.

[0068]

[0066] In certain embodiments, the method according to any of the embodiments above, further comprises any of the following steps:

[0069] (i) implementing a computational guidance mechanism that utilizes predefined physics-based heuristics independent of subject-specific anatomical geometry to generate in-situ computational information, thereby providing a directional bias that functions as a transient numerical accelerator to stabilize convergence and reduce time-to-solution from an uninitialized initial state; or

[0070] (ii) extracting structural constraints in-situ from captured wavefield data to isolate anatomical features and suppress external undesirable artifacts without reliance on external anatomical priors or pre-labeled information, nor any additional measurement,

[0071] or any combination thereof.

[0072]

[0067] In certain embodiments, the method according to any of the embodiments above, is performed using the ultrasonic transducer array and / or system according to any of the embodiments above.

[0073]

[0068] In a fourth aspect, the present invention provides a non-transitory computer-readable medium storing instructions that, when executed by one or more processors of a system fornon-invasive low-frequency acoustic interrogation of tissue within a cranial cavity through an intact skull, cause the system to: (i) receive acoustic signals corresponding to low-frequency acoustic waves propagated through the skull along multiple propagation paths; (ii) extract waveform-level acoustic information from the received acoustic signals, including at least one of temporal information, phase information, spectral content, or multi-path characteristics; (iii) apply a physics-informed computational inversion to the extracted waveform-level acoustic information to reconstruct a physical-domain representation of acoustic-wave propagation through heterogeneous cranial structures based on the extracted waveform-level acoustic information, thereby resolving distortions introduced by the intact skull; and (iv) generate one or more outputs based on the reconstructed physical-domain representation, the outputs comprising a spatial representation and / or one or more derived physiological metrics indicative of structural or functional heterogeneity within the cranial region, and optionally indicative of temporal changes in such heterogeneity over time.

[0074]

[0069] In certain embodiments of the above computer-readable medium, the instructions further cause the system to: (i) generate the outputs based on repeated processing of acoustic signals acquired over time, such that the outputs are indicative of temporal changes within the cranial region; (ii) derive one or more physiological metrics based on the determined acousticwave propagation pattern; (iii) determine the acoustic-wave propagation pattern using a data-driven or machine-learned model trained on waveform-level acoustic data; and / or (iv) perform one or more pre-processing operations on the received acoustic signals prior to extracting the waveform-level acoustic information, the pre-processing operations optionally including at least one of signal conditioning, noise reduction, temporal alignment, amplitude normalization, frequency-domain transformation, or artifact suppression.

[0075]

[0070] In certain embodiments, the computer-readable medium according to any of the embodiments above, is performed using the ultrasonic transducer array and / or system according to any of the embodiments above.

[0076]

[0071] The present invention provides a technological solution to the challenge of acoustic propagation through the heterogeneous structure of the human skull

[0182] , which typically induces significant phase scrambling and multi-path distortion. By employing a specialized transducer configuration and phase-aware, waveform-level signal analysis of low-frequency acoustic waves, the system resolves skull-induced distortions by exploiting, rather than discarding, complex multi-path propagation and interference patterns

[0190] ,

[0072] The system utilizes a distributed computing architecture (as illustrated in Fig. 2 and Fig. 5) to process extracted wave-propagation characteristics, including phase, timing, and spectral characteristics, as input to a physics-informed computational inversion process

[0318] to determine a physically-consistent representation

[0420] of the intracranial environment, which may be rendered as a spatial representation and / or one or more derived physiological metrics. This representation enables the resolution of structural heterogeneities on the order of the wavelength, providing high-fidelity mapping of the cranial interior by accounting for the physical interaction between the acoustic waves and the cranial structures via computational inversion. This process enables the detection of dynamic physiological changes overtime, such as those associated with stroke, traumatic brain injury (TBI), or cerebral edema

[0422] , in realtime or near real-time, independently of the specific algorithmic architecture used for the reconstruction.

[0077]

[0073] Ultrasonic Transducer Element Architecture: Each ultrasonic transducer element

[0116] within the multi-element array

[0114] is configured to maintain high temporal fidelity and phase integrity of acoustic signals at a center frequency below 1 MHz. In specific embodiments, and as shown in the cross-sections of Figs. 3A-3B, the active electro-acoustic transduction layer

[0166] comprises a 1-3 piezocomposite material having an effective acoustic impedance between approximately 8 MRayl and 20 MRayl, thereby reducing impedance mismatch with the heterogeneous cranial structures and improving transmission efficiency.

[0078]

[0074] In some embodiments, the layered acoustic structure of each element

[0116] , including the transduction layer

[0166] , one or more matching layers [162, 164], and a backing layer

[0168] , is engineered through the combined selection of material composition, layer thicknesses, and acoustic impedances to support a miniaturized physical footprint while achieving a broadband fractional bandwidth of at least 80% (measured at about -6 dB). This integrated configuration is optimized for low-frequency operation while providing the sensitivity and axial resolution necessary to capture phase-aware, waveform-level information during propagation through the intact, heterogeneous skull

[0182] ,

[0079]

[0075] In some embodiments, the mechanical and acoustic damping properties of the layered structure are configured to suppress internal reflections, spurious resonances, and parasitic vibrational modes that could otherwise introduce artifacts into the acquired signals. By mitigating these undesired modes, each element

[0116] supports high-fidelity sampling of the intracranial wavefield at or above the spatial Nyquist rate, enabling accurate resolution of skull-induced variations.

[0076] Acoustic Matching Structure: To optimize the transfer of acoustic energy into the heterogeneous structure of the human skull

[0182] , each transducer element

[0116] comprises at least one matching layer [162, 164] having an acoustic impedance between about 2 MRayl and about 6 MRayl. This graded impedance transition is configured to significantly reduce reflections at the transducer-skull interface, thereby facilitating the transmission of low-frequency acoustic waves and increasing the effective fractional bandwidth of the element, e.g., to at least 80%. Crucially, the matching layer is engineered to maintain phase integrity across the active aperture, ensuring that the wide-bandwidth wavefront

[0186] is coupled into the cranial environment without inducing phase distortions that would degrade waveform-level fidelity.

[0080]

[0077] Backing Layer and Signal Purity: On the opposite surface of the transduction layer

[0166] , a backing layer

[0168] is disposed to manage the element's mechanical response and ensure signal purity. The backing layer

[0168] is configured with an acoustic impedance configured to substantially match the impedance of the transduction layer (e.g., within ±50% of the impedance of the 1-3 piezocomposite in some embodiments) and possesses high internal acoustic attenuation. This configuration is specifically utilized to reduce transducer ring-down by rapidly damping residual vibrations and suppressing rear-face reflections. Furthermore, the backing structure and the composite nature of the active layer work in tandem to suppress parasitic lateral modes, ensuring that the transducer’s response is dominated by the primary longitudinal thickness mode. By shortening the pulse duration and eliminating internal resonance artifacts, the backing layer

[0168] preserves the waveform-level temporal fidelity necessary for high-resolution intracranial characterization.

[0081]

[0078] Acceptance Angle and Multi-path Reception: A key physical attribute of the transducer element

[0116] is its broad acceptance angle, which is critical for capturing the complex, non-specular acoustic reflections characteristic of the intracranial environment. The diameter of the transducer is optimized for a wide angular response, preferably exceeding ±30 degrees at the -6dB point.

[0082]

[0079] This broad acceptance angle is essential for the exploitation of multi-path propagation, as shown in Fig. 4B, as it allows each individual element

[0116] to detect acoustic energy arriving from a wide range of incidence angles after scattering within the skull

[0182] and brain tissues

[0184] , Unlike conventional narrow-beam transducers that filter out off-axis signals, the present configuration ensures that diffuse and scattered interference patterns

[0190] arepreserved and captured. This wide-angle reception capability provides the rich, waveform-level data required for the subsequent high-resolution reconstruction of structural heterogeneities.

[0083]

[0080] Spatial Array Configuration and Support Structure: To provide comprehensive coverage of the cranial volume and to resolve the complex wavefield variations induced by the skull, the system is configured to capture acoustic data at a plurality of spatial locations across the cranial surface. In some embodiments, and as shown in Figs. 1A-1B, the multi-element array

[0114] comprises a high-density configuration of approximately 1000 transducer elements

[0116] distributed across the surface sampling manifold. The spatial distribution of these sampling locations- whether achieved through the physical geometry of a static multi -element array, a mechanically repositioned element or subset of elements, or a synthetic-aperture acquisition scheme- is configured to provide a sufficiently dense spatial sampling defined across the three-dimensional surface sampling manifold of the skull. This density is essential for ensuring that multi-path arrivals are captured with sufficient phase coherence to support waveform-level reconstruction, enabling the system to exploit interference patterns

[0190] rather than discard them as noise. In cases of dynamic repositioning or sequential scanning, the system employs precise time-synchronization to maintain this phase coherence across all acquired data points.

[0084]

[0081] In some embodiments, the elements

[0116] are arranged in a helical pattern or a dense conformal grid along a support structure

[0110] , such as a helmet or designated housing. The elliptical or conformal nature of the support structure

[0110] (e.g., a cap, helmet, or wearable frame [HO]) is specifically configured to approximate the local curvature of the skull. Unlike a simple hemispherical structure, which may result in excessive gaps and signal degradation when applied to an elliptical cranium, the present conformal structure

[0110] ensures that each sampling location is maintained within an effective acoustic coupling distance from the skull surface. This minimizes variations in the standoff distance, thereby supporting coupling consistency and temporal alignment of the captured waveforms.

[0085]

[0082] Furthermore, in some embodiments, the effective center-to-center spacing between adjacent sampling locations is configured to be no greater than approximately 15 mm, , less than about 14 mm, 13 mm, 12 mm, 11 mm, 10 mm or less, or between about 6 mm and about 9 mm, or less than 6 mm. This spacing is utilized to satisfy spatial Nyquist criteria for the effective center frequency of the targeted range, ensuring that the intracranial environment is resolved without spatial aliasing. The system is configured to maintain this sampling density independently of the specific mechanical implementation, ensuring that the intracranialrepresentation

[0420] remains physically consistent even when the number of physical sensors is reduced through dynamic repositioning, sequential scanning, or synthetic aperture acquisition.

[0086]

[0083] Data Acquisition and Synthetic Aperture Strategies: To increase the information content extracted from the cranial cavity, the system is configured for multi-channel data acquisition. In some embodiments, the system employs a synthetic aperture configuration (refer to Fig. 6) where a transmitting subset

[0302] of transducer elements is selectively activated as a transmitter to emit an acoustic wavefield

[0306] , while a plurality of other elements, simultaneously or sequentially, function as receivers

[0310] to capture the resulting acoustic energy. In some embodiments, the sequential activation of transmitters combined with phase-synchronized reception

[0308] enables the construction of an effective aperture significantly larger than the physical array.

[0087]

[0084] In some embodiments, the data acquisition unit is configured to capture spatially and temporally diverse waveforms associated with multiple distinct acoustic propagation paths through and within the skull. This includes capturing transmitted, refracted, and multi-path components - specifically signals received by elements positioned on opposing sides of the skull relative to the transmitting subset - thereby utilizing the full angular coverage permitted by the cranial geometry. By leveraging this comprehensive angular data, the system forms or assembles a high-dimensional dataset that accounts for the interactions between various transmitter-receiver combinations, providing the physical basis for resolving intracranial heterogeneities.

[0088]

[0085] Signal Pre-processing and Feature Extraction: Prior to reconstruction, the system performs a sequence of pre-processing operations

[0312] to ensure signal integrity. In some embodiments, and as detailed in the block diagram of Fig. 5, these operations include applying filters configured to compensate for signal distortions and frequency-selective attenuation induced by the skull. Unlike conventional systems that utilize envelope detection, the present system (via ADC

[0236] ) preserves waveform-level acoustic information, including phase and full RF bandwidth. This pre-processing is specifically configured to maintain inter-element phase relationships across the sampling manifold, enabling the extraction of subtle propagation-induced features from multi-path arrivals and diverse wave modes (e.g., longitudinal and shear) as they traverse the cranial structures.

[0089]

[0086] Wave Propagation Modeling and Physical-Domain Reconstruction: Following preprocessing and feature extraction, the system implements a physics-informed computationalframework

[0318] to reconstruct a physical -domain representation

[0420] of the intracranial environment, which may include a discretized spatial map of intrinsic physical parameters (e.g., acoustic impedance, sound velocity, or density). Unlike conventional imaging that relies on simplified time-of-flight approximations, the present framework

[0318] is configured to utilize the waveform-level acoustic information, capturing the interaction of the acoustic wavefield with the intracranial structures. The framework is configured to resolve complex propagation effects, such as refraction and multi-path arrivals

[0190] , by ensuring that the reconstructed representation is physically consistent with the captured wave-propagation characteristics. In some embodiments, this process does not necessarily require a pre-defined initial structural model, a prior anatomical image (e.g., CT or MRI), or a prior estimate of the acoustic velocity field.

[0090]

[0087] In certain embodiments of the method according to any of the embodiments above, the computational inversion process

[0318] is performed without reliance on a prior anatomical model or subject-specific structural imaging data of the cranial cavity. In such embodiments, the reconstruction is derived directly from waveform-level acoustic measurements acquired through the intact skull, without seeding, constraining, or initializing the inversion using preexisting CT, MRI, or any other structural imaging data of the same subject.

[0091]

[0088] In certain embodiments of the method according to any of the embodiments above, the reconstruction of the physical-domain representation is initiated without an initial subjectspecific anatomical model of the skull.

[0092]

[0089] The system achieves a physically consistent reconstruction by utilizing the combined contribution of high-fidelity waveform-level measurements of the acoustic wavefield with physics-informed computational inversion

[0318] , wherein the reconstruction is further characterized by: (i) low-frequency structural information (e.g., reaching a lower bound in the order of 100 kHz) providing penetration and global structural context and ensuring convergence stability by mitigating cycle-skipping; (ii) high fractional bandwidth (at least 80%) supporting inversion stability and enabling high-resolution reconstruction by utilizing the higher spectral components within the bandwidth; and (iii) high-density spatial-temporal sampling enabling the framework to exploit phase-coherence and interference patterns.

[0093]

[0090] To enable this high-fidelity reconstruction within clinical timescales, the processing unit

[0130] utilizes a distributed computing architecture comprising at least one high-performance computing (HPC) resource, which may be disposed within the local processing unit

[0250] , within a remote cloud-based computing environment

[0260] , or distributed betweenboth. In some embodiments, the HPC resource includes an internal GPU array

[0256] dedicated to initial signal conditioning and real-time or near real-time mapping. In some embodiments, for high-resolution three-dimensional (3D) reconstruction, the system interfaces with a remote cloud-based engine

[0260] configured to provide a computing capacity on the order of at least 1 PFLOP / s, or a sufficient computational throughput to enable reconstruction within clinical timescales. This framework exploits phase-coherence and interference patterns, transforming the captured wavefield into a spatially resolved representation

[0320] by accounting for refraction, mode conversion, and multi-path propagation. The aforementioned framework and reconstruction methods are implemented as processor-executable instructions stored on a non-transitory computer-readable medium (CRM).

[0094]

[0091] Multi-Modal Data Integration and Enhanced Resolution: In some embodiments, the physics-informed computational framework

[0318] is further configured to incorporate or be augmented by various enhancement factors to further enhance spatial resolution, reconstruction stability, or clinical context. Such enhancement factors may include, without limitation: prior anatomical knowledge (e.g., from CT or MRI images), initial structural estimates, extrinsic contrast agents, and super-resolution methodologies. The framework is further configured such that said auxiliary data, or the intrinsic exploitation of wavefield interference patterns

[0190] , may optionally be utilized to further enhance spatial resolution. The inclusion of such auxiliary information or resolution-enhancement methods remains compatible with the core inversion process and does not depart from the scope of the invention, providing a modular framework that adapts to diverse clinical needs and multi-modal data availability. In specific embodiments, such auxiliary anatomical information is optional and is not required for initiation or execution of the core reconstruction process described herein.

[0095]

[0092] Distributed Processing and System Synergy: The distributed computing architecture of the processing unit

[0130] is specifically configured to manage and process the massive datasets generated by the high-density transducer array elements

[0116] , The synergy between the local GPU array

[0256] and the remote cloud-based inversion engine

[0136] allows the system to perform physics-informed computational inversion using high-density spatial-temporal sampling that would otherwise be computationally prohibitive. In some embodiments, data communication between the ultrasonic transducer array

[0114] and the processing unit

[0130] is optionally performed via a high -bandwidth wireless communication interface

[0124] (e.g., WiFi, 5G, or other radio-frequency protocols), thereby providing a cable-free interface that facilitates ease of use in clinical settings.

[0093] In certain embodiments, the system is further configured to support high-speed wired communication interfaces [120, 240], including but not limited to USB (e.g., USB-C), Thunderbolt™, Ethernet, or optical fiber links, to ensure maximum data throughput and minimal latency between the acquisition hardware and the processing unit. This architecture supports the extraction of subtle wavefield variations independently of the specific mechanical implementation of the array. By dynamically allocating computational tasks based on current processing loads and connectivity, the system maintains real-time or near real-time updates of the spatial representation during the acoustic interrogation procedure. This ensures that structural or functional heterogeneities are resolved with high precision on the order of the wavelength while the patient is being monitored, providing immediate feedback to the clinical user.

[0096]

[0094] Clinical Monitoring and Diagnostics: In some embodiments, the resulting physicaldomain representation

[0420] , reconstructed via the aforementioned physics-informed computational inversion

[0318] , is utilized to identify and monitor temporal changes within the intracranial environment. By deriving spatial parameter maps or specific physiological indicators from said reconstructed representation, the system is configured to support the detection of clinical conditions and associated findings such as, but not limited to, traumatic brain injury (TBI), ischemic or hemorrhagic stroke

[0412] , cerebral edema

[0422] , or secondary features such as midline shift. Additionally, the representation may provide a basis for the non-invasive assessment of metrics such as intracranial pressure (ICP) or cerebral compliance derived from the reconstructed physical parameters.

[0097]

[0095] Because the inversion-based reconstruction is grounded in measurable physical properties of acoustic-wave propagation and does not require a prior anatomical baseline image of the same patient, it provides an objective longitudinal record based on intrinsic physical parameters, and is therefore independent of operator variability. This ensures that the results are comparable across different sessions and devices, allowing for the serial longitudinal tracking of a patient’s neurological state over time in real-time or near real-time across various clinical settings (e.g., intensive care units or emergency departments).EXAMPLES

[0098]

[0096] To examine the system of the invention, and validate its performance, various multistage experimental protocols will be used.

[0099]

[0097] In a preliminary validation or calibration process, the system will be used to interrogate a heterogeneous phantom representing cranial structures to verify the accuracy of the reconstructed representation and the suppression of skull-induced artifacts. The generated output will then be compared against a ground-truth reference to quantify reconstruction fidelity (see example in Fig. 7A).

[0100]

[0098] In an exemplary subsequent clinical stage, the validated or calibrated system will be used to monitor a subject's cranial cavity for the presence of anomalous regions, such as a hemorrhage, by identifying characteristic acoustic signatures within the reconstructed output (see example in Fig. 7B).

[0101]

[0099] Exemplary steps of the protocol may include, for instance: (i) positioning a wearable acoustic interrogation interface

[0110] on the subject’s head to establish acoustic coupling between the transducer array and the intact skull; (ii) executing a coordinated transmission sequence wherein, in each step of the sequence, at least one transducer element

[0116] emits an acoustic wave while a plurality of other transducer elements

[0116] in the array operate in a receiving mode to capture the resulting wavefields; (iii) repeating the transmission sequence until at least a subset of the transducer elements have acted as emitters, thereby acquiring a comprehensive set of multi-static spatial data; (iv) computationally processing the acquired data to resolve skull-induced distortions; and (v) generating a representation of the cranial cavity and its contents, including a reconstructed representation of the skull, intracranial structures, and physiological parameters.

[0102]

[0100] For purposes of clarity, and in no way limiting the scope of the teachings, unless otherwise indicated, all numbers expressing quantities, percentages or proportions, and other numerical values recited herein, should be interpreted as being preceded in all instances by the term “about”. Accordingly, the numerical parameters recited in the present specification are approximations that may vary depending on the desired outcome. For example, each numerical parameter may be construed in light of the number of reported significant digits and by applying ordinary rounding techniques. The term “about” as used herein means that values of 10% or less above or below the indicated values are also included.

[0101] For purposes of clarity, where ranges were provided using example values, such as for sizes, a range with similar values is included as example.

[0103]

[0102] The words used in this specification to describe the invention and its various embodiments are to be understood not only in the sense of their commonly defined meanings, but to include by special definition in this specification structure, material or acts beyond the scope of the commonly defined meanings. Thus, if an element can be understood in the context of this specification as including more than one meaning, then its use in a claim must be understood as being generic to all possible meanings supported by the specification and by the word itself.

[0104]

[0103] Reference in the specification to “one embodiment”, “an embodiment”, “another embodiment” or "certain embodiments" means that a particular feature, structure, or characteristic described in conjunction with the embodiment can be included in at least one embodiment of the invention. The appearances of the phrase “in one embodiment” in various places in the specification do not necessarily all refer to the same embodiment.

[0105]

[0104] Reference numerals list:

[0106] 100 - system for non-invasive intracranial acoustic interrogation

[0107] 110 - wearable acoustic interrogation interface

[0108] 114 - ultrasonic transducer array

[0109] 116 - individual transducer elements

[0110] 118 - transmitting ultrasonic transducer element

[0111] 120 - signal and / or communication interface

[0112] 122 - cable assembly

[0113] 124 - wireless transceiver (optional)

[0114] 130 - processing unit

[0115] 132 - data acquisition and control module

[0116] 134 - local GPU computing resource

[0117] 136 - cloud-based inversion engine

[0118] 140 - display device

[0119] 142 - visual output

[0120] 162 - outer acoustic matching layer

[0121] 164 - inner acoustic matching layer

[0122] 166 - electro-acoustic transduction layer (e.g., piezoelectric layer)- backing layer

[0123] - element housing

[0124] - electrical connection to signal interface

[0125] - wave propagation

[0126] - skull layer

[0127] - intracranial tissue region

[0128] - initial emitted acoustic wavefront

[0129] - distorted wavefront after skull interaction - scattering / phase aberration effects

[0130] - transducer cables

[0131] - ultrasound front-end subsystem

[0132] - multiplexer (optional)

[0133] - transmit / receive (T / R) switch

[0134] - waveform generator

[0135] - power amplifier / pulser

[0136] - low-noise amplifier (LNA)

[0137] - variable gain amplifier (VGA)

[0138] - programmable gain amplifier (PGA)

[0139] - anal og-to-digi tai converter (ADC)

[0140] - high-speed data interface

[0141] - local processing unit

[0142] - memory (RAM)

[0143] - central processing unit (CPU)

[0144] - graphics processing unit (GPU)

[0145] - remote computing resource

[0146] - display device

[0147] - method for intracranial acoustic interrogation - selection of transmitting elements

[0148] - issuance of excitation commands

[0149] - generation of acoustic wavefields

[0150] - synchronization of reception

[0151] - capture of propagated acoustic signals

[0152] - digitization and buffering— streaming to processing resources

[0153] - determination of acquisition completion

[0154] - physics-informed computational inversion

[0155] - generation of outputs

[0156] - ground truth representation

[0157] - ground-truth pathological feature (e.g., hemorrhage) - reconstructed physical-domain representation

[0158] - reconstructed pathological feature (e.g., hemorrhage)

Claims

CLAIMS:

1. An ultrasonic transducer array configured for non-invasive low-frequency acoustic interrogation of tissue within a cranial cavity through an intact skull, comprising a plurality of ultrasonic transducer elements, wherein:- each ultrasonic transducer element comprises a layered acoustic structure including at least one electro-acoustic transduction layer and at least one acoustic matching layer, the layered acoustic structure being configured to maintain phase coherence and temporal fidelity of acoustic waves at a center frequency below 1 MHz during propagation through the intact skull; and- the plurality of ultrasonic transducer elements is arranged: (i) in a spatial geometry providing angularly diverse acoustic propagation paths through the skull; and (ii) with an inter-element spacing configured to provide spatial sampling of propagated acoustic wavefields sufficient to resolve spatial variations in acoustic-wave propagation caused by the skull.

2. The ultrasonic transducer array of claim 1, wherein the electro-acoustic transduction layer comprises a piezoelectric composite material.

3. The ultrasonic transducer array of claim 1, wherein at least one acoustic matching layer comprises a composite acoustic structure configured to provide a graded or specific acoustic impedance matching between the transduction layer and the skull.

4. The ultrasonic transducer array of claim 3, wherein the composite acoustic structure is a 1-3 composite structure.

5. The ultrasonic transducer array of claim 1, wherein each ultrasonic transducer element further comprises a backing layer having an acoustic impedance configured to substantially match the impedance of the transduction layer to reduce transducer ringdown, and wherein the backing layer possesses high internal acoustic attenuation configured to suppress rear-face reflections.

6. The ultrasonic transducer array of claim 1, wherein at least a subset of the transducer elements further comprises an acoustic lens configured to broaden the angular response of the element.

7. The ultrasonic transducer array of claim 1, wherein the array is configured to preferentially capture transmitted and multi-path acoustic wave components propagating through the skull, including acoustic wave components received by transducer elements positioned on opposing sides of the skull relative to transmitting elements, rather than only reflected acoustic wave components.

8. The ultrasonic transducer array of claim 1, wherein the transducer elements are distributed circumferentially around at least a portion of the head to provide angularly diverse acoustic propagation paths through the skull.

9. The ultrasonic transducer array of claim 1, wherein acoustic waves transmitted by a subset of transducer elements are received by a different or partially overlapping subset of transducer elements than those used for transmission after propagating through the cranial cavity and skull.

10. The ultrasonic transducer array of claim 1, wherein the center frequency is below 1 MHz.

11. The ultrasonic transducer array of claim 1, wherein each ultrasonic transducer element is configured to operate at said low center frequency while maintaining a miniaturized physical footprint and exhibiting a broadband fractional bandwidth of at least 80% measured at -6dB, thereby enabling high-fidelity capture of phase-coherent, waveformlevel information during transmission through the skull.

12. The ultrasonic transducer array of claim 1, wherein each ultrasonic transducer element is configured to transmit and receive acoustic energy across the entire effective acceptance angle permitted by the skull geometry, such that the angular coverage of the transducer element is optimized to match the maximum acoustic transmission window of the intact skull without total internal reflection.

13. The ultrasonic transducer array of claim 1, wherein the plurality of transducer elements is arranged in a high-density configuration to achieve said spatial sampling at or above the Nyquist rate, utilizing a miniaturized physical footprint of each element to maintain said sampling density within the constrained surface area of a human cranium, thereby enabling the resolution of spatial variations in acoustic-wave propagation caused by the skull.

14. The ultrasonic transducer array of claim 1, wherein said spatial sampling is achieved by:(i) a simultaneous acquisition from the plurality of transducer elements, and / or (ii) asequential acquisition from a subset of transducer elements translated across a plurality of spatial positions relative to the skull.

15. A system for non-invasive low-frequency ultrasound interrogation of tissue within a cranial cavity through an intact skull, comprising:i) an ultrasonic transducer array configured to generate and receive low-frequency acoustic waves that propagate through the skull;ii) a data acquisition unit configured to selectively excite one or more subsets of transducer elements and to receive acoustic signals from one or more subsets of transducer elements that are different from or overlapping with the excited subsets; and iii)a processing unit configured to process received acoustic signals based on waveformlevel acoustic information preserved during propagation through the skull, and to reconstruct a physical-domain representation via physics-informed computational inversion of acoustic-wave propagation patterns within the cranial cavity, thereby resolving skull-induced distortions.

16. The system of claim 15, wherein the physics-informed computational inversion is performed without reliance on a prior anatomical model or subject-specific structural imaging data of the cranial cavity.

17. The system of claim 15, wherein the data acquisition unit comprises multiplexing circuitry configured to selectively connect the plurality of transducer elements to a smaller number of system channels, such that the number of transducer elements in the array exceeds the number of active acquisition channels18. The system of claim 15, wherein the ultrasonic transducer array comprises an ultrasonic transducer array as claimed in claim 1.

19. The system of claim 15, wherein the processing unit is configured to optionally augment the physics-informed computational inversion using one or more enhancement factors to further enhance spatial resolution; wherein the enhancement factors include, but are not limited to: prior anatomical data, structural estimates, extrinsic contrast agents, and superresolution methodologies.

20. The system of claim 15, wherein data communication between the transducer array and the processing unit is optionally performed via a high-bandwidth communication interface,which may include a wireless communication interface providing a cable-free configuration, and / or a high-speed wired interface.

21. The system of claim 15, wherein the processing unit is configured to process broadband acoustic signals exhibiting a fractional bandwidth of at least 80% to resolve skull-induced distortions and distinguish between direct-path and multi-path acoustic components.

22. The system of claim 15, wherein the ultrasonic transducer array is integrated into a conformable support structure configured to be worn on a patient’s head.

23. The system of claim 15, wherein the processing unit is configured to operate on broadband waveform-level acoustic information preserved during propagation through the skull, including at least one of phase information, temporal information, spectral content, or multi-path characteristics of the received acoustic signals.

24. The system of claim 15, wherein the processing unit is further configured to generate one or more outputs indicative of spatial variations in acoustic-wave propagation within the cranial cavity, the outputs optionally comprising a visual representation or image derived from the determined acoustic-wave propagation pattern.

25. The system of claim 15, wherein the processing unit is configured to support repeated or continuous acquisition and processing of acoustic signals over time, and to analyze temporal variations, high-precision phase shifts, and / or frequency shifts in the received acoustic signals, such that the generated outputs are indicative of dynamic changes within the cranial cavity.

26. The system of claim 15, wherein the processing unit is further configured to extract Doppler shift information from the received acoustic signals to monitor hemodynamic changes or blood flow velocity within the cranial cavity.

27. The system of claim 15, wherein the processing unit is further configured to derive one or more physiological metrics based on variations in acoustic-wave propagation within the cranial cavity, the physiological metrics optionally including indicators related to intracranial pressure, compliance, or changes in intracranial fluid or tissue properties.

28. The system of claim 15, wherein the conforming article and the plurality of ultrasonic transducers are configured to maintain a stable acoustic coupling with the patient’s headduring repeated or prolonged measurements, such that waveform-level acoustic information is preserved over time.

29. The system of claim 15, wherein the processing unit comprises a distributed computing architecture including at least one high-performance computing (HPC) resource, wherein said HPC resource is disposed within the local processing unit, within a remote cloudbased computing environment, or distributed between both.

30. The system of claim 29, wherein the HPC resource comprises a Graphics Processing Unit (GPU) array configured to provide an aggregate computing capacity dedicated to at least one of: signal conditioning, real-time mapping, or physical-domain reconstruction.

31. The system of claim 29, wherein the remote cloud-based computing resource comprises a scalable engine configured to provide a computing capacity on the order of at least 1 PFLOP / s, or a sufficient computational throughput to enable high-resolution three- dimensional (3D) physical-domain reconstruction of the intracranial environment.

32. The system of claim 15, wherein the processing unit is configured to perform the physics- informed computational inversion in real-time or near real-time, such that the spatial representation is generated during the acoustic interrogation procedure.

33. A method for performing non-invasive low-frequency ultrasound interrogation of tissue within a cranial cavity through an intact skull, in which skull-induced distortion significantly affects acoustic-wave propagation, the method comprising:i) transmitting excitation pulses from one or more selected transmitting elements of a multi-element ultrasonic transducer array having a plurality of elements spatially distributed around at least a portion of the patient’s skull, the excitation pulses being transmitted through an acoustic coupling medium and configured to generate low- frequency acoustic waves that propagate through the skull along multiple distinct propagation paths;ii) receiving the resulting propagated acoustic waves at a subset of spatially distributed transducer elements of the multi-element ultrasonic transducer array, the subset including one or more elements that are different from, overlapping with, or identical to the selected transmitting elements, to capture spatially and temporally diverse waveforms associated with the multiple propagation paths;iii) extracting, from the received propagated acoustic waves, one or more wavepropagation characteristics, including temporal, spectral, phase, and / or amplitude- related information that preserves propagation effects induced by the skull, the extracted characteristics being provided as input to a physics-informed computational inversion process;iv) determining, based on the extracted wave-propagation characteristics, a physically consistent representation of acoustic wave propagation through heterogeneous cranial structures by applying said physics-informed computational inversion that transforms the extracted wave-propagation characteristics into a spatially resolved physicaldomain representation accounting for skull-induced distortion effects; andv) generating one or more outputs based on the determined physically consistent representation of acoustic wave propagation, the outputs comprising a spatial representation and / or one or more derived physiological metrics indicative of structural or functional heterogeneity within the cranial region, and further indicative of temporal changes in such heterogeneity based on repeated or comparative determinations over time.

34. The method of claim 33, wherein reconstruction of the physical-domain representation is initiated without an initial subject-specific anatomical model of the skull.

35. The method of claim 33, wherein transmitting the excitation pulses and receiving the propagated acoustic waves are performed using a synthetic aperture configuration in which subsets of transducer elements of the multi-element ultrasonic transducer array are selectively activated across multiple transmission events to sample multiple distinct acoustic-wave propagation paths through the skull.

36. The method of claim 33, wherein the extracted wave-propagation characteristics preserve phase-coherent, time-resolved acoustic waveform information associated with the propagated acoustic waves, rather than envelope-only or intensity-only measurements.

37. The method of claim 33, wherein the extracted wave-propagation characteristics include effects arising from multiple propagation paths and wave interference produced as the acoustic waves traverse heterogeneous cranial structures, and wherein such multi-path and interference effects are utilized as informative components in determining the physically consistent representation of acoustic-wave propagation.

38. The method of claim 33, wherein the low-frequency acoustic waves have a center frequency below 1 MHz.

39. The method of claim 33, wherein one or more outputs are generated based on repeated determinations of the physically consistent representation of acoustic-wave propagation over time, such that the outputs are indicative of temporal changes within the cranial region associated with clinical conditions, including, but not limited to, stroke, traumatic brain injury (TBI), or cerebral edema.

40. The method of claim 33, wherein the spatial representation is rendered as a visual image representing the determined physically consistent representation of acoustic-wave propagation, and wherein the visual image is optionally updated over time to reflect temporal changes within the cranial region.

41. The method of claim 33, wherein the acoustic coupling medium is provided as part of a wearable interface configured to maintain consistent acoustic coupling during repeated measurements.

42. The method of claim 33, wherein transmitting and receiving are performed through an acoustic coupling interface between the transducer elements and the patient’s head.

43. The method of claim 33, wherein the extracted wave-propagation characteristics include contributions from one or more acoustic wave modes propagating within the skull, including longitudinal waves, shear waves, and surface acoustic waves.

44. The method of claim 33, further comprising performing one or more pre-processing operations on the received acoustic signals prior to extracting waveform-level acoustic information.

45. The method of claim 33, wherein the physical-domain representation corresponds to at least one physical parameter of the cranial region, including acoustic impedance, sound velocity, attenuation, or a related physical property.

46. The method of claim 33, wherein determining the physically consistent representation comprises applying a data-driven or machine-learned computational inversion model trained on waveform-level acoustic data to infer spatial variations within the cranial region corresponding to physical effects of acoustic-wave propagation.

47. The method of claim 33, wherein the physically consistent representation enables resolving spatial variations on the order of the wavelength associated with the center frequency, by exploiting phase-coherent multi-path propagation effects captured in the waveform-level acoustic information.

48. The method of claim 33, which is carried out by the system of claim 13, or a system that comprises the ultrasonic transducer array of claim 1.

49. A non-transitory computer-readable medium storing instructions that, when executed by one or more processors of a system for non-invasive low-frequency acoustic interrogation of tissue within a cranial cavity through an intact skull, cause the system to:i) receive acoustic signals corresponding to low-frequency acoustic waves propagated through the skull along multiple propagation paths;ii) extract waveform-level acoustic information from the received acoustic signals, including at least one of temporal information, phase information, spectral content, or multi-path characteristics;iii)apply a physics-informed computational inversion to the extracted waveform-level acoustic information to reconstruct a physical -domain representation of acoustic-wave propagation through heterogeneous cranial structures based on the extracted waveformlevel acoustic information, thereby resolving distortions introduced by the intact skull; andiv) generate one or more outputs based on the reconstructed physical-domain representation, the outputs comprising a spatial representation and / or one or more derived physiological metrics indicative of structural or functional heterogeneity within the cranial region, and optionally indicative of temporal changes in such heterogeneity over time.

50. The computer-readable medium of claim 49, wherein the instructions further cause the system to generate the outputs based on repeated processing of acoustic signals acquired over time, such that the outputs are indicative of temporal changes within the cranial region.

51. The computer-readable medium of claim 49, wherein the instructions further cause the system to derive one or more physiological metrics based on the determined acoustic-wave propagation pattern.

52. The computer-readable medium of claim 49, wherein the instructions cause the system to determine the acoustic-wave propagation pattern using a data-driven or machine-learned model trained on waveform-level acoustic data.

53. The computer-readable medium of claim 49, wherein the instructions further cause the system to perform one or more pre-processing operations on the received acoustic signals prior to extracting the waveform-level acoustic information, the pre-processing operations optionally including at least one of signal conditioning, noise reduction, temporal alignment, amplitude normalization, frequency-domain transformation, or artifact suppression.

Images