Method and system for tracking tissue motion using covariance and anti-aliasing parameter tools

Abstract

A method of determining a time dependent location of a tissue in response to a stimulation may include receiving ultrasound data reflected from the tissue; calculating a covariance matrix from the ultrasound data as a function of depth between a plurality of pulse periods of the ultrasound data; and calculating one or more feature vectors and related feature values of the covariance matrix, wherein the one or more feature vectors represent a first trace of the tissue location.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of U.S. Provisional Application No. 63 / 500,833, filed May 8, 2023, which is incorporated by reference herein in its entirety. BACKGROUND

[0002] Pneumatic otoscopy is a medical examination that allows determination of the mobility of a patient’s tympanic membrane in response to pressure changes, such as air flow. A healthy tympanic membrane moves in response to pressure. Lack of movement can be due to the presence of fluid in the middle ear. Secretory otitis media is characterized by the presence of fluid beside the tympanic membrane. Thus, pneumatic otoscopy can aid in the diagnosis of secretory otitis media. SUMMARY

[0003] Pneumatic otoscopy using ultrasound can have advantages and challenges. For example, ultrasound typically requires a coupling medium; but filling the ear canal with a coupling medium can not be feasible. However, air-coupled ultrasound can face difficulties. For example, there can typically be a significant mismatch in acoustic impedance between air and the transducer and / or the material under test. The above challenges can result in reduced measurement sensitivity. Furthermore, ultrasound measurements can have a typical low signal-to-noise ratio during or close to the time when the pneumatic stimulus is applied. Thus, determining the position of the tympanic membrane during or close to the time when the pneumatic stimulus is applied can be quite challenging.

[0004] The methods and systems disclosed herein address at least some of the above-mentioned challenges. Autoregression and covariance analysis can be used, alone or in combination, to improve the extraction of tissue motion and position in the region, at least in part, by exploiting periodicity in the received ultrasound signals.

[0005] In one aspect, disclosed herein is a method of determining a time-dependent position of tissue in response to a stimulus, the method comprising: (a) receiving ultrasound data, wherein the ultrasound data is derived from ultrasound waveforms reflected from the tissue; (b) computing a covariance matrix from the ultrasound data from a set of depths and for a plurality of pulse cycles of the ultrasound data; (c) computing one or more eigenvectors and associated eigenvalues of the covariance matrix; and (d) determining a first trace of the position of the tissue associated with content of a dominant eigenvector of the one or more eigenvectors.

[0006] In some embodiments, the method of determining a time-dependent position of tissue in response to a stimulus further comprises analyzing frequency content of the ultrasound data across the plurality of pulse cycles. In some embodiments, the method of determining a time-dependent position of tissue in response to a stimulus further comprises computing a second trace of the position of the tissue based at least in part on the frequency content.

[0007] In some embodiments, the method of determining a time-dependent location of a tissue in response to a stimulus further comprises outputting an indication of a disease state of the tissue in response to the second trace of tissue locations. In some embodiments, the method of determining a time-dependent location of a tissue in response to a stimulus further comprises outputting an indication of a health state of the tissue in response to the second trace of tissue locations. In some embodiments, the method of determining a time-dependent location of a tissue in response to a stimulus further comprises outputting an indication of an indeterminate state of the tissue in response to the second trace of tissue locations.

[0008] In some embodiments, the plurality of pulse cycles are consecutive. In some embodiments, the tissue is a tympanic membrane.

[0009] In some embodiments, the method of determining a time-dependent location of a tissue in response to a stimulus further comprises outputting an indication of a disease state in response to the first trace of tissue locations. In some embodiments, the method of determining a time-dependent location of a tissue in response to a stimulus further comprises outputting an indication of a health state in response to the first trace of tissue locations. In some embodiments, the method of determining a time-dependent location of a tissue in response to a stimulus further comprises outputting an indication of an indeterminate state of the tissue in response to the first trace of tissue locations.

[0010] In some embodiments, each pulse cycle of the plurality of pulse cycles is associated with a correlation covariance matrix from a plurality of correlation covariance matrices. In some embodiments, the method of determining a time-dependent location of a tissue in response to a stimulus further comprises calculating a set of displacement vectors of the plurality of correlation covariance matrices. In some embodiments, the plurality of correlation covariance matrices are associated with pulse cycles of the ultrasound data across consecutive adjacent depths. In some embodiments, the set of displacement vectors of the plurality of correlation covariance matrices are calculated based on a target quality signal ratio. In some embodiments, the target quality signal ratio is a value between 0.00 and 1.00. In some embodiments, the method of determining a time-dependent location of a tissue in response to a stimulus further comprises stitching together segments of the set of displacement vectors to generate a time trace of tissue locations. In some embodiments, the time trace of tissue locations is the first trace.

[0011] In some embodiments, the method of determining the time-dependent position of the tissue in response to the stimulus further comprises analyzing one or more eigenvectors of the covariance matrix. In some embodiments, the one or more eigenvectors each contain an orbital rotation associated with a phase of the ultrasound data. In some embodiments, the method of determining the time-dependent position of the tissue in response to the stimulus further comprises applying a bias to each of the one or more eigenvectors such that the eigenvectors loop around an origin in phase space. In some embodiments, the method of determining the time-dependent position of the tissue in response to the stimulus further comprises calculating a second trace of the tissue position using the resulting biased eigenvectors. In some embodiments, analyzing the one or more eigenvectors further comprises analyzing membrane motion information of the one or more eigenvectors. In some embodiments, analyzing the one or more eigenvectors further comprises repeating the analysis in adjustable 20 ms analysis windows in 10 millisecond (ms) steps. In some embodiments, analyzing the one or more eigenvectors further comprises analyzing a depth range value. In some embodiments, analyzing the one or more eigenvectors further comprises creating a set of principal eigenvectors at each 10 millisecond (ms) step. In some embodiments, analyzing the one or more eigenvectors further comprises creating a set of principal eigenvectors within each 20 ms analysis window.

[0012] In some embodiments, the stimulus is a pneumatic excitation. In some embodiments, the pneumatic excitation is a puff of air.

[0013] In some embodiments, the method of determining the time-dependent position of the tissue in response to the stimulus further comprises analyzing frequency content of the ultrasound data across a plurality of pulse cycles and determining one or more regions in which the first trace contains non-physical tissue positions or movements. In some embodiments, the method of determining the time-dependent position of the tissue in response to the stimulus further comprises analyzing frequency content of the ultrasound data across a plurality of pulse cycles and replacing the first trace with at least a portion of a second trace in the one or more regions. In some embodiments, determining the one or more regions in which the first trace contains non-physical tissue positions or movements comprises utilizing an autoregression. In some embodiments, the autoregression is a third order autoregression. In some embodiments, the autoregression is higher than a third order autoregression.

[0014] In another aspect, disclosed herein is a method of improving accuracy of a determination of a time-dependent position of a tissue in response to a stimulus, the method comprising: applying an autoregression to a signal trace to generate a plurality of poles; excluding or selecting a pole of the plurality of poles using a characteristic of the plurality of poles; and generating an improved signal trace of the tissue position from the remaining or selected poles of the plurality of poles.

[0015] In some embodiments, the autoregression is applied to one or more points of the signal trace. In some embodiments, the autoregression is a third order autoregression. In some embodiments, the autoregression is higher than a third order autoregression. In some embodiments, the order of the autoregression corresponds to the number of separated frequency peaks of the signal trace. In some embodiments, the signal trace comprises a covariance motion detection trace. In some embodiments, the autoregression is applied to one or more eigenvectors of the covariance motion detection trace. In some embodiments, the method of improving accuracy of time-dependent location determination of a tissue in response to a stimulus further comprises using an angle value between each of the plurality of poles and the center of the complex unit circle of the complex plane to estimate a mean frequency of the signal trace. In some embodiments, the angle value is calculated by taking an inverse tangent of a ratio of an imaginary part of the pole to a real part of the pole.

[0016] In some embodiments, the method of improving accuracy of time-dependent location determination of a tissue in response to a stimulus further comprises using a radius value between each of the plurality of poles and the center of the complex unit circle of the complex plane to estimate a variance of a frequency of the signal trace. In some embodiments, the method of improving accuracy of time-dependent location determination of a tissue in response to a stimulus further comprises using a radius value between each of the plurality of poles and the center of the complex unit circle of the complex plane to estimate a velocity represented by the signal trace.

[0017] In some embodiments, the method of improving accuracy of time-dependent location determination of a tissue in response to a stimulus further comprises plotting a velocity trace of the signal trace over time by plotting angles between the plurality of poles and the center of the complex unit circle of the complex plane.

[0018] In some embodiments, the method of improving accuracy of time-dependent location determination of a tissue in response to a stimulus further comprises calculating one or more displacement values between a location of the velocity trace and a location of the signal trace. In some embodiments, the determination of the one or more displacement values comprises integrating a velocity in the velocity trace. In some embodiments, the features of the plurality of poles further comprise a radius, an angle, a velocity, a distance in the complex plane between the plurality of poles, a change in velocity between the plurality of poles, an inertial change in velocity between one or more of the poles, or a momentum of the plurality of poles, or any combination thereof. In some embodiments, the plurality of poles are divided into one or more frequency groups based on frequency.

[0019] In some implementations, generating the optimized signal trace further includes replacing multiple poles of the signal trace with multiple poles from different frequency groups. In some implementations, multiple poles of the signal trace are replaced when the absolute error between one or more of the characteristics of the multiple poles exceeds a threshold. In some implementations, the threshold is a value calculated using one or more of the pole selection algorithms in Table 3. In some implementations, the poles have opposite signs. In some implementations, the threshold is between 0.1 and 5.0. In some cases, the threshold is approximately 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, 5.0, or approximately greater than 5.0. In some implementations, the threshold is π / 3, and the pole signs are opposite.

[0020] On the other hand, this paper discloses a system for determining the time-dependent location of a tissue in response to a stimulus, the system comprising a processor containing executable instructions stored thereon, which are configured, when executed, to: (a) receive ultrasound data, wherein the ultrasound data are derived from ultrasound waves reflected from the tissue; (b) compute a covariance matrix from the ultrasound data from a set of depths and for multiple pulse cycles of the ultrasound data; (c) compute one or more eigenvectors and associated eigenvalues ​​of the covariance matrix; and (d) output a first trace of the tissue location associated with the contents of the principal eigenvector among the one or more eigenvectors.

[0021] In some implementations, the processor is also configured to: (i) analyze the frequency content of ultrasound data across multiple pulse cycles; and (ii) calculate a second trace of tissue location based at least in part on the frequency content.

[0022] In some embodiments, the processor is also configured to output an indication of a disease state in response to a second trace at the tissue location. In some embodiments, the processor is also configured to output an indication of a health state in response to a second trace at the tissue location. In some embodiments, the processor is also configured to output an indication of an undetermined state of the tissue in response to a second trace at the tissue location.

[0023] In some implementations, the system also includes a pneumatic otoscope. In some implementations, the processor is operatively connected to the pneumatic otoscope.

[0024] In some embodiments, the system further includes a capacitively micromachined ultrasonic transducer. In some embodiments, the processor is operatively connected to the capacitively micromachined ultrasonic transducer. In some embodiments, the capacitively micromachined ultrasonic transducer is placed inside an otoscope.

[0025] In some embodiments, multiple pulse cycles are sequential. In some embodiments, the tissue is the tympanic membrane. In some embodiments, the processor is also configured to output an indication of a disease state in response to a first trace of the tissue location. In some embodiments, the processor is also configured to output an indication of a healthy state in response to a first trace of the tissue location. In some embodiments, the processor is also configured to output an indication of an undetermined state of the tissue in response to a first trace of the tissue location.

[0026] In some embodiments, each of the plurality of pulse cycles is associated with a correlation covariance matrix derived from a plurality of correlation covariance matrices. In some embodiments, the system further includes calculating a set of displacement vectors for the plurality of correlation covariance matrices. In some embodiments, the plurality of correlation covariance matrices are associated with pulse cycles of ultrasound data across consecutive adjacent depths. In some embodiments, the set of displacement vectors for the plurality of correlation covariance matrices is calculated based on a target quality signal ratio. In some embodiments, the target quality signal ratio is a value between 0.00 and 1.00. In some embodiments, the processor is also configured to stitch together segments of the displacement vector set to generate a time trace of the tissue location. In some embodiments, the time trace of the tissue location is a first trace.

[0027] In some embodiments, the system's processor is further configured to: (i) analyze one or more eigenvectors of the covariance matrix, wherein each of the one or more eigenvectors contains an orbital rotation associated with the phase of the ultrasound data; (ii) apply a bias to each of the one or more eigenvectors such that the eigenvectors orbit the origin in phase space; and (iii) use the eigenvectors with the resulting bias to calculate a second trace of the tissue location. In some embodiments, the processor is also configured to analyze one or more eigenvectors by analyzing membrane motion information of the one or more eigenvectors.

[0028] In some implementations, the processor is further configured to repeat the analysis in 10 ms steps within an adjustable 20 ms analysis window. In some implementations, the processor is further configured to analyze one or more feature vectors by analyzing depth range values ​​of one or more feature vectors. In some implementations, the processor is further configured to analyze one or more feature vectors by generating a principal feature vector set at each 10 ms step. In some implementations, the processor is further configured to analyze one or more feature vectors by generating a principal feature vector set within each 20 ms analysis window.

[0029] In some implementations, the stimulus is a pneumatic excitation. In some implementations, the pneumatic excitation is a jet of air.

[0030] In some embodiments, the processor is further configured to: (i) determine one or more regions in which the first trace contains a non-physical tissue location or movement; and (ii) replace the first trace with at least a portion of a second trace in one or more regions. In some embodiments, the processor is further configured to analyze the frequency content of the demodulated ultrasound data by applying autoregression. In some embodiments, the autoregression is third-order autoregression. In some embodiments, the autoregression is higher than third-order autoregression.

[0031] On the other hand, this paper discloses a system for improving the accuracy of time-dependent location determination of tissue response to stimuli, the system comprising a processor containing executable instructions stored thereon, which are configured, when executed, to: apply autoregression to a signal trace to generate multiple poles; exclude or select poles among the multiple poles using features of the multiple poles; and generate an improved signal trace of tissue location from the remaining or selected poles among the multiple poles.

[0032] In some embodiments, the processor is further configured to apply autoregression to one or more points of the signal trace. In some embodiments, the autoregression is third-order autoregression. In some embodiments, the autoregression is higher than third-order autoregression. In some embodiments, the order of the autoregression corresponds to the number of separated frequency peaks in the signal trace. In some embodiments, the signal trace includes a covariance motion detection trace. In some embodiments, autoregression is applied to one or more eigenvectors of the covariance motion detection trace.

[0033] In some embodiments, the system's processor is further configured to estimate the average frequency of the signal trace using the angle value between each of the plurality of poles and the center of the complex unit circle in the complex plane. In some embodiments, the processor is further configured to calculate the angle value by taking the arctangent of the ratio of the imaginary part to the real part of the pole. In some embodiments, the processor is further configured to estimate the variance of the signal trace using the radius value between each of the plurality of poles and the center of the complex unit circle in the complex plane. In some embodiments, the processor is further configured to estimate the velocity of the signal trace's frequency using the radius value between each of the plurality of poles and the center of the complex unit circle in the complex plane. In some embodiments, the processor is further configured to plot a velocity trace of the signal trace over time by plotting the angles between the plurality of poles and the center of the complex unit circle in the complex plane. In some embodiments, the processor is further configured to calculate one or more displacement values ​​between the position of the velocity trace and the position of the signal trace. In some embodiments, the processor is further configured to calculate one or more displacement values ​​by integrating the velocity in the velocity trace.

[0034] In some implementations of the system, the characteristics of the multiple poles further include radius, angle, velocity, complex plane distance between the multiple poles, velocity variation between the multiple poles, inertial velocity variation between one or more of the poles, or momentum of the multiple poles, or any combination thereof. In some implementations, the processor is also configured to divide the multiple poles into one or more frequency groups based on frequency. In some implementations, the processor is also configured to generate an optimized signal trace by replacing multiple poles of the signal trace with multiple poles from different frequency groups. In some implementations, the processor is also configured to replace multiple poles of the signal trace when the absolute error between one or more of the characteristics of the multiple poles is higher than a threshold. In some implementations, the threshold is a value calculated using one or more of the pole selection algorithms in Table 3. In some implementations, the signs of the poles are reversed. In some implementations, the threshold is between 0.1 and 5.0. In some cases, the threshold is about 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, 5.0, or about 5.0 or more. In some implementations, the threshold is π / 3, and the pole signs are reversed.

[0035] On the other hand, this paper discloses a system for improving time-dependent location determination, comprising: a covariance analysis model containing executable instructions configured, when executed by a processor, to: (a) derive one or more covariance matrices from ultrasound data, (b) compute one or more eigenvectors of the one or more covariance matrices, and (c) generate traces from the one or more eigenvectors; one or more autoregressive models configured to generate multiple poles; an anti-aliasing model containing executable instructions configured, when executed by a processor, to select one or more regions in the trace that deviate from the phase trajectory; and a signal trace model containing executable instructions configured, when executed by a processor, to: (a) derive one or more covariance matrices from ultrasound data, (b) compute one or more eigenvectors of the one or more covariance matrices, and (c) generate traces from the one or more eigenvectors; one or more autoregressive models configured to generate multiple poles; an anti-aliasing model containing executable instructions configured, when executed by a processor, to: (a) derive one or more covariance matrices from ultrasound data, (b) compute one or more eigenvectors of the one or more covariance matrices, and (c) generate traces from the one or more eigenvectors; and (d) generate traces from the one or more eigenvectors; and (e) generate traces from the one or more eigenvectors; ... These executable instructions are configured to: (d) exclude one or more poles in one or more regions of a trace selected by an anti-aliasing model, wherein the signal trace model is configured to exclude the one or more poles at least in part based on the influence of one or more features of the one or more poles; and (e) select one or more alternative poles of one or more candidate signal traces to replace the one or more poles excluded in (d), wherein the signal trace model is configured to select one or more alternative poles at least in part based on the influence of one or more features of the one or more alternative poles; and an output generator configured to generate an improved signal trace of the tissue location from the one or more alternative poles output by the anti-aliasing model.

[0036] In some implementations, one or more features include the radius of one or more poles to the center of the complex unit circle, the angle between one or more poles and the center of the complex unit circle, the angular velocity between one or more poles, the complex plane distance between one or more poles, the velocity variation between one or more poles, the inertial velocity variation between one or more poles, or the momentum of one or more poles, or any combination thereof.

[0037] Additional aspects and advantages of this disclosure will become readily apparent to those skilled in the art from the following detailed description, which only shows and describes illustrative embodiments of the disclosure. As will be appreciated, this disclosure can have other different embodiments, and certain details can be modified in many obvious ways without departing from this disclosure. Therefore, the drawings and descriptions should be considered illustrative rather than limiting. Incorporation

[0038] All publications, patents, and patent applications mentioned in this specification are incorporated herein by reference to the extent that each individual publication, patent, or patent application is specifically and individually indicated to be incorporated by reference. To the extent that any publication, patent, or patent application incorporated by reference contradicts the disclosure contained in this specification, the specification is intended to supersede and / or take precedence over any such contradictory material. Attached Figure Description

[0039] The novel features of the invention are set forth in the appended claims. A better understanding of the features and advantages of the invention will be obtained by referring to the following detailed description of illustrative embodiments utilizing the principles of the invention, along with the accompanying drawings (also referred to herein as “Figures”): Figure 1 A schematic diagram of the propagation of a burst of pulsed ultrasound is shown.

[0040] Figure 2 Examples of ultrasound data spectral analysis of arterial blood flow are shown. Subfigure (A) shows an example of echo amplitude from a moving target (blood cells) versus depth on the vertical axis and time on the horizontal axis. Subfigure (B) shows ultrasound data spectral analysis of blood flow in the proximal right middle cerebral artery (RMCA), where the right vertical axis represents blood flow velocity in cm / s and the horizontal axis represents time, with a total duration of 4 seconds. Subfigure (C) shows the ultrasound beam intersecting with tissue, simultaneously intersecting the right middle cerebral artery (RMCA), the right anterior cerebral artery (RACA), and the left anterior cerebral artery (LACA).

[0041] Figure 3The following diagram shows ultrasound radiofrequency data from the tympanic membrane. Subplot (A) shows the radiofrequency (RF) echo data from the tympanic membrane (in the human body), where a vertical scale depth of 0 mm corresponds to a distance of 8.5 mm from the transducer, resulting in a TM signal at a depth of approximately 13 to 15 mm. Subplot (B) shows a magnified view of subplot (A) within the area indicated by the arrow. Subplot (C) shows a magnified view of subplot (B) within the area indicated by the arrow.

[0042] Figure 4 It shows Figure 3 An enlarged view of the subplot (C), which highlights three regions and plots three sets of horizontal lines, each about 20 ms long.

[0043] Figure 5 The subgraphs (a), (b), and (c) show the... Figure 4 Regions 1, 2, and 3 in the diagram are deconstructed into a set of 80 overlapping phase trajectories, as shown in subgraphs (d), (e), and (f). Figure 4 Signal amplitudes in regions 1, 2, and 3.

[0044] Figure 6 Examples of non-physical phase behavior that may occur during the tracking of the tympanic membrane are shown, along with instances of unpacking and correcting them.

[0045] Figure 7 This illustrates a practical example of discontinuities generated using blood flow data in covariance techniques.

[0046] Figure 8 Subplot (a) shows the vector set D containing positional changes during each covariance calculation time interval. Subplot (b) shows the curves from subplot (a) magnified and presents the fit between these curves in their overlapping areas. Subplot (c) shows the segment of subplot (b) spliced ​​together to reveal the tympanic membrane motion.

[0047] Figure 9 The phase distribution is shown plotted against the background of the raw RF ultrasound data.

[0048] Figure 10 An example of the quality metric M calculated over windows 20 ms wide, with these windows spaced in 10 ms increments (50% overlap), for a total duration of 1 second is shown.

[0049] Figure 11 It shows following a similar pattern Figure 9 An enlarged example of the covariance trace of the RF phase trajectory of the TM.

[0050] Figure 12It is shown that, except for two points (time ≈ 0.18 s and time ≈ 0.62), the covariance trace remains faithful to the phase trajectory RF data.

[0051] Figure 13 An example of using a second-order AR model to estimate the frequency spectrum of a signal is shown.

[0052] Figure 14 An example of using a first-order AR model to estimate the frequency spectrum of a signal is shown.

[0053] Figure 15 This illustrates how a third-order AR model can be applied to... Figure 11 and Figure 12 The first eigenvector of the covariance shown in the original text then produces the three poles.

[0054] Figure 16 The pole angles over time are shown in both velocity and radians. The pole angles are calculated by taking the arctangent of the ratio of the imaginary part of a given pole to the real part of a given pole.

[0055] Figure 17 A graph showing the pole velocities and positions for all pulse cycles is provided. Once the velocities for all pulse cycles are calculated, they can be integrated over time to obtain the displacement of TM.

[0056] Figure 18 The comparison between the covariance method and the pole 1 of the third-order AR model is shown.

[0057] Figure 19 A visual overview of the process of adjusting the covariance trace based on AR analysis is shown.

[0058] Figure 20 An example implementation of the anti-aliasing (hybrid) algorithm is shown.

[0059] Figure 21 A histogram showing the absolute values ​​of the error between AR velocity and covariance velocity is presented.

[0060] Figure 22 The results of the anti-aliasing procedure are shown.

[0061] Figure 23A , Figure 23B and Figure 23C The diagram shows regions where the trace of pole 1 exhibits discontinuities across three different pulse periods: 251-500, 751-1000, and 1001-1250.

[0062] Figure 24A , Figure 24B , Figure 24C , Figure 24D and Figure 24EFive examples of implementing the AR pole selection algorithm described above are shown.

[0063] Figure 25 A computer system is shown that is programmed or otherwise configured to implement the methods provided herein.

[0064] Figure 26 It is a flowchart of an example method for determining the time-dependent location of an organization's response to a stimulus, based on some implementation schemes.

[0065] Figure 27 This is a flowchart of an example method for improving the accuracy of time-dependent location determination of an organization's response to stimuli, based on some implementation schemes.

[0066] Figure 28 This is a flowchart of an example system for improving time-dependent location determination, based on some implementation schemes.

[0067] Figure 29A and Figure 29B Side and front sectional views of an otoscope placed inside the ear according to some embodiments are shown respectively.

[0068] Figure 30A A side sectional view of a peephole according to some embodiments is shown, and Figure 30B A front sectional view of the peephole tip is shown according to some embodiments.

[0069] Figure 31 This is a flowchart of an example method for determining the time-dependent location of an organization's response to stimulation by an otoscope, based on some implementation schemes. Detailed Implementation

[0070] Although various embodiments of the invention have been shown and described herein, it will be apparent to those skilled in the art that these embodiments are provided by way of example only. Many variations, modifications, and substitutions will occur to those skilled in the art without departing from the invention. It should be understood that various alternatives to the embodiments of the invention described herein may be employed.

[0071] This paper discloses a method and system for air-coupled ultrasound of the tympanic membrane. The method and system disclosed herein can improve the determination of the tympanic membrane position during or near the application of pneumatic stimulation. The transducer and ultrasound system are disclosed. Ultrasound signals can be measured from the inner end of the external auditory canal. Ultrasound data can improve the diagnosis of ear infections and systematize the collection and analysis of diagnostic information for assessing ear infections.

[0072] The methods and systems disclosed herein offer at least some of the following advantages: No or virtually no tissue clutter removal, where tissue is the source of the signal of interest rather than a confounding factor. The target is reached via air in the external auditory canal, rather than through the use of an ultrasound coupling gel or by penetrating intermediate tissue. The pulsed ultrasound transmitter / sensor can be a capacitively micromachined ultrasound transducer (CMUT).

[0073] The methods and systems disclosed herein can be used in combination with, for example, apparatus and methods for characterizing ductile films, surfaces, and subsurface layers, such as those described in commonly owned U.S. Patent No. 7,771,356 and U.S. Patent Publications Nos. 2019 / 0365292, 2018 / 0310917, and 2017 / 0014053, each of which is incorporated herein by reference in its entirety. Furthermore, the machine learning methods and systems disclosed in commonly owned U.S. Patent Publication No. 2020 / 0286227, which is also incorporated herein by reference in its entirety, can also be used in combination with the methods and systems disclosed herein.

[0074] The methods and systems disclosed herein can be used in combination with, for example, the air-coupled capacitive micromachining ultrasonic transducer and its use and manufacturing method disclosed in commonly owned U.S. Patent Publication No. 2021 / 0145406, which is incorporated herein by reference in its entirety.

[0075] The methods and systems disclosed herein can be used in combination with, for example, optical illumination transmission devices and methods, such as those described in commonly owned U.S. Patent Publication No. 2020 / 0107813, which is incorporated herein by reference in its entirety.

[0076] The methods and systems disclosed herein can be used to characterize a variety of biological tissues to provide a range of diagnostic information. Biological tissues may include patient organs. A speculum can be placed within a body cavity to characterize patient tissues. For example, patient organs or body cavities may include: muscles, tendons, ligaments, mouth, tongue, pharynx, esophagus, stomach, intestines, anus, liver, gallbladder, pancreas, nose, larynx, trachea, lungs, kidneys, bladder, urethra, uterus, vagina, ovaries, testes, prostate, heart, arteries, veins, spleen, glands, brain, spinal cord, nerves, etc.

[0077] The methods and systems disclosed herein can be used to characterize the tympanic membrane. For example, the membrane can be characterized to determine ear conditions such as acute otitis media (AOM). Characterization of an ear exhibiting AOM can include detecting the presence of effusion and characterizing the effusion type as serous, mucoid, purulent, or a combination thereof. In AOM, middle ear effusion (MEE) can be caused by infectious agents and can be thin or serous in viral infections, while becoming thicker and purulent in bacterial infections. Therefore, determining the various properties of the fluid near the tympanic membrane can provide information that can be used to characterize the membrane.

[0078] Whenever the terms “at least,” “greater than,” or “greater than or equal to” appear before the first value in a series of two or more values, the terms “at least,” “greater than,” or “greater than or equal to” apply to each value in the series. For example, greater than or equal to 1, 2, or 3 is equivalent to greater than or equal to 1, greater than or equal to 2, or greater than or equal to 3.

[0079] Whenever the terms “not exceeding,” “less than,” or “less than or equal to” appear before the first value in a series of two or more values, the terms “not exceeding,” “less than,” or “less than or equal to” apply to each value in the series. For example, less than or equal to 3, 2, or 1 is equivalent to less than or equal to 3, less than or equal to 2, or less than or equal to 1.

[0080] Some embodiments of the invention herein consider numerical ranges. When a range exists, it includes the endpoints of the range. Furthermore, each subrange and value within a range exists as explicitly listed. The terms “about” or “approximately” can indicate within an acceptable margin of error for a particular value, which will depend in part on how the value is measured or determined (e.g., limitations of the measurement system). For example, “about” can, according to art practice, mean within one or more standard deviations. Alternatively, “about” can indicate a range of up to 20%, up to 10%, up to 5%, or up to 1% of a given value. Unless otherwise stated, when describing a particular value in this application and claims, the term “about” may be assumed to indicate within an acceptable margin of error for that particular value.

[0081] Example blood flow ultrasound Ultrasound can be used in medical applications to understand hemodynamics and / or investigate pathological vascular conditions that cause blood flow behavior to deviate from the expected range within normal limits. One example is detecting blood flow velocities above normal levels in arteries. Higher-than-normal flow velocities can indicate vascular narrowing caused by disease, which reduces the effective blood flow area. Higher velocities within blood vessels can be a warning sign of an impending crisis, such as a heart attack or stroke. Furthermore, higher velocities in blood vessels can provide a basis for diagnosis and patient treatment management.

[0082] In an example method for processing pulsed ultrasound measurements of blood flow signals, the RF echo from each pulse cycle can be recombinantly demodulated so that the reflected signal characteristics can be observed as a DC-centered "baseband" signal. In this example, the frequency content across a set of pulse cycles can be proportional to the blood flow velocity.

[0083] Figure 1 and Figure 2 A visual summary of signal processing for pulsed ultrasound demonstrates this property. Figure 1 The diagram schematically depicts the physical propagation of an ultrasound pulse into tissue and its return echo, which represents the reflection at any given time from a resolution cell or gate centered at a specific echo depth. The width of the resolution gate is cΔt / 2, where Δt is the duration of the pulse train emitted by the transducer. This resolution parameter is correlated with the depth range of scatterers that will contribute to the amplitude of the echo received at the transducer at any time after the pulse train leaves the transducer.

[0084] Figure 1 This diagram illustrates the propagation of a burst of ultrasound in pulsed ultrasound. A burst is the building block of pulsed ultrasound. As shown, the leading edge is located at time T0, and the trailing edge is emitted at T1, with a duration Δt = T1 - T0. The positively slanted solid and dashed lines represent the leading and trailing edges of the ultrasound burst at any given fixed time on the horizontal axis, respectively. These lines illustrate that, taking into account the round trip, the resolution of the burst (called the ultrasound data sample volume size) is cΔt / 2. Therefore, at time T2, echoes arrive from a depth range spanning this sample volume.

[0085] The activity depicted in this image is repeated with each pulse cycle of the ultrasound dataset and represents one of thousands of vertical lines in each motion-time (M-mode) display.

[0086] Figure 2 Subplot (a) shows an example of echo amplitude from a moving target (blood cells) versus depth on the vertical axis and time on the horizontal axis. The vertical axis ranges from 25 to 85 mm depth, and the horizontal axis represents a 4-second duration, indicating a power M-mode ultrasound analysis of blood flowing in the middle and anterior cerebral arteries that supply the brain. Figure 2 Subplot (b) shows the ultrasound spectral analysis of blood flow in the proximal RMCA (subplot (a)). In this example, there is a high-pass filter with a cutoff of approximately 7 cm / s in the processing chain. This filter removes reflections from brain tissue, which are approximately 1000 times larger than the blood flow signal. Without this removal of “tissue clutter,” subplot (a) might not have been possible, as the brain tissue signal could dominate the analysis of the average blood flow velocity at a given time.Figure 2 Subfigure (c) depicts the main axis of the ultrasound beam, showing simultaneous intersections with the right middle cerebral artery (RMCA), right anterior cerebral artery (RACA), and left anterior cerebral artery (LACA). These intersection areas exhibit echoes from blood flow, such as... Figure 2 The subplot (a) is shown. The blood flow velocity in subplot (a) was calculated using autocorrelation techniques, and differs from the location calculations described below using covariance and autoregression.

[0087] Ultrasound measurement of the tympanic membrane Figure 3 The ultrasound radiofrequency data from the tympanic membrane is displayed. Figure 3 Subplot (A) shows RF echo data from the tympanic membrane (in the human body), where a vertical scale depth of 0 mm corresponds to a distance of 8.5 mm from the transducer, making the TM signal at a depth of approximately 13 to 15 mm. The ultrasound pulse has 21 cycles. Note that the bottom of the image is closest to the transducer. Figure 3 The image between the two horizontal lines in subgraph (A) is magnified to generate Figure 3 The subgraph (B), and Figure 3 The image between the two vertical lines in subgraph (B) is further magnified to generate Figure 3 The subgraph (C). Figure 3 The lack of RF data in the horizontal region around 80 ms in subplot (C) is due to reduced reflected energy from the TM. This reduced energy makes TM location tracking more difficult. Therefore, since the signal is reduced in this region, it can be useful to consider all reflected data when determining the TM location. The covariance and autoregressive techniques described in this paper at least partially address this signal region.

[0088] Figure 4 Showing Figure 3 An enlarged view of subplot (C). As shown, three sets of horizontal lines are plotted, each approximately 20 ms long. These three sets are placed in different non-overlapping regions of the TM echo reflection. Each set of 10 lines is an example of sampling the complex RF echo (via Hilbert transform) and determining the covariance matrix from the summation outer product of each of the ten complex echo vectors.

[0089] Figure 5 The following are shown in subgraphs (a), (b), and (c) respectively. Figure 4 Regions 1, 2, and 3 in the diagram are deconstructed into a set of 80 overlapping phase trajectories. Each trajectory consists of values ​​from each of M pulse periods (e.g., M=100 at a pulse repetition frequency of 5 kHz, or a time span of 100 / 5000=20 ms). Within this set of 80 trajectories, there is one trajectory for each gate depth or a / d sample increment.

[0090] Figure 5 Subplots (a), (b), and (c) overlay 80 phase trajectories and show that the phases of the data source (RF echo) are highly consistent.

[0091] Figure 5 Subgraphs (d), (e), and (f) show Figure 4 The signal amplitudes in regions 1, 2, and 3. Figure 5 In subgraph (b), anomalous jaggedness exists in the phase evolution. This can be observed through attention... Figure 5 The signal energy levels in subgraphs (d), (e), and (f) further illustrate this point. Figure 5 In the jagged region of subfigure (b), the energy drops significantly by 15 to 20 dB. This jaggedness in the phase spectrum can be challenging if the unwrapped phase from the Hilbert transform is used to extract the motion dynamics of the scattering target (in this case, the tympanic membrane) without other corrections.

[0092] Method for determining time-dependent position of tissue in response to a stimulus This article discloses methods and systems for determining the time-dependent location of a tissue's response to a stimulus. The tissue can be biological tissue. The tissue can be plant, food, animal, or human. The tissue can be human tissue, such as an organ. The tissue can be a smooth tissue organ or a musculoskeletal organ. The tissue can be, for example, glands such as the heart, liver, spleen, ear organs, gallbladder, pancreas, kidney, bladder, uterus or ovary, prostate, thyroid and parathyroid glands, blood vessels, eye organs, appendix, gastrointestinal tract, or any other organ. The tissue can be the tympanic membrane. The stimulus can be a physical stimulus. The stimulus can be a pressure stimulus, electrical stimulus, electromagnetic stimulus, physical contact stimulus, fluid injection stimulus, or sensory stimulus, or another type of stimulus. The stimulus can be a pressure change stimulus.

[0093] In some cases, the method may include receiving ultrasound data. In other embodiments, the method may include receiving another type of electromagnetic data. The ultrasound data may originate from ultrasound waves reflected from tissue. In other embodiments, the ultrasound data may originate from one or more segments of ultrasound waves reflected from tissue. The method may include calculating a covariance matrix from the ultrasound data. The covariance data may be a function of the depth of multiple pulse cycles of the ultrasound data. The covariance data may be a function of a depth of approximately 2 pulse cycles, 3 pulse cycles, 4 pulse cycles, 5 pulse cycles, approximately 10 pulse cycles, approximately 15 pulse cycles, approximately 20 pulse cycles, approximately 25 pulse cycles, approximately 50 pulse cycles, approximately 100 pulse cycles, approximately 200 pulse cycles, approximately 300 pulse cycles, approximately 400 pulse cycles, approximately 5000 pulse cycles, approximately 1000 pulse cycles, approximately 2000 pulse cycles, approximately 3000 pulse cycles, approximately 4000 pulse cycles, approximately 5000 pulse cycles, or more than approximately 5000 pulse cycles. In some cases, the number of pulse cycles can be between 4500 and 5000. In other cases, the number of pulse cycles can be 4800. The method may also include calculating one or more eigenvectors and associated eigenvalues ​​of the covariance matrix. In some cases, the method may include calculating 1 eigenvector, 2 eigenvectors, 3 eigenvectors, 4 eigenvectors, 5 eigenvectors, 6 eigenvectors, 7 eigenvectors, 8 eigenvectors, 9 eigenvectors, 10 eigenvectors, 11 eigenvectors, 12 eigenvectors, 13 eigenvectors, 14 eigenvectors, 15 eigenvectors, 16 eigenvectors, 17 eigenvectors, 18 eigenvectors, 19 eigenvectors, 20 eigenvectors, 21 eigenvectors, 22 eigenvectors, 23 eigenvectors, 24 eigenvectors, 25 eigenvectors, 26 eigenvectors. The method may include 1, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, or more than 50 feature vectors. The method may also include determining a first trace associated with the content of a principal feature vector among one or more feature vectors.

[0094] In some cases, the method may also include determining the time-dependent location of tissue response to stimulation by analyzing the frequency content of ultrasound data across multiple pulse cycles. In some cases, the frequency content may be a trace. In some cases, the method may also include a second trace that calculates tissue location at least in part based on the frequency content. In some cases, the frequency content may be a measurement of frequency. In some cases, the frequency content may be a change in frequency or a rate of frequency.

[0095] In some embodiments, the method may also include outputting an indication of the disease state of the tissue. In some embodiments, the indication may be responsive to a trace. In some embodiments, the indication may be responsive to a second trace at the tissue location. In other embodiments, the method may also include outputting an indication of the health state of the tissue. In some embodiments, the indication may be responsive to a second trace at the tissue location. In other embodiments, the method may also include outputting an indication of an undetermined state of the tissue. In some embodiments, the indication may be responsive to a second trace at the tissue location. In some embodiments, the indication may be responsive to a first trace at the tissue location. In some embodiments, the indication may be responsive to a third trace at the tissue location.

[0096] In some cases, multiple pulse cycles may be sequential. In other cases, multiple pulse cycles may be non-sequential. In some cases, the tissue may be an organ of the ear, such as the tympanic membrane, ear bones, eustachian tube, cochlea, oval window, or another organ of the ear. In some cases, the tissue may be another human organ, such as the heart, liver, kidneys, stomach, eye, brain, or another organ.

[0097] In some embodiments, the method may also include outputting an indication of the disease state of the tissue. In some embodiments, the indication may be responsive to a trace. In some embodiments, the indication may be responsive to a first trace at the tissue location. In other embodiments, the method may also include outputting an indication of the health state of the tissue. In some embodiments, the indication may be responsive to a first trace at the tissue location. In other embodiments, the method may also include outputting an indication of an undetermined state of the tissue. In some embodiments, the indication may be responsive to a first trace at the tissue location. In some embodiments, the indication may be responsive to a second trace at the tissue location. In some embodiments, the indication may be responsive to a third trace at the tissue location.

[0098] In some cases, each pulse period in a plurality of pulse periods can be associated with a covariance matrix from a plurality of covariance matrices. In some cases, the plurality of covariance matrices can include 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, or 26 covariance matrices. Covariance matrix, 27 covariance matrices, 28 covariance matrices, 29 covariance matrices, 30 covariance matrices, 31 covariance matrices, 32 covariance matrices, 33 covariance matrices, 34 covariance matrices, 35 covariance matrices, 36 covariance matrices, 37 covariance matrices, 38 covariance matrices, 39 covariance matrices, 40 covariance matrices, 41 covariance matrices, 42 covariance matrices, 43 covariance matrices, 44 covariance matrices, 45 covariance matrices, 46 covariance matrices, 47 covariance matrices, 48 ​​covariance matrices, 49 covariance matrices, 50 covariance matrices, or more than 50 covariance matrices.

[0099] In some cases, the method may also include calculating a set of displacement vectors for multiple correlated covariance matrices. In some cases, multiple correlated covariance matrices can be correlated with pulse periods across consecutive adjacent depths in ultrasound data. In some cases, analysis of the covariance matrix across phase trajectories can reveal the principal components of tympanic membrane motion. For example... Figure 5 As shown in subfigure (b), this technique can be helpful in correcting low-energy sawtooth ultrasound signals. The covariance matrix can be constructed by acquiring RF ultrasound samples with Hilbert transform within a region of ultrasound data. This region can be a region of clear echoes. This region can also coincide with the length of the pulse period of the ultrasound data. The covariance matrix can be analyzed using the Hilbert transform.

[0100] The analytical pulse echo signal obtained from the Hilbert transform of a single pulse period (indicated by "k") can be expressed as:

[0101] Where R is the amplitude. ω =2 πf For carrier frequency, φLet z be the signal phase, and z be the depth, corresponding to the m-th a / d sample if discretized. Analytical pulse echo signals can be used to determine the elements of the covariance matrix. In some cases, the matrix can be an N×N covariance matrix. In some cases, the covariance matrix can originate from the same depth but from two different pulse periods. These pulse periods can each be a known number of pulse periods in pulse period k. These pulse periods can be denoted by exponents p and q. The covariance matrix can be added to by a fixed depth z or m. The covariance matrix can be an outer product matrix described by the following equation:

[0102] Where p = {0...N-1} and q = {0...N-1}, and the apostrophe can indicate complex conjugation.

[0103] The complete covariance matrix can be obtained for K(m={m0, m1, m2, ..., mR} from a set of adjacent depths m={m0, m1, m2, ..., mR}. p,q, m Summing is performed. Substituting H into the covariance expression above, we get the following formula:

[0104] It can be further simplified to: .

[0105] The covariance matrix resulting from the contributions across a series of consecutive depths spanning a specific tympanic membrane echo can be expressed by the following formula:

[0106] It can be written more concisely as

[0107] in

[0108] Furthermore, Δ can be at a depth m at two different times. p and q The phase difference of the samples. This can also be called the rate of change of the phase of the ultrasonic signal over time.

[0109] This avoids demodulation because, at the beginning of the description of K, the carrier frequency is actually subtracted by multiplying one exponent term by another exponent term that has been conjugate.

[0110] The covariance matrix can be calculated based on the mathematical description above. The eigenvectors and eigenvalues ​​of the resulting matrix can be expressed by the following formula:

[0111] inν 1 , ν 2 , ν 3 … ν N Let be the feature vector set, and ( λ 1 , λ 2 , λ 3 … ) is the characteristic value.

[0112] In some cases, the shift vectors of multiple correlated covariance matrices are calculated based on the target quality signal ratio. The target quality signal ratio can be a value between 0.00 and 1.00. In some cases, the target quality signal ratio can be 0.01, 0.02, 0.03, 0.04, 0.05, 0.06, 0.07, 0.08, 0.09, 0.1, 0.11, 0.12, 0.13, 0.14, 0.15, 0.16, 0.17, 0.18, 0.19, 0.20, 0.21, 0.22, or 0.2. 3, 0.24, 0.25, 0.26, 0.27, 0.28, 0.29, 0.30, 0.31, 0.32, 0.33, 0.34, 0.35, 0.36, 0.37, 0.38, 0.39, 0.40, 0.41, 0.42, 0.43, 0.44, 0.45, 0.46, 0.47, 0.48, 0. 49, 0.50, 0.51, 0.52, 0.53, 0.54, 0.55, 0.56, 0.57, 0.58, 0.59, 0.60, 0.61, 0.62, 0.63, 0.64, 0.65, 0.66, 0.67, 0.68, 0.69, 0.70, 0.71, 0.72, 0.73, 0.74, 0 0.75, 0.76, 0.77, 0.78, 0.79, 0.80, 0.81, 0.82, 0.83, 0.84, 0.85, 0.86, 0.87, 0.88, 0.89, 0.90, 0.91, 0.92, 0.93, 0.94, 0.95, 0.96, 0.97, 0.98, 0.99, or 1.00.

[0113] In some cases, the method may also include concatenating segments of the displacement vector set to generate a time trace of the tissue location. In some cases, the time trace may be a first trace. In some cases, the time trace may be a second trace. In some cases, the time trace may be a third trace.

[0114] In some cases, methods for determining the time-dependent location of a tissue's response to a stimulus may also include analyzing one or more eigenvectors of the covariance matrix. In some cases, one or more eigenvectors may each contain orbital rotations associated with the phase of the ultrasound data. Note that there are multiple ways to express the eigenvectors and eigenvalues ​​of the covariance operator. In some cases, membrane motion may be contained within a first eigenvector. In some cases, membrane motion may be contained within a second eigenvector. In some cases, membrane motion may be contained within a third eigenvector. In some cases, membrane motion may be contained within different eigenvectors.

[0115] In some cases, feature vectors can have similar phase trajectory behavior across multiple depths, such as Figure 5 Subgraphs (a), (b), and (c) are shown. In some cases, the overall signal energy may be concentrated in only a few eigenvectors. In some cases, the eigenvectors { ν 1 , ν 2 , ν 3 … ν N} can form an orthonormal basis set. In some cases, the ratio of each eigenvalue to the sum of all eigenvalues ​​can be a fraction of the signal class energy represented in the relevant eigenvector. In some cases, eigenvalue energy can be contained within an eigenvector. In some cases, at locations where signal energy decays (e.g. Figure 5 The subgraph (b) can show the simultaneous decrease in the energy fraction contained in the first eigenvector. In some cases, to quantize it, the target signal quality ratio can be expressed as: .

[0116] Within the range of pulse cycles included in the covariance calculation, M can be estimated at different depths along the ultrasonic beam. When no target is present, M can be a value between 0 and 1. In some cases, the variance of M can be higher compared to when a target is present. These aspects of quality-to-performance behavior are seen in, for example... Figure 10 middle.

[0117] A physical model of tympanic membrane movement can show a continuous cycle of phase without jumps or discontinuities. Figure 6 Subgraphs (a), (b), and (c) illustrate instances of non-physical phase behavior that may occur when tracking the tympanic membrane, as well as instances of unpacking and correcting it.

[0118] Figure 6 The subgraph (a) shows the subgraph from Figure 4An example of an eigenvector representing the covariance of region 2. In some cases, the eigenvector may show, for example, a decrease in amplitude. The observed decrease may occur because the intensity of reflection from the tympanic membrane is lower than that from a relatively stationary (non-moving) neighboring tissue (such as the tympanic membrane umbilicus). Figure 6 Subgraph (a) shows an instance of one of the orbital periods of the path, which can be offset. Figure 6 Subgraph (b) shows the relationship with Figure 6 The subgraph (a) has the same eigenvectors, but includes, for example, a pre-inserted bias, such that all the trajectories of the signal are around the origin of the complex plane. Figure 6 The subgraph (c) will, for example Figure 6 The cumulative phase change of the path in subgraph (a) is shown as trace 1, and... Figure 6 The cumulative phase change of the path in subgraph (b) is shown as trace 2. The difference between trace 2 and trace 1 can be seen, for example, in the dashed trace 3. Because from... Figure 6 subgraph (a) to Figure 6 The bias correction of the subgraph (b) may have a difference of, for example, 2π.

[0119] In some cases, methods for determining the time-dependent location of an organization's response to a stimulus may also include applying a bias to each of one or more eigenvectors such that the eigenvectors orbit around the origin in phase space. For example... Figure 6 As shown in subgraph (b), while adding a correction bias to the data can be a solution to the non-physical phase extraction problem, this may add its own artifacts to the data.

[0120] Figure 7 This demonstrates that, for example, an ultrasonic signal can be a linear sum of phasors in the complex plane. As shown, if the phasor orbit center is offset away from the origin of the complex plane, the motion associated with a particular phasor may be masked. Figure 7 Subgraph (a) shows phasor 2 (which can be the signal of interest), whose orbit can be far from the origin due to the bias from phasor 1. Figure 7 Subgraph (b) shows that the orbit of phasor 2 (which can be the signal of interest) can be around the origin of the complex plane. Figure 7 Subgraph (c) shows the results using ultrasound signal data from brain tissue. Figure 7 Examples of the behavior of two phasors in subgraph (a). In some cases, phasor 1 can be a slowly moving phasor, and the particles in the blood flow of the middle cerebral artery are phasors 2, which can be a rapidly moving phasor. Figure 7 The subplot (d) shows the brain / particle data using the same data. Figure 7The example of subgraph (b) can be filtered out by a high-pass filter (clutter filter) to remove the slow brain phasor, leaving only the embolus signal as a residual phasor orbiting around the recovery point. Figure 7 The subgraph (e) shows the subgraph based on having Figure 7 The example embolus location calculation for the data with the subgraph (a) type bias can inaccurately show very little movement of the embolus moving in the bleeding flow. Figure 7 The subgraph (f) can display, for example, based on having Figure 7 Subgraph (b) and Figure 7 The calculation of the thrombus position based on the characteristics of the subgraph (d) can lead to improved accuracy in calculating thrombus movement.

[0121] like Figure 5 As shown in subfigure (e), in some cases, the signal-to-noise ratio may decrease in the region near the target. The target could be, for example, a tympanic membrane. In some cases, performing covariance analysis over the entire same region containing the target can reduce broadband noise and make motion paths and dynamics more apparent. In some cases, the noise process covariance (data without a target) results in the use of more eigenvectors to describe the process compared to the target covariance (data with a target). In some cases, covariance analysis with a target can result in one or a single-digit component describing the target dynamics. In some cases, the principal eigenvector (with the largest eigenvalue) can describe the target motion, which has contributions from all a / d samples within the target echo. In some cases, broadband noise can often be analyzed separately from the signal.

[0122] In some cases, methods for determining the time-dependent location of a tissue's response to a stimulus may also include using the resulting biased eigenvectors to compute a second trace of the tissue location. In some cases, analyzing one or more eigenvectors may also include analyzing membrane motion information from one or more eigenvectors. In some cases, analyzing one or more eigenvectors may also include repeating the analysis in 10-millisecond (ms) steps. In some cases, the 10-ms step may be within an adjustable 20-ms analysis window. In some cases, analyzing one or more eigenvectors may also include analyzing depth range values. In some cases, analyzing one or more eigenvectors may also include creating a principal eigenvector set at each 10-ms step. In some cases, analyzing one or more eigenvectors may also include creating a principal eigenvector set within each 20-ms analysis window.

[0123] In some cases, the principal eigenvectors of covariance analysis can be repeatedly calculated across the depth and time ranges of the signal. Figure 3 The subgraph (A) is shown. Figure 3In the example shown in subplot (A), in some cases, the data stream can last for a duration of 1 second during tympanic membrane data capture. For example, this could be data from approximately 5000 pulse cycles for which an analysis of covariance is to be performed. In some cases, the analysis can be repeated in 10 ms steps. In some cases, the analysis can also be repeated within a 20 ms window from the signal start edge to the latest signal edge. In some cases, the analysis can also be repeated at the M1 value at the center of the tympanic membrane profile. For example, relative to... Figure 3 The subgraph (A) and Figure 3 The outline in subgraph (B), Figure 10 The first 80 M1 values ​​at each time point. In some cases, the stimulus can be an excitation. In some cases, the excitation can be a physical, electrical, or electromagnetic excitation. In some cases, the stimulus can be a pneumatic excitation. In some cases, the pneumatic excitation can be a jet of air.

[0124] For example, the set of principal feature vectors obtainable at each of 100 windows within a 1-second timeframe, incremented by 10 ms (one principal feature vector per window position), can be represented as follows: .

[0125] The membrane motion information contained in the eigenvector can be obtained by the following expression:

[0126] in U It is the "phase unwrapping operator". The expression can be the phase of the unwrapped eigenvector multiplied by a wavelength of the carrier frequency for each 4π radian phase change. It can be a vector that has a value for each pulse cycle included in the covariance analysis window.

[0127] Similar to the eigenvector set mentioned above, there exists a displacement vector set, which can be represented by the following formula: .

[0128] Figure 8 Subgraph (a) shows an instance of a vector set D containing positional changes during each covariance calculation time interval. These individual segments can show the relative positional changes produced by the tympanic membrane. In some cases, they can be "stitched" to their adjacent neighboring segments. For example, Figure 8 The subplot (b) is magnified to show these curves and to demonstrate their fit to each other in their overlapping regions. Figure 8 Subgraph (c) shows an example of splicing the segments together end to end, thus revealing the movement of the tympanic membrane.

[0129] Figure 9The phase distribution is shown, for example, against the background of raw RF ultrasound data.

[0130] Figure 10 An example of the quality metric M is shown, calculated over a 20 ms wide window spaced in 10 ms increments over a total 1-second duration. For example, the background noise could be close to 0.2, compared to a target of close to 1.0, with occasional decreases over both the time span and the depth span equal to the volume of the ultrasound sample. The top path could be the location of the maximum M value across the depth at each 10 ms increment.

[0131] System for determining time-dependent position of tissue in response to a stimulus This article discloses a system for determining the time-dependent location of a tissue's response to a stimulus. The tissue can be biological tissue. The tissue can be plant, food, animal, or human. The tissue can be human tissue, such as an organ. The tissue can be a smooth tissue organ or a musculoskeletal organ. The tissue can be, for example, glands such as the heart, liver, spleen, ear organs, gallbladder, pancreas, kidney, bladder, uterus or ovary, prostate, thyroid and parathyroid glands, blood vessels, eye organs, appendix, gastrointestinal tract, or any other organ. The tissue can be the tympanic membrane. The stimulus can be a physical stimulus. The stimulus can be a pressure stimulus, electrical stimulus, electromagnetic stimulus, physical contact stimulus, fluid injection stimulus, or sensory stimulus, or another type of stimulus. The stimulus can be a pressure change stimulus.

[0132] The system may include a processor containing executable instructions stored thereon, which, when executed, are configured to receive ultrasound data. In other embodiments, the processor is configured to receive another type of electromagnetic data. The ultrasound data may originate from ultrasound waveforms reflected from tissue. In other embodiments, the ultrasound data may originate from one or more segments of ultrasound waveforms reflected from tissue. The processor may be configured to calculate a covariance matrix from the ultrasound data. The covariance data may be a function of the depth of multiple pulse periods of the ultrasound data. The covariance data can be a function of a depth of approximately 2, 3, 4, 5, 10, 15, 20, 25, 50, 100, 200, 300, 400, 500, 1000, 2000, 3000, 4000, 5000, 1000, 2000, 3000, 4000, 5000, or more than approximately 5000 pulse periods. In some cases, the number of pulse periods can be between 4500 and 5000. In some cases, the number of pulse periods can be 4800. The processor can be configured to compute one or more eigenvectors and associated eigenvalues ​​of the covariance matrix. In some cases, the processor can be configured to compute 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, or 25 eigenvectors. The processor can contain 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, or more than 50 feature vectors. The processor can also be configured to determine a first trace associated with the contents of a principal feature vector among one or more feature vectors.

[0133] In some cases, the system may also include a processor configured to determine the time-dependent location of tissue response to stimulation by analyzing the frequency content of ultrasound data across multiple pulse cycles. In some cases, the frequency content may be a trace. In some cases, the processor may be configured to calculate a second trace of tissue location based at least in part on the frequency content. In some cases, the frequency content may be a measurement of frequency. In some cases, the frequency content may be a change in frequency or the rate of frequency.

[0134] In some embodiments, the system may further include a processor configured to output an indication of the disease state of the tissue. In some embodiments, the indication may be responsive to a trace. In some embodiments, the indication may be responsive to a second trace at the tissue location. In other embodiments, the system may further include a processor configured to output an indication of the health state of the tissue. In some embodiments, the indication may be responsive to a second trace at the tissue location. In other embodiments, the system may further include a processor configured to output an indication of an undetermined state of the tissue. In some embodiments, the indication may be responsive to a second trace at the tissue location. In some embodiments, the indication may be responsive to a first trace at the tissue location. In some embodiments, the indication may be responsive to a third trace at the tissue location.

[0135] In some cases, multiple pulse cycles may be sequential. In other cases, multiple pulse cycles may be non-sequential. In some cases, the tissue may be an organ of the ear, such as the tympanic membrane, ear bones, eustachian tube, cochlea, oval window, or another organ of the ear. In some cases, the tissue may be another human organ, such as the heart, liver, kidneys, stomach, eye, brain, or another organ.

[0136] In some embodiments, the system may further include a processor configured to output an indication of the disease state of the tissue. In some embodiments, the indication may be responsive to a trace. In some embodiments, the indication may be responsive to a first trace at the tissue location. In other embodiments, the system may further include a processor configured to output an indication of the health state of the tissue. In some embodiments, the indication may be responsive to a first trace at the tissue location. In other embodiments, the system may further include a processor configured to output an indication of an undetermined state of the tissue. In some embodiments, the indication may be responsive to a first trace at the tissue location. In some embodiments, the indication may be responsive to a second trace at the tissue location. In some embodiments, the indication may be responsive to a third trace at the tissue location.

[0137] In some cases, each pulse period in a plurality of pulse periods can be associated with a covariance matrix from a plurality of covariance matrices. In some cases, the plurality of covariance matrices can include 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, or 26 covariance matrices. Covariance matrix, 27 covariance matrices, 28 covariance matrices, 29 covariance matrices, 30 covariance matrices, 31 covariance matrices, 32 covariance matrices, 33 covariance matrices, 34 covariance matrices, 35 covariance matrices, 36 covariance matrices, 37 covariance matrices, 38 covariance matrices, 39 covariance matrices, 40 covariance matrices, 41 covariance matrices, 42 covariance matrices, 43 covariance matrices, 44 covariance matrices, 45 covariance matrices, 46 covariance matrices, 47 covariance matrices, 48 ​​covariance matrices, 49 covariance matrices, 50 covariance matrices, or more than 50 covariance matrices.

[0138] In some cases, the system may also include a processor configured to compute a set of displacement vectors for multiple correlated covariance matrices. In some cases, the multiple correlated covariance matrices can be correlated with pulse periods across consecutive adjacent depths of ultrasound data. In some cases, analysis of the covariance matrices across phase trajectories can reveal the principal components of tympanic membrane motion. For example... Figure 5 As shown in subfigure (b), this technique can be helpful in correcting low-energy sawtooth ultrasound signals. The covariance matrix can be constructed by acquiring RF ultrasound samples with Hilbert transform within a region of ultrasound data. This region can be a region of clear echoes. This region can also coincide with the length of the pulse period of the ultrasound data. The covariance matrix can be analyzed using the Hilbert transform.

[0139] The analytical pulse echo signal obtained from the Hilbert transform of a single pulse period (indicated by "k") can be expressed as:

[0140] Where R is the amplitude. ω =2 πf For carrier frequency, φLet z be the signal phase, and z be the depth, corresponding to the m-th a / d sample if discretized. Analytical pulse echo signals can be used to determine the elements of the covariance matrix. In some cases, the matrix can be an N×N covariance matrix. In some cases, the covariance matrix can originate from the same depth but from two different pulse periods. These pulse periods can each be a known number of pulse periods in pulse period k. These pulse periods can be denoted by exponents p and q. The covariance matrix can be added to by a fixed depth z or m. The covariance matrix can be an outer product matrix described by the following equation:

[0141] Where p = {0...N-1} and q = {0...N-1}, and the apostrophe can indicate complex conjugation.

[0142] The complete covariance matrix can be obtained for K(m={m0, m1, m2, ..., mR} from a set of adjacent depths m={m0, m1, m2, ..., mR}. p,q, m Summing is performed. Substituting H into the covariance expression above, we get the following formula:

[0143] It can be further simplified to: .

[0144] The covariance matrix resulting from the contributions across a series of consecutive depths spanning a specific tympanic membrane echo can be expressed by the following formula:

[0145] It can be written more concisely as

[0146] in

[0147] Furthermore, Δ can be at a depth m at two different times. p and q The phase difference of the samples. This can also be called the rate of change of the phase of the ultrasonic signal over time.

[0148] This avoids demodulation because, at the beginning of the description of K, the carrier frequency is actually subtracted by multiplying one exponent term by another exponent term that has been conjugate.

[0149] The covariance matrix can be calculated based on the mathematical description above. The eigenvectors and eigenvalues ​​of the resulting matrix can be expressed by the following formula:

[0150] inν 1 , ν 2 , ν 3 … ν N Let be the feature vector set, and ( λ 1 , λ 2 , λ 3 … ) is the characteristic value.

[0151] In some cases, the shift vectors of multiple correlated covariance matrices are calculated based on the target quality signal ratio. The target quality signal ratio can be a value between 0.00 and 1.00. In some cases, the target quality signal ratio can be 0.01, 0.02, 0.03, 0.04, 0.05, 0.06, 0.07, 0.08, 0.09, 0.1, 0.11, 0.12, 0.13, 0.14, 0.15, 0.16, 0.17, 0.18, 0.19, 0.20, 0.21, 0.22, or 0.2. 3, 0.24, 0.25, 0.26, 0.27, 0.28, 0.29, 0.30, 0.31, 0.32, 0.33, 0.34, 0.35, 0.36, 0.37, 0.38, 0.39, 0.40, 0.41, 0.42, 0.43, 0.44, 0.45, 0.46, 0.47, 0.48, 0. 49, 0.50, 0.51, 0.52, 0.53, 0.54, 0.55, 0.56, 0.57, 0.58, 0.59, 0.60, 0.61, 0.62, 0.63, 0.64, 0.65, 0.66, 0.67, 0.68, 0.69, 0.70, 0.71, 0.72, 0.73, 0.74, 0 0.75, 0.76, 0.77, 0.78, 0.79, 0.80, 0.81, 0.82, 0.83, 0.84, 0.85, 0.86, 0.87, 0.88, 0.89, 0.90, 0.91, 0.92, 0.93, 0.94, 0.95, 0.96, 0.97, 0.98, 0.99, or 1.00.

[0152] In some cases, the system may also include a processor configured to stitch together segments of the displacement vector set to generate a time trace of the tissue location. In some cases, the time trace may be a first trace. In some cases, the time trace may be a second trace. In some cases, the time trace may be a third trace.

[0153] In some cases, systems for determining the time-dependent location of a tissue's response to a stimulus may also include a processor configured to analyze one or more eigenvectors of the covariance matrix. In some cases, the one or more eigenvectors may each contain orbital rotations associated with the phase of the ultrasound data. Note that there are multiple ways to express the eigenvectors and eigenvalues ​​of the covariance operator. In some cases, membrane motion may be contained within a first eigenvector. In some cases, membrane motion may be contained within a second eigenvector. In some cases, membrane motion may be contained within a third eigenvector. In some cases, membrane motion may be contained within different eigenvectors.

[0154] In some cases, feature vectors can have similar phase trajectory behavior across multiple depths, such as Figure 5 Subgraphs (a), (b), and (c) are shown. In some cases, the overall signal energy may be concentrated in only a few eigenvectors. In some cases, the eigenvectors { ν 1 , ν 2 , ν 3 … ν N} can form an orthonormal basis set. In some cases, the ratio of each eigenvalue to the sum of all eigenvalues ​​can be a fraction of the signal class energy represented in the relevant eigenvector. In some cases, eigenvalue energy can be contained within an eigenvector. In some cases, at locations where signal energy decays (e.g. Figure 5 The subgraph (b) can show the simultaneous decrease in the energy fraction contained in the first eigenvector. In some cases, to quantize it, the target signal quality ratio can be expressed as: .

[0155] Within the range of pulse cycles included in the covariance calculation, M can be estimated at different depths along the ultrasonic beam. When no target is present, M can be a value between 0 and 1. In some cases, the variance of M can be higher compared to when a target is present. These aspects of quality-to-performance behavior are seen in, for example... Figure 10 middle.

[0156] A physical model of tympanic membrane movement can show a continuous cycle of phase without jumps or discontinuities. Figure 6 Subgraphs (a), (b), and (c) illustrate instances of non-physical phase behavior that may occur when tracking the tympanic membrane, as well as instances of unpacking and correcting it.

[0157] Figure 6 The subgraph (a) shows the subgraph from Figure 4An example of an eigenvector representing the covariance of region 2. In some cases, the eigenvector may show, for example, a decrease in amplitude. The observed decrease may occur because the intensity of reflection from the tympanic membrane is lower than that from a relatively stationary (non-moving) neighboring tissue (such as the tympanic membrane umbilicus). Figure 6 Subgraph (a) shows an instance of one of the orbital periods of the path, which can be offset. Figure 6 Subgraph (b) shows the relationship with Figure 6 The subgraph (a) has the same eigenvectors, but includes, for example, a pre-inserted bias, such that all the trajectories of the signal are around the origin of the complex plane. Figure 6 The subgraph (c) will, for example Figure 6 The cumulative phase change of the path in subgraph (a) is shown as trace 1, and... Figure 6 The cumulative phase change of the path in subgraph (b) is shown as trace 2. The difference between trace 2 and trace 1 can be seen, for example, in the dashed trace 3. Because from... Figure 6 subgraph (a) to Figure 6 The bias correction of the subgraph (b) may have a difference of, for example, 2π.

[0158] In some cases, systems for determining the time-dependent location of an organization's response to a stimulus may also include a processor configured to apply a bias to each of one or more eigenvectors such that the eigenvectors orbit the origin in phase space. For example... Figure 6 As shown in subgraph (b), while adding a correction bias to the data can be a solution to the non-physical phase extraction problem, this may add its own artifacts to the data.

[0159] Figure 7 This demonstrates that, for example, an ultrasonic signal can be a linear sum of phasors in the complex plane. As shown, if the phasor orbit center is offset away from the origin of the complex plane, the motion associated with a particular phasor may be masked. Figure 7 Subgraph (a) shows phasor 2 (which can be the signal of interest), whose orbit can be far from the origin due to the bias from phasor 1. Figure 7 Subgraph (b) shows that the orbit of phasor 2 (which can be the signal of interest) can be around the origin of the complex plane. Figure 7 Subgraph (c) shows the results using ultrasound signal data from brain tissue. Figure 7 Examples of the behavior of two phasors in subgraph (a). In some cases, phasor 1 can be a slowly moving phasor, and the particles in the blood flow of the middle cerebral artery are phasors 2, which can be a rapidly moving phasor. Figure 7 The subplot (d) shows the brain / particle data using the same data. Figure 7The example of subgraph (b) can be filtered out by a high-pass filter (clutter filter) to remove the slow brain phasor, leaving only the embolus signal as a residual phasor orbiting around the recovery point. Figure 7 The subgraph (e) shows the subgraph based on having Figure 7 The example embolus location calculation for the data with the subgraph (a) type bias can inaccurately show very little movement of the embolus moving in the bleeding flow. Figure 7 The subgraph (f) can display, for example, based on having Figure 7 Subgraph (b) and Figure 5 The calculation of the thrombus position based on the characteristics of the subgraph (d) can lead to improved accuracy in calculating thrombus movement.

[0160] like Figure 3 As shown in subfigure (e), in some cases, the signal-to-noise ratio may decrease in the region near the target. The target could be, for example, a tympanic membrane. In some cases, performing covariance analysis over the entire same region containing the target can reduce broadband noise and make motion paths and dynamics more apparent. In some cases, the noise process covariance (data without a target) results in the use of more eigenvectors to describe the process compared to the target covariance (data with a target). In some cases, covariance analysis with a target can result in one or a single-digit component describing the target dynamics. In some cases, the principal eigenvector (with the largest eigenvalue) can describe the target motion, which has contributions from all a / d samples within the target echo. In some cases, broadband noise can often be analyzed separately from the signal.

[0161] In some cases, systems for determining the time-dependent location of tissue response to stimuli may also include a processor configured to compute a second trace of the tissue location using the resulting biased eigenvectors. In some cases, analyzing one or more eigenvectors may also include analyzing membrane motion information from the one or more eigenvectors. In some cases, analyzing one or more eigenvectors may also include repeated analysis in 10-millisecond (ms) steps. In some cases, the 10-ms step may be within an adjustable 20-ms analysis window. In some cases, analyzing one or more eigenvectors may also include analyzing depth range values. In some cases, analyzing one or more eigenvectors may also include creating a principal eigenvector set at each 10-ms step. In some cases, analyzing one or more eigenvectors may also include creating a principal eigenvector set within each 20-ms analysis window.

[0162] In some cases, the principal eigenvectors of covariance analysis can be repeatedly calculated across the depth and time ranges of the signal. Figure 3 The subgraph (A) is shown. Figure 3In the example shown in subplot (A), in some cases, the data stream can last for a duration of 1 second during tympanic membrane data capture. For example, this could be data from approximately 5000 pulse cycles for which an analysis of covariance is to be performed. In some cases, the analysis can be repeated in 10 ms steps. In some cases, the analysis can also be repeated within a 20 ms window from the signal start edge to the latest signal edge. In some cases, the analysis can also be repeated at the M1 value at the center of the tympanic membrane profile. For example, relative to... Figure 3 The subgraph (A) and Figure 10 The outline in subgraph (B), Figure 8 The first 80 M1 values ​​at each time point. In some cases, the stimulus can be an excitation. In some cases, the excitation can be a physical, electrical, or electromagnetic excitation. In some cases, the stimulus can be a pneumatic excitation. In some cases, the pneumatic excitation can be a jet of air.

[0163] For example, the set of principal feature vectors obtainable at each of 100 windows within a 1-second timeframe, incremented by 10 ms (one principal feature vector per window position), can be represented as follows: .

[0164] The membrane motion information contained in the eigenvector can be obtained by the following expression:

[0165] in U It is the "phase unwrapping operator". The expression can be the phase of the unwrapped eigenvector multiplied by a wavelength of the carrier frequency for each 4π radian phase change. It can be a vector that has a value for each pulse cycle included in the covariance analysis window.

[0166] Similar to the eigenvector set mentioned above, there exists a displacement vector set, which can be represented by the following formula: .

[0167] Figure 8 Subgraph (a) shows an instance of a vector set D containing positional changes during each covariance calculation time interval. These individual segments can show the relative positional changes produced by the tympanic membrane. In some cases, they can be "stitched" to their adjacent neighboring segments. For example, Figure 8 The subplot (b) is magnified to show these curves and to demonstrate their fit to each other in their overlapping regions. Figure 9 Subgraph (c) shows an example of splicing the segments together end to end, thus revealing the movement of the tympanic membrane.

[0168] Figure 10The phase distribution is shown, for example, against the background of raw RF ultrasound data.

[0169] Autoregressive method An example of the quality metric M is shown, calculated over a 20 ms wide window spaced in 10 ms increments over a total 1-second duration. For example, the background noise could be close to 0.2, compared to a target of close to 1.0, with occasional decreases over both the time span and the depth span equal to the volume of the ultrasound sample. The top path could be the location of the maximum M value across the depth at each 10 ms increment.

[0170] Figure 11 Further correction of the unwrapped phase can be advantageous. For example, near pressure transitions, the phase may exhibit anomalies, such as phase jumps, indicating rapid, non-physical movement of the tympanic membrane. This anomaly is referred to as inaccuracy or "sawtooth" in the resulting motion detection results. In many cases, this sawtooth is erroneous. Optionally, these non-physical movements can be addressed using the anti-aliasing methods and systems described below.

[0171] Figure 9 It shows following a pattern similar to Figure 12 An enlarged example of the covariance trace of the RF phase trajectory of the TM.

[0172] Figure 12 This shows that, except for two points (time ≈ 0.18 s and time ≈ 0.62 s), the covariance trace remains faithful to the phase trace RF data. These two covariance error omissions occur during the pressure transition region and are the two locations with the lowest signal energy (indicated by triangles). Figure 13 As shown, covariance can "jump" from one phase trajectory to another, thus deviating from the correct path. This can occur during the pressure transition region, when the signal energy is at its lowest.

[0173] Processing ultrasound data can include analyzing the frequency content of RF echoes over multiple pulse cycles. The frequency of slow-time ultrasound migration can be proportional to velocity, thus allowing frequency estimation techniques to be applied across slow time periods to track the velocity of the reflector over time. Autoregressive (AR) models are a technique that can be used to estimate the frequency spectrum of a signal.

[0174] In some cases, methods for determining the time-dependent location of a tissue response to a stimulus may also include analyzing the frequency content of ultrasound data across multiple pulse cycles. In some cases, the method may also include determining one or more regions where the first trace contains a non-physical tissue location or movement. In some cases, methods for determining the time-dependent location of a tissue response to a stimulus may also include analyzing the frequency content of ultrasound data across multiple pulse cycles. In some cases, the method may also include replacing the first trace in one or more regions with at least a portion of a second trace. In some cases, determining that the first trace may contain one or more regions where the first trace contains a non-physical tissue location or movement may include utilizing autoregression. In some cases, the autoregression may be third-order autoregression. In some cases, the autoregression may be higher than third-order autoregression. In some cases, the autoregression may be first-order autoregression. In some cases, the autoregression may be second-order autoregression. In some cases, the autoregression may be fourth-order autoregression. In some cases, the autoregression may be fifth-order autoregression. In some cases, the autoregression may exceed fifth-order autoregression.

[0175] The methods disclosed herein for improving the accuracy of time-dependent location determination of tissue response to stimuli may include generating a frequency spectrum of a signal trace. The method may also include applying autoregression to generate one or more poles. The number of poles generated may be one, two, three, or more than three. The method may also include using features of one or more poles to exclude or select poles from the one or more poles. The number of poles that can be excluded or selected may be one, two, three, or more than three. The method may also include generating an improved signal trace of the tissue location from the remaining or selected poles from the one or more poles.

[0176] IEEE Transactions on Ultrasonics, Ferroelectrics, and An example of estimating the frequency spectrum of a signal using a second-order AR model is shown. The left subplot shows the artificial signal, which is the sum of a 3 Hz sine wave and a 5 Hz sine wave. The middle subplot shows the frequency spectrum of the signal, which is estimated using the illustrated equation. As expected, the middle subplot shows that the estimated frequency spectrum has peaks at 3 Hz and 5 Hz. The coefficients a(1) and a(2) in the equation are AR coefficients, which are obtained using the Burg algorithm. The right subplot shows that the poles of the middle equation contain information about the frequency estimation. The angle of each pole on the complex unit circle can correspond to a peak in the frequency spectrum, where + / - π Equal to + / - Nyquist frequency. Radius of each pole. r It reflects the variance or width of the frequency spectrum and can be compared with 1 - r Proportional.

[0177] AR models can be used Frequency ControlFigure 13 The method described in Young Bok Ahn and Song Bai Park, "Estimation of mean frequency and variance of ultrasonic Doppler signal by using second-order autoregression," May 1991, doi: 10.1109 / 58.79600, Vol. 38, No. 3, pp. 172-182, is adopted and is incorporated herein by reference for all purposes.

[0178] In some cases, AR models can be used to estimate the frequency content of a signal. The order of the AR model determines how many distinct frequency peaks it can separate. For example, in... Figure 14 In this case, the input signal has two frequencies to be separated, so the second-order AR model returns two frequency peaks.

[0179] IEEE Transactions on Sonics and Ultrasonics This example demonstrates the use of a first-order AR model to estimate the frequency spectrum of a signal. A first-order AR model can return a single frequency peak in spectral estimation; therefore, if the input signal contains two frequencies, the AR model can return the average of those two frequencies.

[0180] AR models can be used Figure 13 The method described in C. Kasai, K. Namekawa, A. Koyano and R. Omoto, "Real-Time Two-Dimensional Blood Flow Imaging Using an Autocorrelation Technique," May 1985, doi: 10.1109 / T-SU.1985.31615, Vol. 32, No. 3, pp. 458-464, is incorporated herein by reference for all purposes.

[0181] Figure 14 and Figure 15 The example demonstrates the application of AR frequency estimation to simple artificial signals, but the technique can also be applied to slow-time ultrasound signals to estimate the velocity of the TM.

[0182] The AR processing techniques described in this article are not limited to first-order or second-order models. In some cases, autoregression can be applied to one or more points in the signal trace. In some cases, autoregression can be applied to two, three, four, five, or more than five points in the signal trace. In some cases, autoregression can be third-order. In some cases, autoregression can be higher than third-order. In some cases, the order of autoregression can correspond to the number of frequency peaks separated in the signal trace. In some cases, the signal trace may include a covariance motion detection trace. In some cases, autoregression can be applied to one or more eigenvectors of the covariance motion detection trace. In some cases, autoregression can be applied to two, three, four, five, or more than five eigenvectors of the covariance motion detection trace. In some cases, methods to improve the accuracy of determining the time-dependent location of a tissue response to a stimulus may also include estimating the average frequency of the signal trace. In some cases, estimating the average frequency of a signal trace may involve using the angle between each of one or more poles and the center of the complex unit circle in the complex plane. In other cases, the angle can be calculated by taking the arctangent of the ratio of the imaginary part of the pole to the real part of the pole.

[0183] In some cases, methods to improve the accuracy of determining the time-dependent location of a tissue's response to a stimulus may also include estimating the frequency variance of the signal trace. The variance can be estimated using the radius value between each of one or more poles and the center of the complex unit circle in the complex plane. In some cases, methods to improve the accuracy of determining the time-dependent location of a tissue's response to a stimulus may also include estimating velocity. In some cases, the velocity can be represented by the signal trace using the radius value between each of one or more poles and the center of the complex unit circle in the complex plane. In some cases, the number of poles may be one, two, three, four, five, or more than five.

[0184] In some cases, methods to improve the accuracy of determining the time-dependent location of an organization's response to a stimulus may also include plotting the velocity traces of the signal trace over time. In some cases, the velocity traces of the signal trace can be plotted over time by plotting the angles between one or more poles and the center of the complex unit circle in the complex plane.

[0185] The radius of each pole can be used, for example, by using the following equation (where T is the pulse period and r is the pole radius) to determine the variance of the estimated frequency spectrum.

[0186]

[0187] Figure 15The diagram shows the poles of the spectral estimation obtained, for example, when a 12-point third-order AR model is applied to the first eigenvector of the covariance. Pole 1 can primarily estimate the low-frequency velocity of the TM, while poles 2 and 3 can primarily estimate the velocity from high-frequency noise.

[0188] Figure 15 The poles shown can be obtained by taking the root of the denominator of the third-order estimate of the power spectral density.

[0189]

[0190] To determine the roots (“poles”) of the denominator, the zeros of the third-order polynomial in the denominator of the above equation can be solved using the cubic formula shown below. These zeros of the cubic polynomial are the poles of the third-order AR model, and these three poles provide frequency estimation for ultrasonic data processing.

[0191]

[0192]

[0193]

[0194]

[0195]

[0196] In the third-order AR model, pole 1 corresponds to root 1 of the cubic formula above. Pole 2 and pole 3 are the "angular conjugate poles" of pole 1, just as root 2 and root 3 are conjugate combinations of root 1.

[0197] like Figure 16 As shown, applying the 12-point third-order AR model to the first eigenvector of the covariance yields three poles.

[0198] like Figure 17 As shown, the velocity over time can be determined by plotting the angle of a given pole across all pulse periods. This angle can be converted to velocity using the equation shown above. If the velocity estimated through pole 1 is integrated over time to obtain the displacement ( Figure 16 If the integral result is then compared with the location of pole 1, the trace can be traced back to the covariance location.

[0199] In some cases, methods for improving the accuracy of determining the time-dependent location of an organization's response to a stimulus may also include calculating one or more displacement values ​​between the location of the velocity trace and the location of the signal trace. In some cases, determining one or more displacement values ​​may include integrating the velocity in the velocity trace. In some cases, the characteristics of one or more poles may also include radii, angles, velocities, complex-plane distances between one or more poles, velocity variations between one or more poles, inertial velocity variations between one or more poles, or momentum of one or more poles, or any combination thereof. In some cases, one or more poles may be divided into one or more frequency groups based on frequency. In some cases, one or more poles may be divided into two frequency groups. In some cases, one or more poles may be divided into three frequency groups. In some cases, one or more poles may be divided into more than three frequency groups.

[0200] Table 1 below shows the data from ultrasound migration sequence data { z 1 , z 2 , z 3 , …, z N A summary of exemplary formulas for calculating first-, second-, and third-order autoregressive models (extremes).

[0201]

[0202] Figure 17 An exemplary pole angle plot is shown in both velocity and radians over time. The pole angle can be calculated by taking the arctangent of the ratio of the imaginary part of a given pole to the real part of a given pole. This angle can be plotted over time and then converted to velocity using the equations shown above.

[0203] Figure 18 A graph showing the pole velocities and positions for all pulse cycles is displayed. Once the velocities for all pulse cycles are calculated, they can be integrated over time to obtain the displacement of TM.

[0204] System for monitoring movement of tissue in response to a stimulus The comparison between the covariance method and the pole 1 of the third-order AR model is shown.

[0205] In some cases, generating an optimized signal trace may further include replacing one or more poles in the signal trace with one or more poles from different frequency groups. In some cases, one or more poles in the signal trace may be replaced when the absolute error between one or more of the characteristics of one or more poles is higher than a threshold. In some embodiments, the threshold is a value calculated using one or more of the pole selection algorithms in Table 3. In some embodiments, the pole signs are reversed. In some embodiments, the threshold is between 0.1 and 5.0. In some cases, the threshold is approximately 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, 5.0, or approximately greater than 5.0. In some cases, one or more poles in the signal trace may be replaced when the absolute error between one or more of the characteristics of one or more poles is lower than a threshold. In some embodiments, the threshold is a value calculated using one or more of the pole selection algorithms in Table 3. In some embodiments, the pole signs are reversed. In some embodiments, the threshold is between 0.1 and 5.0. In some cases, the threshold is approximately 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, 5.0, or approximately greater than 5.0. In some cases, the threshold can be π / 3, and the pole signs can be opposite. In some cases, the threshold can be π / 2, π / 3, π / 4, π / 5, π / 6, π / 7, π / 8, π / 9, π / 10, or less than π / 10. In some cases, the pole signs can be the same.

[0206] Figure 15 In some cases, the system disclosed herein may include a processor configured to apply autoregression to one or more points in a signal trace. In some cases, autoregression may be applied to two, three, four, five, or more than five points in the signal trace. In some cases, the autoregression may be third-order autoregression. In some cases, the autoregression may be higher than third-order autoregression. In some cases, the order of the autoregression may correspond to the number of separated frequency peaks in the signal trace. In some cases, the signal trace may include a covariance motion detection trace. In some cases, autoregression may be applied to one or more eigenvectors of the covariance motion detection trace. In some cases, autoregression may be applied to two, three, four, five, or more than five eigenvectors of the covariance motion detection trace. In some cases, the system including a processor configured to improve the accuracy of determining the time-dependent location of a tissue response to a stimulus may also include estimating the average frequency of the signal trace. In some cases, estimating the average frequency of a signal trace may involve using the angle between each of one or more poles and the center of the complex unit circle in the complex plane. In other cases, the angle can be calculated by taking the arctangent of the ratio of the imaginary part of the pole to the real part of the pole.

[0207] In some cases, a system including a processor configured to improve the accuracy of determining the time-dependent location of a tissue response to a stimulus may also include a processor configured to estimate the frequency variance of the signal trace. The variance can be estimated by the processor using the radius value between each of one or more poles and the center of the complex unit circle in the complex plane. In some cases, a system including a processor configured to improve the accuracy of determining the time-dependent location of a tissue response to a stimulus may also include a processor for estimating velocity. In some cases, the velocity can be represented by the signal trace using the radius value between each of one or more poles and the center of the complex unit circle in the complex plane. In some cases, the number of poles can be one, two, three, four, five, or more than five.

[0208] In some cases, systems including processors configured to improve time-dependent location accuracy may include processors configured to determine the velocity of tissue response to stimuli. Velocity determination may also include plotting a velocity trajectory of the signal trace over time. In some cases, the velocity trajectory of the signal trace can be plotted over time by plotting the angle between one or more poles and the center of the complex unit circle in the complex plane.

[0209] The radius of each pole can be used, for example, by using the following equation (where T is the pulse period and r is the pole radius) to determine the variance of the estimated frequency spectrum.

[0210]

[0211] Figure 15 The diagram shows the poles of the spectral estimation obtained, for example, when a 12-point third-order AR model is applied to the first eigenvector of the covariance. Pole 1 can primarily estimate the low-frequency velocity of the TM, while poles 2 and 3 can primarily estimate the velocity from high-frequency noise.

[0212] Figure 15 The poles shown can be obtained by taking the root of the denominator of the third-order estimate of the power spectral density.

[0213]

[0214] To determine the roots (“poles”) of the denominator, the zeros of the third-order polynomial in the denominator of the above equation can be solved using the cubic formula shown below. These zeros of the cubic polynomial are the poles of the third-order AR model, and these three poles provide frequency estimation for ultrasonic data processing.

[0215]

[0216]

[0217]

[0218]

[0219]

[0220] In the third-order AR model, pole 1 corresponds to root 1 of the cubic formula above. Pole 2 and pole 3 are the "angular conjugate poles" of pole 1, just as root 2 and root 3 are conjugate combinations of root 1.

[0221] like Figure 16 As shown, applying the 12-point third-order AR model to the first eigenvector of the covariance yields three poles.

[0222] like Figure 17 As shown, the velocity over time can be determined by plotting the angle of a given pole across all pulse periods. This angle can be converted to velocity using the equation shown above. If the velocity estimated through pole 1 is integrated over time to obtain the displacement ( Hybrid method and system If the integral result is then compared with the location of pole 1, the trace can be traced back to the covariance location.

[0223] In some cases, a system including a processor configured to improve the accuracy of time-dependent positions may include a processor configured to calculate one or more displacement values ​​between the position of a velocity trace and the position of a signal trace. In some cases, determining one or more displacement values ​​may include integrating the velocity in the velocity trace. In some cases, the characteristics of one or more poles may also include radii, angles, velocities, complex-plane distances between one or more poles, velocity variations between one or more poles, inertial velocity variations between one or more poles, or momentum of one or more poles, or any combination thereof. In some cases, one or more poles may be divided into one or more frequency groups based on frequency. In some cases, one or more poles may be divided into two frequency groups. In some cases, one or more poles may be divided into three frequency groups. In some cases, one or more poles may be divided into more than three frequency groups.

[0224] In some cases, generating an optimized signal trace may further include replacing one or more poles in the signal trace with one or more poles from different frequency groups. In some cases, one or more poles in the signal trace may be replaced when the absolute error between one or more of the characteristics of one or more poles is higher than a threshold. In some implementations, the threshold is a value calculated using one or more of the pole selection algorithms in Table 3. In some implementations, the pole signs are reversed. In some implementations, the threshold is between 0.1 and 5.0. In some cases, the threshold is approximately 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, 5.0, or approximately greater than 5.0. In some cases, one or more poles in the signal trace may be replaced when the absolute error between one or more of the characteristics of one or more poles is lower than a threshold. In some cases, the threshold may be π / 3, and the pole signs may be reversed. In some cases, the threshold may be π / 2, π / 3, π / 4, π / 5, π / 6, π / 7, π / 8, π / 9, π / 10, or less than π / 10. In some cases, the pole signs can be the same.

[0225] Figure 19 Hybrid approaches that combine the fidelity of covariance with the ability of AR estimation to reliably follow phase trajectories during pressure transitions can be advantageous. Hybrid methods or systems can combine covariance and autoregressive processing algorithms into a single approach.

[0226] In some cases, AR traces can be used to adjust the covariance. This adjustment can be applied to regions where the covariance does not follow a "physical" phase trajectory (i.e., a phase trajectory that conforms to a physical interpretation of the membrane movement).

[0227] Figure 20 An exemplary visual overview of the process of adjusting covariance traces based on AR analysis is provided. As shown, the covariance trace can transition from one phase trace to another. The region of covariance error pressure transition can be an error region. The error region can be replaced with segments of the AR trace to create an anti-aliased trace. This anti-aliased trace can represent the combination of covariance and AR processing.

[0228] Figure 21 An example implementation of the anti-aliasing (hybrid) algorithm is shown. Three different TM displacement traces are shown (top). A third-order AR trace can be combined with a covariance trace to form an anti-aliasing trace. The velocity traces of AR processing and covariance are shown in the middle subplot. The absolute error between the AR velocity and the covariance velocity is shown, for example, in the bottom subplot. When the difference between the absolute errors of the AR trace and the covariance trace is greater than... π When the velocity sign is opposite to 3, the covariance displacement trace can be repaired using AR traces.

[0229] Figure 22 An exemplary histogram showing the absolute value of the error between AR velocity and covariance velocity is displayed.

[0230] Pole selection Exemplary results of the anti-aliasing procedure are shown. AR traces and covariance traces can be combined to form anti-aliased traces. This hybrid trace is faithful to the RF phase trajectory data and does not miss pressure transitions.

[0231] The above method can be executed at a time very close to the time of covariance calculation, or it can be executed as a post-processing routine.

[0232] This paper discloses a system for improving time-dependent location determination. The system may include a covariance analysis model. The covariance analysis model may contain executable instructions that, when executed by a processor, can be configured to derive one or more covariance matrices from ultrasound data. The processor can be configured to derive one, two, three, four, five, or more than five covariance matrices from the ultrasound data. The processor can also be configured to compute one or more eigenvectors of the one or more covariance matrices. The processor can also be configured to generate traces from the one or more eigenvectors. The system may also include one or more autoregressors configured to generate one or more poles. The autoregressors can be configured to generate one, two, three, or more than three poles. The system may also include an anti-aliasing model containing executable instructions that, when executed by a processor, can be configured to select one or more regions in a trace that deviate from the phase trajectory. The processor can be configured to select one, two, three, four, five, or more than five regions in the trace that deviate from the phase trajectory. The system may also include a signal trace model containing executable instructions that, when executed by a processor, can be configured to exclude one or more poles in one or more regions of the trace selected by the anti-aliasing model. The signal trace model can be configured to exclude one or more poles at least partially based on the influence of one or more characteristics of the one or more poles. The signal trace model can also be configured to select one or more alternative poles of one or more alternative signal traces to replace the excluded one or more poles. In some cases, the signal trace model can be configured to select one or more alternative poles at least partially based on the influence of one or more characteristics of the one or more alternative poles. In some cases, the signal trace model can be configured to select one, two, three, or more than three alternative poles. In some cases, the system may also include an output generator configured to generate an improved signal trace of the tissue location from one or more alternative poles output from the anti-aliasing model.

[0233] In some cases, one or more features may include the radius of one or more poles to the center of the complex unit circle, the angle between one or more poles and the center of the complex unit circle, the angular velocity between one or more poles, the complex plane distance between one or more poles, the change in velocity between one or more poles, the change in inertial velocity between one or more poles, or the momentum of one or more poles, or any combination thereof.

[0234] Figure 17 In some cases, selecting the first pole (referred to as "pole 1" in this paper) in autoregression may be disadvantageous; however, it may not be the pole that returns the minimum absolute error. This paper discloses an improved method for pole selection in autoregression. For example, Figures 23A-23C An exemplary velocity and position trajectory is demonstrated when pole 1 of the AR model is always selected. In some cases, discontinuities may occur in the velocity and position trajectories. Figure 23A ).

[0235] Figure 23B , Figure 23C and Figure 23A Examples of regions showing discontinuities in the Pole 1 trace across three different pulse periods, namely 251-500, 751-1000, and 1001-1250, are shown.

[0236] Figure 23A (Left) shows an exemplary pole 1 location trace drawn on the RF phase trajectory and the pole 1 location trace after removing discontinuous poles. Figure 24A (Right) shows exemplary autoregressive poles 1, 2, and 3 plotted on the complex unit circle. For example, different poles can be chosen when pole 1 is discontinuous.

[0237] The decision to select pole 1, pole 2, or pole 3 can be made using one or more of the discriminators listed in Tables 2 and 3 below.

[0238] Table 2: Discriminator of AR Pole Selection Algorithm

[0239] Table 3: Description of the AR Pole Selection Algorithm

[0240] Figure 24B , Figure 24C , Figure 24D , Figure 24E and Computer system Five examples of implementing the AR pole selection algorithm described above are shown.

[0241] Figure 25 This disclosure provides computer systems programmed to implement the methods of this disclosure. Method and system applicationsA computer system 2501 is illustrated, which is programmed or otherwise configured to perform methods for determining the time-dependent location of tissues as disclosed herein. The computer system 2501 can adjust various aspects of the otoscope and system disclosed herein for determining the time-dependent location of tissues, such as sending stimulation, sending and / or receiving ultrasound signals, and performing various data analysis operations and sub-operations. The computer system 2501 can be a user's electronic device or a computer system remotely located relative to an electronic device. The electronic device can be a mobile electronic device.

[0242] Computer system 2501 includes a central processing unit (CPU, also referred to herein as a “processor” and “computer processor”) 2505, which may be a single-core or multi-core processor, or multiple processors for parallel processing. Computer system 2501 also includes memory or storage location 2510 (e.g., random access memory, read-only memory, flash memory), electronic storage unit 2515 (e.g., hard disk), communication interface 2520 for communicating with one or more other systems (e.g., network adapter), and peripheral devices 2525 (such as cache, other memory, data memory, and / or electronic display adapter). Memory 2510, storage unit 2515, interface 2520, and peripheral devices 2525 communicate with CPU 2505 via a communication bus (solid line) such as a motherboard. Storage unit 2515 may be a data storage unit (or data warehouse) for storing data. Computer system 2501 may be operatively coupled to computer network (“network”) 2530 with the aid of communication interface 2520. Network 2530 may be the Internet, the Internet, and / or an extranet, or an intranet and / or extranet communicating with the Internet. In some cases, network 2530 is a telecommunications and / or data network. Network 2530 may include one or more computer servers capable of enabling distributed computing such as cloud computing. In some cases, with the assistance of computer system 2501, network 2530 may implement a peer-to-peer network that enables devices coupled to computer system 2501 to act as clients or servers.

[0243] CPU 2505 can execute a series of machine-readable instructions, which can be embodied in a program or software. The instructions can be stored in a memory location such as memory 2510. The instructions can be directed to CPU 2505, which can then program or otherwise configure CPU 2505 to implement the methods of this disclosure. Examples of operations performed by CPU 2505 can include fetching, decoding, executing, and writing back.

[0244] CPU 2505 may be part of a circuit, such as an integrated circuit. One or more other components of system 2501 may be included in this circuit. In some cases, this circuit is an application-specific integrated circuit (ASIC).

[0245] Storage unit 2515 may store files, such as drivers, libraries, and saved programs. Storage unit 2515 may store user data, such as user preferences and user programs. In some cases, computer system 2501 may include one or more additional data storage units located outside computer system 2501, such as on a remote server located in communication with computer system 2501 via an intranet or the Internet.

[0246] Computer system 2501 can communicate with one or more remote computer systems via network 2530. For example, computer system 2501 can communicate with a user's remote computer system (e.g., providing processed data, providing updates, providing instructions, troubleshooting, etc.). Examples of remote computer systems include personal computers (e.g., portable PCs), tablet PCs (e.g., Apple® iPad, Samsung® Galaxy Tab), telephones, smartphones (e.g., Apple® iPhone, Android-enabled devices, Blackberry®), or personal digital assistants. Users can access computer system 2501 via network 2530.

[0247] The methods described herein can be implemented using machine-executable code (e.g., a computer processor) stored in an electronic storage location of computer system 2501 (such as storage in memory 2510 or electronic storage unit 2515). The machine-executable or machine-readable code can be provided in software form. During use, the code can be executed by processor 2505. In some cases, the code can be retrieved from storage unit 2515 and stored in memory 2510 for easy access by processor 2505. In some cases, electronic storage unit 2515 can be excluded, and machine-executable instructions can be stored in memory 2510.

[0248] The code can be pre-compiled and configured for use with a machine having a processor suitable for executing the code, or the code can be compiled during runtime. The code can be supplied in a programming language, which can be selected to enable the code to be executed in a pre-compiled or compiled manner.

[0249] Aspects of the systems and methods provided herein (such as computer system 2501) can be embodied in programming. These aspects of the technology can be considered "products" or "manufactured goods," which are typically in the form of machine (or processor) executable code and / or associated data carried on or embodied in a type of machine-readable medium. Machine-executable code can be stored on electronic storage units such as memory (e.g., read-only memory, random access memory, flash memory) or hard disks. "Storage" type media can include any or all tangible memory of a computer, processor, etc., or its associated modules (such as various semiconductor memories, tape drives, disk drives, etc.), which can provide non-temporary storage for software programming at any time. All or part of the software can sometimes be communicated via the Internet or various other telecommunications networks. For example, such communication enables the loading of software from one computer or processor into another computer or processor, for example, from a management server or host computer into a computer platform for an application server. Therefore, another type of medium that can carry software elements includes physical interfaces between local devices, light waves, radio waves, and electromagnetic waves used via wired and optical ground networks, and via various air links. Physical elements carrying such waves (such as wired or wireless links, optical links, etc.) can also be considered as media carrying software. As used herein, unless limited to non-transitory, tangible "storage" media, terms such as "computer or machine-readable medium" refer to any medium involved in providing instructions to a processor for execution.

[0250] Therefore, machine-readable media (such as computer-executable code) can take various forms, including but not limited to tangible storage media, carrier media, or physical transmission media. Non-volatile storage media include, for example, optical discs or disks (such as any storage device in any computer), such as those used to implement the database shown in the accompanying drawings. Volatile storage media include dynamic memory, such as the main memory of such a computer platform. Tangible transmission media include: coaxial cables; copper wires and optical fibers, which include conductors that form a bus within a computer system. Carrier transmission media can take the form of electrical or electromagnetic signals, or the form of sound or light waves (such as sound or light waves generated during radio frequency (RF) and infrared (IR) data communication). Therefore, common forms of computer-readable media include, for example, floppy disks, flexible disks, hard disks, magnetic tapes, any other magnetic media, CD-ROMs, DVDs or DVD-ROMs, any other optical media, punched card tapes, any other physical storage media with a perforated pattern, RAM, ROM, PROM and EPROM, FLASH-EPROM, any other memory chips or cartridges, carrier waves for transmitting data or instructions, cables or links for transmitting such carrier waves, or any other medium from which a computer may read programming code and / or data. Many of these forms of computer-readable media may involve loading one or more sequences of one or more instructions onto a processor for execution.

[0251] Computer system 2501 may include or communicate with an electronic display 2535, the electronic display 2535 including a user interface (UI) 2540 for providing indications of, for example, the state or condition of the tympanic membrane, one or more traces related to tissue location or movement, etc. Examples of UI include, but are not limited to, graphical user interfaces (GUIs) and web-based user interfaces.

[0252] The methods and systems disclosed herein can be implemented by one or more algorithms. The algorithms can be implemented in software when executed by the central processing unit 2505. For example, the algorithm may compute a covariance matrix from ultrasound data from a set of depths and across multiple pulse cycles of ultrasound data; compute one or more eigenvalues ​​of the covariance matrix; analyze the frequency content of the ultrasound data across multiple pulse cycles; use the frequency content to compute a second trace of tissue location; determine one or more regions where the first trace contains non-physical tissue location or movement; replace the first trace with at least a portion of the second trace in one or more regions; and so on.

[0253] Although preferred embodiments of the invention have been shown and described herein, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. The invention is not intended to be limited to the specific examples provided herein. While the invention has been described with reference to the foregoing description, the description and illustration of embodiments herein are not intended to be construed as limiting. Numerous variations, modifications, and alternatives will now occur to those skilled in the art without departing from the invention. Furthermore, it should be understood that all aspects of the invention are not limited to the specific depictions, configurations, or relative proportions described herein, and depend on a variety of conditions and variables. It should be understood that various alternatives to the embodiments of the invention described herein may be employed in carrying out the invention. Therefore, it is contemplated that the invention should also cover any such alternatives, modifications, variations, or equivalents. The appended claims are intended to define the scope of the invention and thereby cover the methods and structures within the scope of these claims and their equivalents.

[0254] Figure 29A The methods and systems disclosed herein can be used to characterize many biological tissues to provide a variety of diagnostic information. Biological tissues may include patient organs. A speculum may be placed within a body cavity to characterize patient tissues. For example, patient organs or body cavities may include muscles, tendons, ligaments, mouth, tongue, pharynx, esophagus, stomach, intestines, anus, liver, gallbladder, pancreas, nose, larynx, trachea, lungs, kidneys, bladder, urethra, uterus, vagina, ovaries, testes, prostate, heart, arteries, veins, spleen, glands, brain, spinal cord, nerves, etc.

[0255] The methods and systems disclosed herein can be used to classify tympanic membranes. For example, membranes can be classified to determine ear conditions such as acute otitis media (AOM), chronic otitis media, otitis media with effusion, and / or chronic suppurative otitis media. Classification of an ear exhibiting AOM may include detecting the presence of effusion and characterizing the effusion type as serous, mucoid, purulent, or a combination thereof. In AOM, middle ear effusion (MEE) can be caused by infectious agents and can be thin or serous in viral infections, while becoming thicker and purulent in bacterial infections. Therefore, determining the various properties of the fluid adjacent to the tympanic membrane can provide information that can be used to characterize the membrane. For example, an ultrasonic transducer can be mounted on a speculum and positioned within the ear canal. An excitation generator can apply impact pressure to the tympanic membrane, the transducer can guide ultrasound to the tympanic membrane, and the reflected ultrasonic energy can be measured from the surface of the tympanic membrane. The phase change of reflected ultrasound during and / or after the application of non-contact excitation can indicate elasticity, which can be correlated with the type of fluid behind the tympanic membrane; for example, air indicates ear health, clear fluid indicates viral infection, or turbid fluid indicates bacterial infection.

[0256] Figure 29B A side cross-sectional view of an otoscope 650 placed inside an ear 651 is shown according to some embodiments. Figure 29A A front cross-sectional view of an otoscope 650 of this disclosure according to some embodiments is shown. Region 150 (as enlarged view) Figure 29B The diagram shows a cross-sectional view of the middle ear and tympanic membrane 130 of the subject being examined. The tympanic membrane 130 can be probed by an ultrasonic beam 128 from an ultrasonic transducer. The transducer can be mounted on the inner surface of the endoscopic tip 124. The endoscopic tip can be detached from the otoscope 650 via an endoscopic mounting adapter 126. The endoscopic tip can be operatively coupled to or include an excitation generator. In some cases, the excitation generator can generate a pressure excitation. In some cases, the excitation generator can generate a pressure excitation that is a sonic excitation, a subsonic excitation, or a supersonic excitation. The pressure excitation generated by the excitation generator can be an impact step or Δ (impact) generation, a sinusoidal pressure excitation, a square wave excitation, or any combination of these excitations, and the excitation can be gated burst or continuous. The pressure excitation may or may not be provided with static positive and negative pressure bias.

[0257] In some cases, the excitation generator produces pressure excitation, such as airflow. For example, the otoscope mounting adapter 126 and the spectral tip 124 may have a common internal volume. This common internal volume can provide dynamic pressure from the excitation generator coupled to the ear canal via coupler 122, where air pressure causes displacement of the tympanic membrane 130. The excitation generator can generate pressure changes that are coupled into the ear canal via the spectral tip 126.

[0258] In some cases, the excitation generator can be an airbag operated by an operator to apply force to a membrane or surface; it is an air displacement generator that produces alternating pressure, stepped pressure, or jets. The excitation generator output may or may not be sealed to the peripheral area of ​​the surface and may use an ejection gas such as atmospheric air or other suitable gas.

[0259] In some cases, the excitation generator can produce sound, subsonic, or supersonic excitations. For example, the excitation generator can produce sub-audio frequencies below 20 Hz, audio frequencies from 20 Hz to 20 kHz, or ultrasonic frequencies above 20 kHz. In one case, sound, subsonic, or supersonic excitations can be generated by a piezoelectric transducer. A piezoelectric transducer converts an electrical signal into a physical displacement, which in turn induces a pressure wave. In another case, sound, subsonic, or supersonic excitations can be generated by a cMUT transducer. In yet another case, an audio loudspeaker with a voice coil actuator can be used to generate the excitation.

[0260] An ultrasonic transducer, different from and other than the excitation generator, may be housed inside the otoscope. This ultrasonic transducer may include any transducer element disclosed herein, or a variation, example, or embodiment of an ultrasonic transducer. In some cases, the ultrasonic transducer and the excitation generator may be the same element.

[0261] like Figure 30A As shown, the otoscope may include a handle 601 for positioning the endoscope 602. The otoscope may include a video display 603. The display may show the user an optical image of the membrane to be characterized. The display 603 may display ultrasound images. The display 603 may provide a user interface therein for controlling various aspects of the otoscope 650 and / or the analysis of ultrasound data. The otoscope may include an onboard digital processing unit, for example, within the handle 601 of the device. The otoscope may be connected to a remote device, such as a server, remote storage, remote processing unit, etc. The analysis of ultrasound data may be performed on-board or remotely.

[0262] Figure 30B A side cross-sectional view of a viewer 701 according to some embodiments is shown. Example 1: Covariance matrix analysis across phase trajectories A cross-sectional view of the trocar tip 702 according to some embodiments is shown. In some cases, an ultrasonic transducer 703 is housed within the trocar 701. The trocar may be disposable. The trocar may include an ultrasonic transducer 703 positioned near the tip region 702 of the trocar 701. The trocar may include a lens assembly 704 that may help provide the user with an optical image to guide the positioning of the ultrasonic transducer. In some cases, the ultrasonic transducer 703 may be located at the center of the trocar, so that optical sensing is performed around the ultrasonic transducer 704. The ultrasonic transducer may be supported by a mesh 705. The mesh 705 may allow electrical signals to be transmitted to a digital processing device.

[0263] The ultrasonic transducer described herein may include a base 706. The base may be mounted within an otoscope via a plate 704. The plate 704 may allow the transducer to be centered or nearly centered within the opening of the otoscope. The plate 704 may be optically transparent. In one case, the plate 704 is glass. The plate 704 may include one or more openings that may allow pressure excitation to be transmitted from the interior of the otoscope tip to the exterior of the otoscope tip. The plate 704 may include one or more conductive portions to allow the supply of drive voltage and / or current to the transducer. The plate 704 may include one or more insulating layers. The plate 704 itself may include conductive or insulating portions. The plate 704 may be insulating, with conductive portions mounted thereon.

[0264] In another context, the methods and systems disclosed herein can be used to characterize animal or human organs, such as the eye. For example, an excitation generator can apply impact pressure to the eye, a transducer can direct ultrasound to the eye, and the reflected ultrasound energy can be measured from the eye surface. The phase change of the reflected ultrasound during and / or after the application of non-contact excitation can indicate elasticity, which can be correlated with interocular pressure used to measure or diagnose glaucoma.

[0265] In another context, the methods and systems disclosed herein can be used to characterize the lungs of animals or humans. For example, audio frequencies (e.g., 3-20 Hz) from the chest can be demodulated from a transducer. The transducer can be integrated into a stethoscope-like device in which the transducer can be moved upward along the chest during a "percussion test" (auscultation) to identify changes in reflected ultrasound, which can indicate fluids (e.g., mucus or water) in the lungs. In some cases, multiple transducers or transducer arrays can be provided and placed or worn on the chest. Phase changes in the reflected ultrasound during application can indicate changes in fluid viscosity, which can be correlated with lung diseases such as pneumonia, lung cancer, chronic obstructive pulmonary disease (COPD), idiopathic pulmonary fibrosis (IPF), etc.

[0266] The methods and systems disclosed herein can be used, for example, to characterize food. For instance, an excitation generator can apply impact pressure to the surface of a food (such as a vegetable or fruit) and can apply ultrasonic energy to the food to measure the time-dependent surface response of the fruit or vegetable, thereby determining elasticity or other physical properties that may be correlated with the ripeness of the fruit or vegetable. For example, food can be placed in a support and its surface can be excited with an ejected gas such as air, and the surface deflection response can be used to estimate ripeness or other properties. For example, the excitation can be a gas that can be transmitted at supersonic speeds and / or at a grazing angle to the food surface, or one or more foods can be placed in a chamber with variable pressure to measure low-frequency surface responses to pressure, such as deflection relative to pressure. For example, the excitation can be applied to a surface and the response measured on different surfaces of the same article, such as measuring propagating surface waves or shear waves passing through the object being characterized.

[0267] The methods and systems disclosed herein can be used to characterize industrial processes. For example, the small size of the transducers disclosed herein allows application to any industrial process where larger transducers, ultrasound, or other modalities (such as LiDAR) are disabled due to their large size. The high resolution achieved by the transducers of this disclosure over a short range (e.g., less than 25-35 mm) with a movement of 10-20 micrometers (e.g., via Doppler integration) allows the invention to be applied to a variety of industrial processes where analysis must be performed without physical contact with the analyte. For example, an excitation generator can apply impact pressure to the surface of a manufactured part, for example, to determine the consistency of a viscous fluid (such as a lubricant), and ultrasonic energy can be applied to the part to measure the time-dependent surface response of the viscous fluid, thereby determining elastic or other physical properties that may be correlated with the quality of the lubricant. The transducer can be used to measure paint thickness by comparing painted portions of an object with unpainted sections. The transducer can be used to measure whether a painted object is dry by comparing it with a similar object recently painted with the same paint. Transducers can be used as part of a manufacturing process to identify objects as part of a count of manufactured objects. Transducers can be used to measure changes in the density or composition of an object by comparing it to an object that has undergone a process with an object prior to that process (e.g., cooked food, pickling, etc.). Other industrial examples can include ranging applications, ultrasonic time-of-passage gas flow meters for measuring dynamic gas flow rates, wind speed measurement applications, and a variety of other ultrasound-based sensing applications.

[0268] The methods and systems disclosed herein may include using an ultrasonic transducer; pointing the tip of a probe into a lumen adjacent to the membrane; directing a disturbance to the surface of the membrane; measuring reflected ultrasonic signals from the membrane surface; and characterizing the membrane’s viscosity or elasticity and reflected ultrasound in response to the disturbance. Example

[0269] Figure 26 In some unrestricted instances, the covariance matrix is ​​used to determine the time-dependent location of the tissue response to the stimulus 2600 ( Figure 31 In some unrestricted instances, the covariance matrix is ​​used to determine the time-dependent location 3100 of the tissue response to the pneumatic excitation generated by the otoscope 3104. ω ).

[0270] Ultrasonic data derived from the ultrasound waves reflected from tissue is used for analysis 2601. From the ultrasound data, a covariance matrix is ​​calculated from a set of depths and over multiple pulse periods of the ultrasound data, and one or more eigenvectors 2602 are calculated from the covariance matrix. The covariance matrix is ​​constructed by acquiring RF ultrasound samples with Hilbert transform within the clear echo region of the ultrasound data. The covariance matrix is ​​analyzed using the Hilbert transform; for example, the analytical pulse-echo signal obtained from the Hilbert transform of a single pulse period (indicated by "k") is represented as:

[0271] Where R is the amplitude. φ =2 πf For carrier frequency, p,q,m Let z be the signal phase, and z be the depth corresponding to the m-th a / d sample. Analytical pulse echo signals are used to determine the elements of the covariance matrix. The covariance matrix originates from the same depth but from two distinct pulse periods. These pulse periods are each a known number of pulse periods in pulse period k and are denoted by exponents p and q. The covariance matrix is ​​generated by adding an arithmetic term from a fixed depth z or m. The covariance matrix can be an outer product matrix described by the following equation:

[0272] Where p = {0...N-1} and q = {0...N-1}, and the apostrophe indicates complex conjugation.

[0273] The complete covariance matrix is ​​given by K(m = {m0, m1, m2, ..., mR} from a set of adjacent depths m = {m0, m1, m2, ..., mR}. ν Summing is performed. Substituting H into the covariance expression above, we get the following formula:

[0274] It can be further simplified to: .

[0275] The covariance matrix resulting from the contributions across a series of consecutive depths spanning a specific tympanic membrane echo is expressed by the following equation:

[0276] It can be written more concisely as

[0277] in

[0278] And Δ is at a depth m at two different times. p and qThe phase difference of the samples. This is also known as the rate of change of the phase of the ultrasound signal over time. This avoids demodulation because, in the initial stages of describing K, the carrier frequency is effectively subtracted by multiplying one exponential term by another exponential term that has been conjugate.

[0279] The covariance matrix is ​​calculated based on the mathematical description above. For the resulting matrix, the eigenvectors and eigenvalues ​​are represented by the following formula:

[0280] in , ν 1 , ν 2 … ν 3 λ N Let be the feature vector set, and ( , λ 1 , λ 2 ν 3 … ) is the characteristic value.

[0281] The overall signal energy is concentrated in only a few eigenvectors. Eigenvectors { , ν 1 , ν 2 … ν 3 Example 2: Using autoregressive deconstruction of signals N The set constitutes a normalized orthogonal basis set; the ratio of each eigenvalue to the sum of all eigenvalues ​​is the fraction of the signal class energy represented in the relevant eigenvector. In the case of signals from the tympanic membrane, we observe eigenvalue behavior and find that the overall eigenvalue energy (>99%) is contained within a single eigenvector, and the energy fraction contained in the first eigenvector decreases simultaneously at the location of signal energy depletion. Specifically, the target signal quality ratio is expressed as: .

[0282] Within the range of pulse cycles included in the covariance calculation, M is estimated at different depths along the ultrasound beam. When no target is present, M can be a value between 0 and 1. In some cases, the variance of M can be higher compared to when a target is present.

[0283] The behavior of the noisy process covariance (without a target) necessitates the use of numerous covariance eigenvectors to describe the diverse motions observed throughout the entire sample volume. This contrasts with the behavior of the target covariance (with a target), which requires only one or a few eigenvectors to describe a limited number of motions observed throughout the entire sample volume. The principal eigenvector (maximum λ) contains a description of the target motion, which has contributions from all a / d samples within the target echo. Target motion information is embedded in the covariance principal eigenvector. Eigenvector analysis involves repeated analysis in 10-millisecond (ms) steps within an adjustable 20 ms analysis window. Analyzing the eigenvectors also includes analyzing depth range values ​​and creating a set of principal eigenvectors at each 10-millisecond (ms) step.

[0284] The set of principal feature vectors (one principal feature vector per window position) obtained at each of 100 windows within a 1-second time range, incremented by 10 ms, is represented as follows: .

[0285] The membrane motion information contained in the eigenvector is obtained by the following expression:

[0286] in U It is the "phase unwrapping operator". This expression indicates that the phase of the unwrapped eigenvector is multiplied by a wavelength of the carrier frequency for each 4π radian phase change. It is a vector that has a value for each pulse cycle included in the covariance analysis window.

[0287] Similar to the eigenvector set mentioned above, there exists a displacement vector set, which can be represented by the following formula: .

[0288] A vector set D containing positional changes during each covariance calculation time interval shows the relative positional changes produced by the tympanic membrane and is “stitched” to its adjacent segments, indicating tympanic membrane movement. A first trace 2603 is then determined to be the tissue location associated with the contents of one or more eigenvalues.

[0289] Figure 27 In some unrestricted instances, autoregressive (AR) models are used to estimate the frequency spectrum of the signal trace.2701 Autoregression is applied to the eigenvectors of the covariance analysis to estimate the frequency spectrum of the signal trace. Example 3: Selecting AR polesFor example, the autoregression is third-order autoregression, or in some unrestricted instances, higher-order autoregression. One or more autoregressive models are applied to generate one or more poles, and one or more features of the poles (e.g., calculated frequency characteristics) are used to exclude or select one or more poles. The pulse repetition frequency is, for example, 5 kHz. The offset frequency of the ultrasound data is, for example, + / - 2.5 kHz. The angle of the complex unit circle representing the ultrasound offset frequency is, for example, + / - π. The angle also represents, for example, the velocity of sound in air, + / - 227.9 mm / s. The velocity estimate is calculated by multiplying the radian angle of each pole on the complex unit circle by 72.54 (mm / s) / radians. The variance of the estimated frequency spectrum is determined using the radius of each pole.

[0290] The variance of the estimated frequency spectrum is calculated using the following equation:

[0291] Where T is the pulse period and r is the pole radius.

[0292] A third-order autoregressive model was applied to the first eigenvector of the covariance from Example 1. Pole 1 was determined to primarily estimate the low-frequency velocity of the ultrasound data, and poles 2 and 3 were found to primarily estimate the velocity from the high-frequency ultrasound data.

[0293] The third-order autoregressive poles are obtained by taking the root of the denominator of the third-order estimate of the power spectral density using the following power spectral density equation:

[0294] Use the zeros of the third-order polynomial in the denominator of the power spectral density equation with the poles of the following third-order equation:

[0295]

[0296]

[0297]

[0298]

[0299] The poles of the third-order equation are used to determine the frequency estimation of the ultrasound data, as well as the poles of the third-order autoregression.

[0300] The poles of the third-order equation in the autoregressive model are applied to the first eigenvector of the covariance result to obtain three poles. For example, pole 1 is a frequency estimator with low-pass filtering characteristics. For example, poles 2 and 3 are frequency estimators with high-pass filtering characteristics.

[0301] The velocity over time is determined by plotting the angle of a given pole across all pulse cycles. The pole angle is calculated by taking the arctangent of the ratio of the imaginary part of the given pole to the real part of the given pole. The poles of the third-order equation and the power spectral density equation are applied to convert the angle of the given pole into velocity. For example, the velocity estimated through pole 1 is integrated over time to obtain the displacement. The pole 1 location trace is compared with the covariance location trace to determine the magnitude of the displacement. One or more poles are selected or excluded to generate an optimized signal trace 2703 for the tissue location from the remaining or selected poles.

[0302] Figure 28 In a non-restrictive instance, a pole (e.g., pole 1) is found to be discontinuous, and another pole is chosen to improve the trajectory. Figure 28 Multiple autoregressive algorithms are applied to select the poles. Figure 28 Ultrasonic measurements, for example, consist of 5000 pulse cycles. The pulse repetition frequency is, for example, 5 kHz. The total measurement time is one second. A 12-point autoregressive analysis is applied to the first 4800 pulse cycles, excluding the data from the last 200 pulse cycles. The first pulse cycle (e.g., pole 1) is automatically selected. The remaining pulse cycles, for example, have poles 1, 2, and 3 selected as poles. One or more discriminators listed in Tables 2 and 3 are applied to each pole in each of the 4799 pulse cycles following the first pulse cycle. A predictive anti-aliasing model is applied to select regions in the trace that do not follow the desired phase trajectory. Example 4: Anti-aliasing method operating on covariance and eigenvector motion detection ).

[0303] Figure 28 In some non-limiting instances, autoregression is applied to regions of covariance error leakage pressure transitions in the trace to create anti-aliasing repair traces. Figure 28 Determine the difference between the absolute errors of the covariance velocity trajectory and the autoregressive trajectory. In regions where the difference in absolute error is greater than π / 3 and the velocities have opposite signs, replace the covariance displacement trajectory with the autoregressive trajectory. Figure 28 The output is an optimized signal trace from the tissue location of the poles in the anti-aliasing model. Figure 21 ).

[0304] Measure the absolute value of the error between the autoregression rate and the covariance rate, and record it in the anti-aliasing repair plot. ​ ).

Claims

1. A method for determining the time-dependent location of an organization's response to a stimulus, the method comprising: (a) Receiving ultrasound data, wherein the ultrasound data originates from ultrasound waves reflected from the tissue; (b) Calculate the covariance matrix from the ultrasound data from a set of depths and for multiple pulse periods of the ultrasound data; (c) Calculate one or more eigenvectors and related eigenvalues ​​of the covariance matrix; as well as (d) Determine a first trace of the organizational location associated with the content of the principal eigenvector among the one or more eigenvectors.

2. The method according to claim 1, further comprising: (i) Analyze the frequency content of the ultrasound data across the multiple pulse cycles; as well as (ii) Calculate a second trace of the tissue location based at least in part on the frequency content.

3. The method of claim 1, further comprising outputting an indication of a disease state, a healthy state, or an undetermined state of the tissue in response to the second trace at the tissue location.

4. The method according to claim 1, wherein the plurality of pulse cycles are sequential.

5. The method according to claim 1, wherein the tissue is a tympanic membrane.

6. The method of claim 1, further comprising outputting an indication of a disease state, a healthy state, or an undetermined state of the tissue in response to the first trace at the tissue location.

7. The method of claim 1, wherein each of the plurality of pulse periods is associated with a correlation covariance matrix from a plurality of correlation covariance matrices, and wherein the method further comprises calculating a set of displacement vectors of the plurality of correlation covariance matrices.

8. The method of claim 7, wherein the plurality of correlated covariance matrices are associated with the pulse periods of the ultrasound data across consecutive adjacent depths.

9. The method of claim 7, wherein the displacement vector set of the plurality of correlated covariance matrices is calculated based on the target quality signal ratio.

10. The method of claim 9, wherein the target quality signal ratio is a value between 0.00 and 1.

00.

11. The method of claim 9, further comprising splicing together segments of the displacement vector set to generate a time trace of the tissue location, wherein the time trace of the tissue location is the first trace.

12. The method according to claim 1, further comprising: Analyze one or more eigenvectors of the covariance matrix, wherein each of the one or more eigenvectors contains an orbital rotation associated with the phase of the ultrasound data.

13. The method of claim 12, further comprising: A bias is applied to each of the one or more eigenvectors such that the eigenvectors orbit the origin in phase space; as well as The second trace of the tissue location is calculated using the eigenvectors of the obtained bias.

14. The method according to claim 12 or 13, wherein analyzing the one or more feature vectors further includes analyzing membrane motion information of the one or more feature vectors.

15. The method of claim 14, wherein analyzing the one or more feature vectors further comprises repeating the analysis in an adjustable 20 ms analysis window at 10 ms steps.

16. The method of claim 15, wherein analyzing the one or more feature vectors further includes analyzing depth range values.

17. The method of claim 15, wherein analyzing the one or more feature vectors further comprises creating a set of principal feature vectors at each 10ms step.

18. The method of claim 15, wherein analyzing the one or more feature vectors further comprises creating a master feature vector set within each 20ms analysis window.

19. The method according to claim 1, wherein the stimulus is a pneumatic excitation.

20. The method of claim 1, wherein the pneumatic actuation is jet propulsion.

21. The method according to claim 2, further comprising: Determine that the first trace contains one or more regions that are not physically located or moved; as well as In one or more regions, the first trace is replaced by at least a portion of the second trace.

22. The method of claim 2, wherein (i) includes utilizing autoregression.

23. The method of claim 22, wherein the autoregression is third-order or higher-order autoregression.

24. A method for improving the accuracy of time-dependent location determination of a tissue response to a stimulus, the method comprising: Autoregression is applied to the signal trace to generate multiple poles; Use the characteristics of the plurality of poles to exclude or select poles among the plurality of poles; as well as An improved signal trace for the tissue location is generated from the remaining or selected poles among the plurality of poles.

25. The method of claim 24, wherein the autoregression is applied to one or more points of the signal trace.

26. The method of claim 24, wherein the autoregression is third-order or higher-order autoregression.

27. The method of claim 26, wherein the order of the autoregression corresponds to the number of separate frequency peaks of the signal trace.

28. The method of claim 24, wherein the signal trace comprises a covariance motion detection trace.

29. The method of claim 28, wherein the autoregression is applied to one or more feature vectors of the covariance motion detection trace.

30. The method of claim 24 or claim 28, further comprising estimating the average frequency of the signal trace using the angle value between each of the plurality of poles and the center of the complex unit circle of the complex plane.

31. The method of claim 30, wherein the angle value is calculated by taking the arctangent of the ratio of the imaginary part through the pole to the real part through the pole.

32. The method of claim 24 or claim 28, further comprising using the radius value between each of the plurality of poles and the center of the complex unit circle of the complex plane to estimate the variance of the frequency of the signal trace or the velocity represented by the signal trace.

33. The method of claim 32 further comprises drawing a velocity trace of the signal trace over time by plotting the angle between the plurality of poles and the center of the complex unit circle of the complex plane.

34. The method of claim 33, further comprising calculating one or more displacement values ​​between the position of the velocity trace and the position of the signal trace.

35. The method of claim 34, wherein determining the one or more displacement values ​​comprises integrating the velocity in the velocity trajectory.

36. The method according to any one of claims 24 to 35, wherein the features of the plurality of poles further include the radius, the angle, the velocity, the complex plane distance between the plurality of poles, the velocity variation between the plurality of poles, the inertial velocity variation between one or more of the poles, or the momentum of the plurality of poles, or any combination thereof.

37. The method of claim 24, wherein the plurality of poles are divided into one or more frequency groups based on frequency.

38. The method of claim 24, wherein generating the optimized signal trace further comprises replacing the plurality of poles of the signal trace with a plurality of poles of different frequency groups.

39. The method of claim 24, wherein when the absolute error between one or more of the features of the plurality of poles is higher than a threshold, the plurality of poles of the signal trace are replaced.

40. The method of claim 39, wherein the threshold is π / 3 and the poles have opposite signs.

41. A system for determining the time-dependent location of an organization's response to a stimulus, the system comprising a processor, the processor containing executable instructions stored thereon, the executable instructions being configured, upon execution, to: (a) Receiving ultrasound data, wherein the ultrasound data originates from ultrasound waves reflected from the tissue; (b) Calculate the covariance matrix from the ultrasound data from a set of depths and for multiple pulse periods of the ultrasound data; (c) Calculate one or more eigenvectors and related eigenvalues ​​of the covariance matrix; as well as (d) Determine a first trace of the organizational location associated with the content of the principal eigenvector among the one or more eigenvectors.

42. The system of claim 41, wherein the processor is further configured to: (i) Analyzing the frequency content of the ultrasound data across the multiple pulse cycles; and (ii) Calculate a second trace of the tissue location based at least in part on the frequency content.

43. The system of claim 41, wherein the processor is further configured to output an indication of a disease state, a healthy state, or an undetermined state of the tissue in response to the second trace at the tissue location.

44. The system of claim 41, wherein the system further comprises a pneumatic otoscope.

45. The system of claim 44, wherein the processor is operatively connected to the pneumatic otoscope.

46. ​​The system of claim 41 further includes a capacitively micromachined ultrasonic transducer.

47. The system of claim 46, wherein the processor is operatively connected to the capacitively micromachined ultrasonic transducer.

48. The system of claim 46, wherein the capacitively micromachined ultrasonic transducer is placed inside an otoscope.

49. The system of claim 41, wherein the plurality of pulse cycles are sequential.

50. The system of claim 41, wherein the tissue is a tympanic membrane.

51. The system of claim 41, wherein the processor is further configured to output an indication of a disease state, a healthy state, or an undetermined state of the tissue in response to the first trace at the tissue location.

52. The system of claim 41, wherein each of the plurality of pulse periods is associated with a correlation covariance matrix from a plurality of correlation covariance matrices, and wherein the system further comprises calculating a set of displacement vectors of the plurality of correlation covariance matrices.

53. The system of claim 52, wherein the plurality of correlated covariance matrices are associated with the pulse periods of the ultrasound data across consecutive adjacent depths.

54. The system of claim 53, wherein the set of displacement vectors of the plurality of correlated covariance matrices is calculated based on the target quality signal ratio.

55. The system of claim 54, wherein the target quality signal ratio is a value between 0.00 and 1.

00.

56. The system of claim 54, wherein the processor is further configured to concatenate segments of the displacement vector set to generate a time trace of the tissue location, wherein the time trace of the tissue location is the first trace.

57. The system of claim 41, wherein the processor is further configured to: (i) Analyze one or more eigenvectors of the covariance matrix, wherein each of the one or more eigenvectors contains an orbital rotation associated with the phase of the ultrasound data; (ii) Apply a bias to each of the one or more eigenvectors such that the eigenvectors orbit the origin in phase space; and (iii) Calculate the second trace of the tissue location using the eigenvectors of the obtained bias.

58. The system of claim 57, wherein the processor is further configured to analyze the one or more feature vectors by analyzing membrane motion information of the one or more feature vectors.

59. The system of claim 58, wherein the processor is further configured to repeat the analysis in 10 ms steps within an adjustable 20 ms analysis window.

60. The system of claim 59, wherein the processor is further configured to analyze the one or more feature vectors by analyzing depth range values ​​of the one or more feature vectors.

61. The system of claim 59, wherein the processor is further configured to analyze the one or more feature vectors by generating a set of principal feature vectors at each 10 ms step.

62. The system of claim 59, wherein the processor is further configured to analyze the one or more feature vectors by generating a set of principal feature vectors within each 20 ms analysis window.

63. The system of claim 41, wherein the stimulus is a pneumatic excitation.

64. The system of claim 41, wherein the pneumatic actuation is jet propulsion.

65. The system of claim 41, wherein the processor is further configured to: Determining that the first trace encompasses one or more regions that are not physically located or moved; and In one or more regions, the first trace is replaced by at least a portion of the second trace.

66. The system of claim 41, wherein the processor is further configured to analyze the frequency content of the re-demodulated ultrasound data by applying autoregression.

67. The system of claim 66, wherein the autoregression is third-order or higher-order autoregression.

68. A system for improving the accuracy of time-dependent location determination of an organization's response to a stimulus, the system comprising a processor, the processor containing executable instructions stored thereon, the executable instructions being configured, when executed, to: Autoregression is applied to the signal trace to generate multiple poles; Use the characteristics of the plurality of poles to exclude or select poles among the plurality of poles; as well as An improved signal trace for the tissue location is generated from the remaining or selected poles among the plurality of poles.

69. The system of claim 68, wherein the processor is further configured to apply the autoregression to one or more points of the signal trace.

70. The system of claim 69, wherein the autoregression is third-order or higher-order autoregression.

71. The system of claim 70, wherein the order of the autoregression corresponds to the number of frequency peaks of the separated signal traces.

72. The system of claim 68, wherein the signal trace comprises a covariance motion detection trace.

73. The system of claim 72, wherein the autoregression is applied to one or more feature vectors of the covariance motion detection trace.

74. The system of claim 68 or claim 72, wherein the processor is further configured to estimate the average frequency of the signal trace using the angle value between each of the plurality of poles and the center of the complex unit circle of the complex plane.

75. The system of claim 74, wherein the processor is further configured to calculate the angle value by taking the arctangent of the ratio of the imaginary part of the pole to the real part of the pole.

76. The system of claim 68 or claim 72, wherein the processor is further configured to estimate the variance of the signal trace using the radius value between each of the plurality of poles and the center of the complex unit circle of the complex plane.

77. The system of claim 68 or claim 72, wherein the processor is further configured to estimate the variance of the frequency of the signal trace or the velocity represented by the signal trace using the radius value between each of the plurality of poles and the center of the complex unit circle of the complex plane.

78. The system of claim 77, wherein the processor is further configured to draw a velocity trace of the signal trace over time by plotting the angle between the plurality of poles and the center of the complex unit circle of the complex plane.

79. The system of claim 78, wherein the processor is further configured to calculate one or more displacement values ​​between the position of the velocity trace and the position of the signal trace.

80. The system of claim 79, wherein the processor is further configured to calculate the one or more displacement values ​​by integrating the velocity in the velocity trace.

81. The system according to any one of claims 68 to 80, wherein the features of the plurality of poles further include the radius, the angle, the velocity, the complex plane distance between the plurality of poles, the velocity variation between the plurality of poles, the inertial velocity variation between one or more of the poles, or the momentum of the plurality of poles, or any combination thereof.

82. The system of claim 68, wherein the processor is further configured to divide the plurality of poles into one or more frequency groups based on frequency.

83. The system of claim 68, wherein the processor is further configured to generate the optimized signal trace by replacing the plurality of poles of the signal trace with a plurality of poles of different frequency groups.

84. The system of claim 83, wherein the processor is further configured to replace a plurality of poles of the signal trace when the absolute error between one or more of the features of the plurality of poles is higher than a threshold.

85. The system of claim 84, wherein the threshold is π / 3 and the poles have opposite signs.

86. A system for improving time-dependent location determination, the system comprising: A covariance analysis model, comprising executable instructions that, when executed by a processor, are configured as follows: (a) Derive one or more covariance matrices from the ultrasound data. (b) Calculate one or more eigenvectors of the one or more covariance matrices, and (c) Generate a trace from the one or more feature vectors; One or more autoregressive systems are configured to generate multiple poles; An anti-aliasing model comprising executable instructions that, when executed by a processor, are configured to select one or more regions in the trace that deviate from the phase trajectory; A signal trace model containing executable instructions that, when executed by a processor, are configured as follows: (d) Excluding one or more poles in one or more regions of the trace selected by the anti-aliasing model, wherein the signal trace model is configured to exclude the one or more poles at least in part based on the influence of one or more characteristics of the one or more poles, and (e) Select one or more alternative poles of one or more candidate signal traces to replace the one or more poles excluded in (d), wherein the signal trace model is configured to select the one or more alternative poles at least in part based on the influence of one or more features of the one or more alternative poles; and An output generator is configured to generate improved signal traces of tissue locations from the one or more alternative poles output from the anti-aliasing model.

87. The system of claim 86, wherein the one or more features include the radius of the one or more poles to the center of the complex unit circle, the angle between the one or more poles and the center of the complex unit circle, the velocity of the angle between the one or more poles, the complex plane distance between the one or more poles, the velocity variation between the one or more poles, the inertial velocity variation between one or more of the poles, or the momentum of the one or more poles, or any combination thereof.

Images