Determination of structural properties of object

Abstract

The present invention generally relates to systems and methods for measuring structural characteristics of an object. The object is subjected to an energy application process and provides an objective quantitative measurement of structural characteristics of the object. The system may include a device, such as a percussion instrument, that can be reproducibly placed on a subject undergoing such measurements for reproducible positioning. A system and method are provided for analyzing characteristics of measurements using machine learning to create a system for predicting anomalies from measurements.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This Patent Cooperation Treaty international application claims the benefit of and priority to the following U.S. Provisional Patent Applications: U.S. Provisional Patent Applications Serial Nos. 63 / 313,405, 63 / 313,407, 63 / 313,409. The contents of the claimed U.S. Provisional Patent Applications are incorporated herein by reference in their entirety.

[0003] Copyright Notice

[0004] A portion of the disclosure of this patent document contains material which is subject to copyright protection. The copyright owner has no objection to the facsimile reproduction by anyone of the patent document or the patent disclosure, as it appears in the U.S. Patent and Trademark Office patent file or records, but otherwise reserves all copyright rights whatsoever. Technical Field

[0005] The present invention generally relates to the assessment of structural properties of an object. In particular, the present invention relates to the use of controlled energy application to an object to assist in the assessment of structural properties reflecting the integrity of the object, and the use of machine learning to facilitate the assessment of structural properties reflecting the integrity of the object using controlled energy application to an object. Background Art

[0006] Every object, such as any anatomical or non-anatomical structure, including industrial or mechanical structures, exhibits certain structural properties that may change over time when such structures are used in any way, including those that simply remain in place in the environment. For changes that are easily discernible visually or revealed by simple tests, these changes can be easily measured. However, when these changes are not easily discernible visually or revealed by simple tests, more complex tests are required. Tests to find these changes are important to the health and life of the structure because such changes may eventually develop into a form of defect that cannot be repaired if not checked or treated. In order to determine the characteristics of a structure, a variety of methods can be used, but if these changes are internal, most tests are likely to be destructive or invasive.

[0007] When an object is subjected to a collision force, a stress wave is transmitted through the object. The stress wave causes deformation of the internal structure of the object. As the object deforms, the object acts in part as a shock absorber, dissipating a portion of the mechanical energy associated with the collision. The ability of an object to dissipate mechanical energy (often referred to as the "damping capacity" of the object) depends on several factors, including the type and structural integrity of the material from which the object is formed.

[0008] There are instruments that can measure the damping capacity of an object. An example of such an instrument is described in U.S. Patent No. 6,120,466 ("the '466 patent"). The instrument disclosed in the '466 patent provides an objective quantitative measurement of the damping capacity of an object, called the loss factor. The energy of an elastic wave decays relatively quickly in a material with a higher loss factor, while the energy of an elastic wave decays relatively slowly in a material with a lower loss factor. Summary of the invention

[0009] The present invention relates to a system and method for measuring and evaluating the structural characteristics of an object, whether anatomical or non-anatomical, in a non-invasive manner and / or using non-destructive measurement methods. The structural characteristics of the object can be identified based on the measurements of the same or other objects previously made using the system and captured by the system. The system can include a device capable of applying energy to an object, which can be an anatomical or mechanical object; measuring a response over a period of time, such as energy reflected from the object due to energy application (e.g., tapping the object), or measuring a response such as deceleration information of an energy application tool such as a tapping rod, such as energy return; recording or compiling these measurements for analysis by a computing system; creating a response distribution map, such as an energy return curve or energy return map (ERG), force return map (FRG), displacement return map (FRG) or other physical return value, as a time distribution map or frequency distribution map to evaluate the characteristics of the object being measured. The response can generally be generated by measuring, for example, a force, energy, displacement or other physical return value on a sensing mechanism or element over a period of time.

[0010] In general, the system may include a program logic module that is trained on a large waveform data set to obtain an optimized decomposition of the waveform. In an exemplary aspect of the invention, the waveform may be a measurement signal from a percussion measurement device that percusses an object, such as an ERG, FRG, DRG, or other physical return value signal, or a simulation of these signals, and may generally be grouped together based on at least one common characteristic. The program logic module may then apply an algorithm to form an initial guess that a signal (e.g., a waveform) is decomposed into its component quantum signals, perform optimization to minimize the difference between the initial guess decomposition and the original signal, identify and resolve potential errors or defects in the decomposition, and perform subsequent rounds of guess decomposition and optimization to form an optimized decomposition. Then, by encoding in an algorithm, such as in a machine learning or deep learning algorithm, the optimized decomposition, its properties, and a method for obtaining the optimized decomposition may be incorporated into the system so that the system can more efficiently and accurately decompose new signals encountered, such as signals from a percussion measurement device that generates signals from a physical object.

[0011] In some embodiments, the object can be real, artificial or simulated oral tissue, such as teeth, oral prostheses, appliances, implants or splints, and / or related tissue or orthopedic implants. In general, a percussion measurement device can apply mechanical energy to an object by percussion and measure the force / energy / displacement, etc. returned to the percussion measurement device, for example, by measuring the force, energy, displacement or other physical return value on a sensing mechanism over a period of time. The collected measurement value can also be deceleration information from the energy application tool after the energy application process.

[0012] In other embodiments, the object can be a machine, an industrial structure or a complex, which may include but is not limited to a polymer composite structure, which includes a honeycomb or layered honeycomb or a metal composite structure; an aircraft structure, a car, a ship, a bridge, a tunnel, a train, a building, an industrial structure, which includes but is not limited to a power generation facility, an arched structure or other similar physical structures.

[0013] In general, some sensing mechanisms may be fixedly limited to detecting response signals in a single direction, for example, only in the direction of compression for a piezoelectric force sensor (i.e., because piezoelectric elements generally only produce signals in response to compression). This typically results in at least some loss of signal in a waveform or other energy distribution where the amplitude includes a negative portion relative to some baseline, such as the application of energy to an object caused by a tap, as any response from the object that causes the object to vibrate, move away from the sensing mechanism, or lose mechanical contact or connection, the sensing mechanism may not detect a portion of the signal due to insufficient compression detected by the sensing mechanism in the appropriate direction. As a result, the ERG produced by the device may be partially incomplete.

[0014] For objects that are in their original state or have not been additionally perturbed from their original state, such as teeth or mechanical structures, such as undamaged canines or single-rooted incisors, signals such as ERG, FRG, DRG, etc., including response profiles from the object or deceleration information of an energy application tool over a certain time interval, can generally be in the form of a Gaussian with a single peak, similar to the upper half of a sine wave (i.e., a sine curve or a quasi-sine curve shape). However, for objects such as non-original, cracked or otherwise defective teeth, their response profiles may be different and present differences. For example, objects with different degrees of perturbation may change from a Gaussian distribution with a single peak to multiple peaks or other deviations from a Gaussian distribution. In some cases, the changing response can provide good insight into the type of perturbation that may be related to the defect type, defect location, etc. In other cases, certain response signals may mask certain perturbations, and additional analysis techniques may be required, such as decomposing the actual response signal collected into more basic sub-signals to evaluate the type of perturbation, thereby obtaining knowledge or additional information about the true structural characteristics. For sine waves, there are many known mathematical formulas that can be used to help clarify. However, as mentioned above, the Gaussian shape only resembles the top half of a sine wave, which can make it difficult to extract information using such readily available mathematical formulas. Furthermore, fundamentally, tapping the system may only produce Gaussian curves, which are sine waves with the bottom missing. In some cases, additional processing may also be required to assist in finding the missing bottom half of the simulation.

[0015] The inventors have discovered that using artificial intelligence, a signal can be decomposed into a series of one or more sub-signals. In many cases, these sub-signals can directly correspond to the oscillation and / or resonant frequencies induced in the defect during the percussion process. In addition, the properties of these sub-signals, such as frequency, amplitude, and exponential decay rate, can narrow down or uniquely identify certain physical and / or clinical phenomena.

[0016] In one exemplary aspect of the invention, the system may be adapted to accommodate or analyze signals resulting from measurements of taps on an object, wherein the sensing mechanism causes at least a portion of the return signal to be limited or lost.

[0017] In some exemplary embodiments, the system may utilize machine learning methods to process and / or analyze signals that inherently lack negative amplitude portions, and attempt to reconstruct or treat the signal as a response with the missing portion rather than a complete response. In the dental field, machine learning algorithms and methods may be trained on a large set of collected impact data (e.g., time-energy profiles) from a variety of teeth with different characteristics, such as different types of teeth (e.g., incisors, bicuspids, canines, molars, etc.), size, number of roots, varying degrees of physical damage (e.g., cracks, cavities, etc.), degree or type of restoration (e.g., artificial crowns, fillings, etc.), age, etc., to train the algorithms and methods to decompose newly encountered signals into component sub-signals, such as a collection of sinusoidal sub-signals that form a signal or an approximation thereof.

[0018] The data sets may also be grouped in other ways, such as by the location of the taps on the object (e.g., for buccal or mesial, distal or proximal teeth, etc.), the location of the object relative to other reference points (e.g., mandibular vs. maxillary in the mouth), the amount / frequency / number of taps, or any other appropriate type of grouping. In such an embodiment, the signal may be analyzed and processed as at least one sinusoidal signal with portions of the signal missing from it (e.g., the lower half of the sinusoid below a given threshold is missing from the signal, forming a Gaussian-like shape over the time frame of the measurement). This may allow for better interpretation of the data, as the entire response may be considered (or at least an approximation thereof to account for the missing portions of the signal), rather than treating the response as only the portion that is present in the signal (i.e., assuming the signal is intact).

[0019] In another aspect of the invention, a machine learning algorithm or set of machine learning algorithms and system methods that may be generally separate or disjoint from the machine learning algorithms discussed in connection with signal decomposition or other uses, may be trained on a large collection of signals that may be annotated with features (which may be determined by an "expert" or other trusted characterizer or by a machine learning algorithm) to identify, guess the degree of probability and / or associate a measured signal with a particular feature, or select an appropriate further analysis or algorithm manipulation approach to produce a useful output for a user to utilize or interpret for, e.g., selecting an appropriate further diagnosis, monitoring, treatment plan, etc. Such a machine learning algorithm may be used, for example, to analyze or find correlations between clinical signals and annotated data sets, such as in a heat map or other comparison or analysis approach of a numerical metric.

[0020] In one exemplary aspect of the invention, a signal may be identified, analyzed and / or processed as a sum or aggregation of multiple different sub-signals resulting from the interaction of energy applied to a tooth with structural properties of the tooth and / or surrounding tissue / structure. Without being bound by any particular theory, a signal may generally represent multiple different sub-signals, each generated by a separate physical structure or feature of the tooth or surrounding the tooth, and each sub-signal may generally take the form of an approximation of a sinusoid, such as, but not limited to, a decaying sinusoid (e.g., a sinusoid that decays exponentially in response to dissipation of the material / structure / motion of the object), resulting in a sum or aggregation of sinusoids to form the overall shape of the signal (or the detected portion, i.e., without missing the "bottom" portion below a given threshold for detection or measurement). In some embodiments, a signal may be decomposed into a "sparse" or finite number of sinusoids (i.e., a bounded number of sinusoids to give the original signal or an approximation thereof), without being bound by any particular theory, thereby recognizing that each sinusoid is caused by at least one element of the tooth, its restoration (if any), and / or surrounding tissue.

[0021] In some embodiments, it will generally be recognized that the dominant Gaussian-like form in the signal from a tooth, particularly from an original or undamaged tooth, may be primarily due to the response of the PDL (periodontal ligament) absorbing the kinetic energy from the energy application tool. For an undamaged healthy tooth, the impact energy generated by chewing is attenuated by the PDL at the healthy bone-natural tooth interface. Even in cases where the PDL may be missing or defective, the portion of the signal referred to as the PDL portion will refer to the signal generated due to the direct or indirect anchoring of the tooth or implant in the bone, which will generally produce a pendulum-like response. In general, the PDL portion of the signal may constitute the majority of the amplitude of the signal and can generally be interpreted as a carrier wave that can be used to isolate and / or separate the response from other elements of the tooth or surrounding tissue, for example by subtracting the PDL portion from the signal.

[0022] In some embodiments, the PDL signal may be simply a generated dominant signal with no specific relationship to the PDL.

[0023] In other embodiments, for example, the tap measurement device may measure mechanical devices or other objects, such as dental implants, industrial equipment or devices, etc., and the dominant signal may generally come from, for example, the device being anchored to the surrounding environment to produce a oscillatory response.

[0024] In some exemplary embodiments, performing a sinusoidal decomposition of a signal using machine learning may generally include de-tapering the signal (e.g., removing deformation of the signal near the x-axis of the signal, such as deformation caused by an energy application tool sticking to the tooth), finding initial guesses for larger and / or more pronounced sinusoidal components in the signal (e.g., sinusoids typically generated by a PDL), finding initial guesses for remaining sinusoidal components of the signal (e.g., sinusoids from cracks, damage, delamination, or other features), including guesses for frequencies (e.g., by performing a Fourier transform-like operation, such as a fast Fourier transform (FFT)) on the signal after subtracting an initial PDL sinusoidal guess, performing an optimization to minimize the difference between the composite sum of the original signal and the initial guess sinusoids to generate candidate sinusoidal decompositions, identifying / repairing decomposition defects or errors (e.g., decomposition results that are untrustworthy or negatively indicated) to rerun the above decomposition steps as needed to remove them, and selecting the best or desired candidate from the resulting decompositions. The resulting decomposition and related data / results / visualizations can then be displayed or output in a human-readable form so that they can be used or interpreted by a human physician or other user, e.g., for clinical diagnosis, monitoring, and / or treatment planning. The system-generated decomposition can also typically include determining or calculating uncertainty measures at various steps of the decomposition, such as calculating error values ​​at various steps or specific calculated values ​​in the decomposition.

[0025] In another exemplary aspect of the invention, the system can generate or calculate various numerical metrics based on the decomposition of the signal that show statistically significant differences in the data set, so that these numerical metrics can be used in probability distributions or heat maps to assist in predicting or detecting different physical characteristics or attributes by comparison with numerical metrics derived from the decomposition in a clinical setting. For example, a numerical metric generated from a data set containing known physical characteristics (e.g., type of dental lesion) can be used for comparison using the same type of numerical metric derived from a clinical measurement (e.g., from a clinical signal or a physical parameter from a clinical measurement), and the probability or degree of match can be determined by the system to output a likelihood of a match with that particular physical characteristic.

[0026] Machine learning algorithms can also be used for comparison to assist in improving efficiency and probabilistic matching with known data sets. For example, basic machine learning methods such as Bayesian networks for kernel density estimation and / or calibration curve adjustment can be used. After the machine learning system is fully developed and / or trained, deep learning methods can also be used to generate more advanced comparison methods.

[0027] In some embodiments, the system can detect digital metrics in clinical measurements that can instruct or suggest that the user change certain physical parameters of the clinical measurement, such as changing the parameters of a tapping measurement device, to generate better or more accurate data. For example, some digital metrics can indicate or suggest tapping with a different force, frequency, or location on the object to elucidate additional information or improve the quality of the measurement.

[0028] Generally, when an implant replaces a natural tooth due to injury or disease, the ligament is usually lost. However, the systems and methods of the present invention can also be used to evaluate the structural properties of implant structures using abutments. Some materials used for abutments, such as composites, gold, and zirconia, can produce certain sub-signals similar to the PDL response.

[0029] In addition, the present systems and methods as described above and below can be used to measure the dynamic response when force is applied to abutment materials, and can also be used to predict suitability or compatibility before implantation, or to select appropriate materials to protect natural teeth near the implant, and to better select materials to minimize inconsistencies between the way implants and natural teeth respond to impact.

[0030] In some aspects, the invention relates to a system for compiling test results from multiple objects, which may or may not include test results for one object tested over a period of time. In some embodiments, each test result can be generated using an instrument having a housing with an open end or a closed end with an energy application tool capable of applying energy to the object to generate a response, such as a percussive response, which can reveal structural properties of the object without substantially affecting existing structural properties of the object.

[0031] For example, the device (i.e., a knock measurement device) may include a housing having an open end and a longitudinal axis, including an energy application tool mounted in the housing for moving from a rest configuration to an activated configuration using a drive mechanism supported in the housing. The housing may include an object contact portion at its open end, or a sleeve may protrude a distance from the open end of the housing, and may include an object contact portion at its open end suitable for placing the device on at least a portion of the object and suitable for activating the drive mechanism, so that the energy application tool strikes the object when the object contact portion of the housing or sleeve is placed on at least a portion of the object, and is used to measure the response after the impact and generate a response to time curve or graph. The response can be captured by a computer coupled to the device as, for example, a response to time curve.

[0032] The system and method may include an apparatus having an energy application tool, such as a striking tool, capable of applying energy to an object to produce a response, such as an impact response, which can reveal structural properties of the object without substantially affecting existing structural properties of the object. The energy application tool can be programmed to strike the object a certain number of times per minute at a substantially constant rate for a certain time interval during a test. The system can measure the impact response, such as energy reflected from the object due to energy application, such as by striking or applying energy, or deceleration information of the energy application tool, i.e., energy return, over a time interval. The response can be fed to a computer and the information recorded or compiled for system analysis, which can include creating an impact response profile, such as an ERG, FRG, DRG, etc., or a frequency-based profile, such as a profile based on energy / force reflected from the object over a time interval, and / or evaluating, for example, an impact response profile (e.g., a time response profile) to determine structural properties of the object, such as vibration damping capacity; acoustic damping capacity; defects, including, for example, inherent defects in the bone or material constituting the object; cracks, microcracks, fractures, microcracks; glue seals; glue failures; adhesive failures; microleakage; lesions; decay; structural properties of a substrate or environment, such as that in which the object may be anchored or exist, overall structural integrity, or overall structural stability.

[0033] For example, the system of the present invention may include a device for performing tapping on an object. The tapping measurement device used in the present invention can have different configurations, and the test results generated from certain configurations can generate better models than other configurations. Generally speaking, the device includes a tapping instrument that can be repeatedly placed directly on the object subjected to such measurement to perform repeatable measurements.

[0034] In some exemplary embodiments of the present invention, the device used may include a shell with a hollow interior and an open end, through which energy can be applied by an energy application tool, including any tool capable of applying any type of energy (including mechanical energy, acoustic energy or electromagnetic energy) to an object. According to some embodiments, a tool capable of applying mechanical energy to an object, such as a knock rod or an impact rod, may be positioned or mounted inside the shell to reach the object being measured. According to other embodiments, an electromagnetic energy source of any frequency, such as light energy, may be located inside the shell, for example. According to yet another example, an acoustic energy source such as an ultrasonic transducer or any acoustic energy source may be positioned inside the shell.

[0035] The device of the present invention can be, for example, a percussion instrument, which can include a handpiece having a housing with a longitudinal axis, an open end and an energy application tool, such as a percussion rod or impact rod installed in the housing, for axial movement along the longitudinal axis of the housing, or for oscillating movement around the longitudinal axis of the housing. In some embodiments, the housing can include an object contact portion that can be reproducibly placed in contact with an object subjected to such measurement. In other embodiments, the housing can include at least a portion of a sleeve portion extending a certain distance from the housing, which can be reproducibly placed in contact with an object subjected to such measurement. Energy application tools such as percussion rods can have a certain length and be located in the housing, and can be programmed to impact an object a certain number of times per minute at substantially the same speed, and record or compile the deceleration information of the tool or the response of the object to the impact, for analysis by the system. In some embodiments, the device and the hardware can communicate via a wired connection. In other embodiments, the device and the hardware can communicate via a wireless connection.

[0036] Systems and methods for measuring and capturing measurements of an object may include a device having an energy application tool capable of applying energy to the object to generate measurements. For example, a percussion measurement device may be useful in the present invention and may have different configurations and forms, such as a desktop or portable device, such as a handheld device, and test results generated from certain configurations may generate better models than other configurations.

[0037] In some embodiments, energy application tools such as tapping rods have a retracted or resting form or configuration and an extended or activated form or configuration length, and if the energy application tool is a tapping rod, the retracted form is retracted from the open end of the housing or substantially extends with the open end. The movement of energy application tools such as tapping rods can be achieved by a drive mechanism installed inside the housing, and the drive mechanism is used to drive the tapping rod axially in the housing between the aforementioned retracted position and the extended position during operation. In the extended position, the free end of the tapping rod can extend or protrude from the open end of the housing. In other embodiments, energy application tools such as tapping rods can be in a form substantially parallel to the longitudinal axis of the housing, wherein the tip portion of the tool is substantially perpendicular to the longitudinal axis of the housing in the resting form or configuration, and moved to an activated form or configuration, wherein the energy application tool is at an acute angle to the longitudinal axis of the housing, and the tip of the tool is kept substantially perpendicular to the longitudinal axis of the housing by swinging back and forth around a pivot point on the longitudinal axis. In other words, the energy application tool can oscillate from a position substantially parallel to the longitudinal axis of the housing to a position at an acute angle to the longitudinal axis of the housing at the pivot point. The energy application tool may be held horizontally or in other positions during the measurement process and may have a tip portion that is substantially perpendicular to the main portion of the tool and maintains a constant length when resting or striking. Movement of an energy application tool such as a tapping rod may be achieved by a drive mechanism mounted inside the housing for driving the tapping rod from a position substantially parallel to the longitudinal axis of the housing to a position at a pivot point that is at an acute angle to the axis and back again, while the tip swings up and down in sequence. Movement of an energy application tool such as a tapping rod may be achieved by a drive mechanism mounted inside the housing for driving the energy application tool.

[0038] The drive mechanism can be, for example, an electromagnetic mechanism, and can include an electromagnetic coil. In some embodiments, the drive mechanism can include a permanent magnet fixed to the rear end of an energy application tool such as a knock rod, and the magnetic coil can be axially located behind the permanent magnet. If the device is a handpiece, the magnetic coil forms a structural unit together with the rear portion of the handpiece housing and any power supply lines, which can be integrally operated and can be connected to the rest of the device, for example, by a suitable releasable connection, such as a screw connection or a plug-in connection. This releasable connection may be helpful for cleaning, repair, etc. In other embodiments, the drive mechanism can be an electromagnetic mechanism, and can include an electromagnetic coil and a permanent magnet fixed to the rear end of an energy application tool such as a knock rod through an interface such as a coil mounting. Coils such as electromagnetic coils can, for example, be located axially behind the permanent magnet. The electromagnetic coil can also act directly on metal or conductive parts, such as ferromagnetic parts. Other forms of linear motors can also be used.

[0039] The energy application tool, such as a tap rod, may be located at the front of the housing, and the mounting mechanism for the tap rod may include frictionless bearings. These bearings may include one or more axial openings so that adjacent chambers formed by the housing and the tap rod communicate with each other to exchange air, depending on the amount of information expected in the test.

[0040] For a given energy application tool, for example, for a physical tool such as a tap rod, changes in the impact force can be achieved by, for example, changing the voltage, current, or both, changing the coil drive time (changing the length of time the coil is energized or activated), changing the speed at which the tap rod travels toward the object during impact, changing the coil delay time (changing the time between drive activities), changing the number of times the coil is energized (i.e., changing the number of drive pulses applied), the polarity of the coil, and / or combinations / multiples of the above.

[0041] According to some embodiments, the drive mechanism may include a measuring device, such as a piezoelectric force sensor, located within the handpiece housing for coupling to an energy application tool such as a tap rod. The measuring device may be adapted to measure the deceleration of the tap rod upon impact with an object during operation, or any vibrations induced by the tap rod on a specimen. The piezoelectric force sensor may detect changes in the properties of an object and may objectively quantify its internal characteristics. The data transmitted by the piezoelectric force sensor may be processed by a system program, which will be discussed further below.

[0042] According to other embodiments, the drive mechanism may include a linear variable differential transformer that is suitable for sensing and / or measuring the displacement of an energy application tool such as a knock rod before, during and after the application of energy. The linear variable differential transformer can be a non-contact linear displacement sensor. The sensor can utilize inductive technology and is therefore able to sense any metal target. In addition, the non-contact displacement measurement can allow a computer to determine velocity and acceleration before impact, thereby eliminating the effects of gravity from the results. The communication between the drive mechanism and the energy application can be wired or wireless.

[0043] The object contact portion located at the open end of the shell may or may not include a sleeve. In some embodiments, the open end of the shell can directly contact the object during the measurement process, thereby stabilizing the device on the object. In other embodiments, the sleeve can be attached to and / or surround at least a length of the free end of the shell, and if the tapping rod moves axially, the sleeve can extend from the shell at a distance that is substantially coextensive with the end of the tapping rod in its extended form. Therefore, the length of the sleeve can depend on the required protruding length of the extended tapping rod. The free end of the sleeve can abut against the object being measured. During the measurement process, the sleeve can directly contact the object, thereby stabilizing the device on the object. In other embodiments, additional features can be included to further stabilize the device, and some repeatability of the placement of the device on the object can also be built in, as described below.

[0044] In other exemplary embodiments of the present invention, the device can be as described in the above exemplary embodiments, except that the sleeve can include a fin protruding from at least a portion of its end, so that when the open end of the sleeve contacts at least a portion of the surface of the object to be measured, the fin can be placed on a portion of the top of the object. The fin and the sleeve together can assist in the repeatable positioning of the handpiece relative to the object; therefore, the results are more repeatable than without the fin. In rare cases, the fin may not protrude at all, so as to test at a lower position of the object. The fin can be substantially parallel to the longitudinal axis of the sleeve. In one aspect, the surface of the fin in contact with the object can be contoured, concave or convex to better position on the top of the object, such as a tooth. In another aspect, the surface of the fin in contact with the object can be flat to adapt to the topography of the object, such as a flat surface. In another aspect, the surface of the fin in contact with the object includes a groove to accommodate an object with an uneven surface. In addition, the fin can be adapted to be placed in substantially the same position on the top of the object each time it is repeated. In some embodiments, the fin can be substantially parallel to the longitudinal axis of the sleeve.

[0045] In rare cases, where the wings may interfere with a stable position on, for example, a dental implant transfer abutment, a sleeve portion without wings may be used for a more stable placement below the abutment.

[0046] In another exemplary embodiment of the present invention, the sleeve may include not only a wing, but also a feature, such as a ridge, protrusion or other feature substantially orthogonal to the surface of the wing located on the side adapted to face the object surface. For example, for teeth, the ridge or protrusion may be nested between adjacent teeth or other orthogonal surfaces, and thereby may assist in preventing any substantial lateral or longitudinal movement of the wing on the object surface and / or further assist in repeatability. The wing may have sufficient length or width, depending on the length or width of the top portion of the object, so that the ridge or protrusion may be properly arranged during operation. Likewise, the wing and feature also contribute to obtaining repeatable results compared to having no wing.

[0047] In the above exemplary embodiments, the device can be any form factor, as described above, including a handpiece having a longitudinal housing for housing the above device parts, or a tabletop device or any portable form. For example, during testing, any portable form or handpiece can be held at any angle to the horizontal plane.

[0048] The stabilization of the instrument by the fin or fins and / or assemblies can minimize any jerking movements of the operator that could confuse the test results, for example, any inherent defects in the bone structure or physical or industrial structure could be masked by the jerking movements of the tester. This type of defect detection is important because the location and extent of the defect can significantly affect the stability of the implant or the physical or industrial structure. Often when a lesion such as an apical or apical defect is detected in an implant, for example, the stability of the implant may be affected in the presence of both apical and apical defects. In the past, there was no other way to collect this type of information except for expensive strong radiation treatments. Using the present device, this type of information can be collected and this collection can be done in an unobtrusive, non-invasive manner without the need for radiation.

[0049] In another exemplary embodiment of the present invention, for example, an inclinometer may be present as part of the electronic control system of any of the above exemplary embodiments, and the inclinometer may trigger an audible warning when the device is outside of the operating angle range; for example, for a strike bar, when it is approximately plus or minus 45 degrees from horizontal, and for example, approximately plus or minus 30 degrees from horizontal, it may trigger a warning to return the device to a more horizontal orientation.

[0050] In another exemplary embodiment of the present invention, any or all of the above exemplary embodiments may also include a force sensor, which is not used to sense or measure the force applied by the energy application tool on the object during the test, nor is it used to sense or measure the response after the energy application tool hits, but is used to sense and / or monitor that the sleeve portion applies an appropriate contact force on the object being measured. As described above, for example, during the measurement process, the device may contact the object with the end of the shell or the sleeve portion. The contact force may vary from operator to operator. It is hoped that no matter who the operator is, the force is continuously applied within a certain range and the range is not too large. The force sensor can be included in the device for sensing the force and can be accompanied by a visual signal, voice or digital readout. The sensor can also be used to ensure that the correct alignment with the object is obtained during the measurement process. A sensor such as a force sensor can be physically close to and / or in contact with and / or physically coupled to at least a portion of the device other than the energy application tool; for example, if the open end of the sleeve portion includes an object contact portion, the sensor can be physically close to and / or in contact with and / or physically coupled to the shell and / or sleeve portion.

[0051] Typically, the sensor can surround the energy application tool without being in physical contact with the tool. For example, the sensor can be positioned so that the energy application tool, or even the physical tool, can pass through it to impact the object being measured. The sensor can include a strain gauge, a piezoelectric element, a sensing pad, or any other sensor that can be sandwiched in between. A sensor such as a force sensor can be arranged anywhere inside the housing, physically close to and / or in contact with and / or physically coupled to at least a portion of the device other than the energy application tool; for example, if the open end of the sleeve portion includes an object contact portion, it can be physically close to and / or in contact with and / or physically coupled to the housing and / or sleeve portion as described above. In some embodiments of the present invention, the sensor may include at least one strain gauge for sensing. The strain gauge can be attached or mounted on a cantilever between the device housing and the sleeve portion so that when the object contact portion of the sleeve portion is pressed against the object, it also deforms the cantilever measured by the strain gauge, thereby providing a force measurement value. In some embodiments, multiple strain gauges mounted to a single cantilever or multiple cantilever separated cantilevers can be used. The cantilever can also, for example, exist on a component separate from the rest of the housing or sleeve portion, such as on a mounting device. In other embodiments of the present invention, the sensor may include a sensing pad that may be positioned between a rigid surface and a sliding portion so that the force is measured when the sliding portion is pressed or squeezed toward the rigid surface. According to some embodiments, the rigid surface may be, for example, a coil interface in a drive mechanism that holds an electromagnetic coil in a housing of an apparatus of any of the above or below exemplary embodiments. The sliding portion may be a force transmission sleeve-like component or member that is disposed within the housing and coupled to an object contact portion of the sleeve portion and adapted to slide inside the housing when the object contact portion of the sleeve portion applies a force to the object. In some embodiments, it may be arranged within the sleeve portion. The sliding distance may be very small, for example, about (millimeter or mm) 0.3 mm to about 1 mm, more, for example, about 0.5 mm. The sensing pad may include a layer structure that may generally be referred to as a "shunt mode FSR" (force sensing resistor) that may change resistance according to the force applied to the pad to provide force measurement. According to other embodiments, the force transmission sleeve-like component or member may be biased forward by a spring so that when a force is applied to the object by the object contact portion of the sleeve portion, the force transmission sleeve-like component or member may transmit the force to the spring. According to one aspect, force sensing can be accomplished by a linear position sensor, knowing that, for example, if the force transmitting sleeve-like portion is at position X, a force of Y must be applied (overcoming the reaction force of the spring) to move it to that position. According to another aspect, force sensing can be performed by an optical sensor for optically sensing the position of the moving part as it is pushed against the spring.In other embodiments of the invention, the relative position of the object contacting portion of the sleeve portion on the object can be determined by having one or more strain gauges, which can be attached to a moving part, such as a force sensor sleeve-like part, at one end and connected to a static element, such as a housing, at the other end. In other embodiments of the invention, the device may include a piezoelectric element for directly measuring force. In another embodiment of the invention, a Hall effect sensor can be used to detect changes in the magnetic field when a magnet (attached to the moving element) moves relative to the position of the sensor. In other embodiments of the invention, a capacitive linear encoder system similar to that in a digital caliper can be used to measure force.

[0052] In addition to monitoring and sensing the contact force applied by the operator to the object when the object contacting portion or sleeve portion of the housing contacts the object, the sensor may also be configured to activate the device when the correct amount of force is applied by the sleeve portion to the object.

[0053] Although the sensor is not physically or mechanically coupled to the energy application tool, it can communicate electronically with the energy application tool and can act as an on / off switch for the device or instrument, as described above. For example, when an appropriate force is applied to the object through the object contact portion of the housing or sleeve, it can trigger the activation mechanism of the device or instrument to activate the movement of the energy application tool to start measurement. Therefore, as described above, no external switch or button is required to activate the opening and closing of the system. The indication of the appropriate force can be indicated by a visual or audible signal.

[0054] The sleeve portion may be mounted to a force transfer sleeve-like assembly or force transfer member which forms a permanent part of the front of the housing or protrudes from it and shields an energy application tool such as a tap rod from damage when the sleeve portion is not present. For example, the sleeve portion may form part of a disposable assembly, as described below. The force transfer sleeve-like component or member is seated around an energy application tool such as a tap rod; and may surround the energy application tool and be retained by the housing at the front and mounted at the front of an electromagnetic coil at the rear. The force transfer sleeve-like component or member may be adapted to slide a small amount, and in doing so may act on a force sensor such as a force sensitive resistor located between the rear surface of the force transfer sleeve-like component or member and the coil mounting. When the object contact portion of the sleeve portion is pushed against the object being measured, an energy application tool such as a tap rod may be triggered; for example, a tooth and force may be detected. When the correct force within a certain range is detected, the instrument is turned on to start measuring.

[0055] As described above, in all embodiments of the sensor, the sensor may be arranged to form a channel through which an energy application tool, such as a tap rod, may pass to impact the object being measured, ie around the tap rod.

[0056] If the device is oriented so that when the pushing force is sensed on the object contact portion of the sleeve portion, the operating axis is greater than about 45 degrees from the horizontal, more for example greater than about 30 degrees, it may cause a speaker located on the device (e.g., a printed circuit board (PCB) within the device) to emit a warning sound. In this case, the tapping action will not begin until the device returns to an acceptable angle. In some cases, if the tapping action has already begun when the above-mentioned deviation range is detected, the device may not actually stop operating but simply issue an alarm so that correction can be made.

[0057] In another exemplary embodiment of the present invention, in any of the above exemplary embodiments, the system and method may also include a device capable of operating by holding the device at an angle different from the horizontal direction and adjusting the energy application process to simulate a substantially horizontal position during measurement, and a system that can apply optimal energy to the object in all situations can be provided. In some embodiments, the device can apply substantially the same impact force to the object at various angles from the perpendicular direction of the object surface, as if the device is operating so that the propagation direction is perpendicular to the object surface. Therefore, regardless of whether the device is operating at an angle of approximately plus or minus 45 degrees, more for example approximately plus or minus 30 degrees, in a direction perpendicular to the object surface, the device can still generate approximately the same amount of equivalent impact force, for example, approximately 20-30 Newtons, to obtain optimal results. The system may include a visual indicator, such as an LED when the handpiece is held at an angle that does not support reliable measurement. In this case, the LED may turn red or any other preset color to indicate this situation to remind the user to readjust the angle at which the handpiece is held.

[0058] The systems and methods of the present invention can, for example, increase the flexibility of operation, for example, to accommodate anatomically and non-anatomically difficult to reach objects, to detect any abnormalities that may be present in the object to produce more reproducible measurements, and also to better detect any abnormalities that may be present in the object. The device can include a housing having a hollow interior and an open end, through which an energy application tool passes to reach the object being measured, the energy application tool including any tool capable of applying any type of energy to the object, such as a tool capable of applying mechanical energy to the object, such as a tapping rod located inside the housing, such as electromagnetic energy of any frequency, such as light, sound waves (such as acoustic energy).

[0059] For example, the system may include a device for performing a tapping action on an object. The device may be placed, the device having a shell with a hollow interior and an open end, through which energy may be applied by an energy application tool, including any tool capable of applying any type of energy (including mechanical energy, acoustic energy, or electromagnetic energy) to the object. In some embodiments, a tool capable of applying mechanical energy to an object, such as a tapping rod, may be positioned through the interior of the shell to reach the object being measured. In other embodiments, an electromagnetic energy source of any frequency, such as light energy, may be located inside the shell, for example. In yet another example, an acoustic energy source, such as an ultrasonic transducer, or any acoustic energy source may be located inside the shell.

[0060] During the measurement process, the energy application tool can be held horizontally, or in other positions, and can have a tip portion that is substantially perpendicular to the main portion of the tool and maintains a constant length whether at rest or during impact. In the latter embodiment, if the tool is a mechanical tool, such as a striker bar, it may or may not include a removable tool tip that is substantially perpendicular to the longitudinal axis of the tool and housing.

[0061] An energy application tool such as a tap rod can be programmed to strike an object a certain number of times per minute at substantially the same rate, and the deceleration information can be recorded or compiled for system analysis as described above. If the sleeve portion is a material with certain damping properties, then in addition to assisting in positioning the device, the sleeve portion can also assist in attenuating any vibration caused by the impact so that it does not interfere with sensitive measurements.

[0062] For electromagnetic energy, the energy application may be in the form of pulses or bursts of energy that may be programmed to strike the object a certain number of times per minute with approximately the same amount of energy each time, and the effects on the object may be recorded or compiled for system analysis. In some cases, repeated impacts may provide an average measurement that may better represent the actual underlying property. If the sleeve portion is of a material with certain damping properties, then in addition to assisting in positioning the device, the sleeve portion may also assist in attenuating any vibrations caused by the impact so that they do not interfere with sensitive measurements.

[0063] When, for example, a mechanical energy applying tool is activated, for example, by pressing a finger switch on the device, or when a certain amount of force is applied by the object contact portion of the sleeve or housing, as described above, the electromagnetic coil within the device will push the energy applying tool, such as a tap rod, to extend toward the object being measured at a certain speed, and, for example, use the impact force to strike or impact the object or sample multiple times per measurement cycle. In some cases, such as in a controlled laboratory environment where a large number of signals need to be generated, when the handpiece may be located in a mounting, the handpiece can be configured to be activated without waiting for the appropriate amount of contact force to be applied by the object contact portion of the sleeve or housing. The impact force on the object may generate a stress wave that travels through the energy applying tool, such as the tap rod, and the deceleration of the tool, such as the tap rod, when it strikes the object can be measured using a measuring or sensing device or a mechanical device located within the device, and the deceleration of the tool can be sent to the rest of the system for analysis. The system can measure an impact response, such as energy reflected from an object due to the application of energy, over a certain time interval, for example by striking or applying energy, which can include, for example, creating an impact response distribution map, such as a signal or frequency-response distribution map based on the response from the object during the time interval, and / or evaluating the impact response distribution map, such as the signal, to determine the damping capacity or other characteristics of the object. The measuring device or sensing mechanism can detect the characteristics of the effect from the collision of the energy application tool with the object. Typically, the measuring device or sensing mechanism can be physically coupled to, functionally coupled to, or otherwise in contact with the energy application tool so that it can detect the characteristics of the collision. The coupling can be wired or wireless.

[0064] As described above, after colliding with an object, an energy application tool such as a tap rod decelerates. The deceleration of an energy application tool such as a tap rod can be measured by a measuring device or sensing mechanism (such as an accelerometer inside the device). For example, an accelerometer within a device coupled to the energy application tool can be adapted to measure the deceleration of the energy application tool when colliding with an object during operation, the impact response from the object, any vibration caused by the impact, or a signal corresponding to the stress wave generated. The measuring device or sensing mechanism can detect changes in the properties of the object and can objectively quantify its internal characteristics. As described above, the data transmitted by the measuring device or sensing mechanism can be processed by a system program.

[0065] The above-described measuring mechanism may also be adapted with a similar sensor arrangement to tools other than the mechanical energy applying tool as described above, for example, when such an energy applying tool performs a striking action.

[0066] In some embodiments, the inclinometer may include an accelerometer that is, for example, a 3-axis device that measures gravity on all three axes (X, Y, and Z axes). In some embodiments of the present invention, a device such as a handpiece may include software for measuring the value of gravity (G-force) on the Y axis (i.e., vertical). For example, if the G-force on the Y axis is greater than, for example, a threshold of plus or minus 15 degrees, the handpiece may emit an audible sound (e.g., a beep) or a light signal (e.g., a flash), or a light of a certain color. If the G-force on the Y axis is greater than a threshold of 30 degrees, the handpiece may emit a faster beep, or if it is a light signal such as a flash, it may be a light that flashes faster. The accelerometer may sample, for example, every 100 ms. Five consecutive valid readings (500 milliseconds) may be required to trigger the threshold, thereby emitting a beep or flash, etc. The thresholds for the 15 and 30 degree thresholds may be determined empirically.

[0067] For example, for a device not having features of the present invention, during operation, if the equivalent impact force at 15 degrees from horizontal is about 26 Newtons, the equivalent impact force at horizontal is about 32 Newtons, and at negative 15 degrees from horizontal, the impact force may be about 35 Newtons. With the present invention, all impact forces at all of the above angles may be about 25 Newtons, or programmed to any optimal impact force applied.

[0068] As described above, the system can be turned on and off with or without an external switch or remote control. In some embodiments, the energy application process of the handpiece can be triggered by a mechanical mechanism, such as a switch mechanism. In one aspect, a finger switch can be located in a convenient location on the handpiece for easy activation by the operator. In another aspect, the switch mechanism can be triggered by applying pressure to the object via a sleeve. In other embodiments, the energy application process of the handpiece can be triggered via voice control or foot control or buttons of a computer software user interface.

[0069] Typically, any external switching device such as a toggle switch, rocker switch or push button switch may restrict the way the operator can hold the instrument and therefore, for example, may restrict the positioning of the instrument (if handheld) on the object during measurement so that the operator can easily access the switching device to switch it on and / or off.

[0070] In some embodiments, in order to obtain greater flexibility in positioning the instrument, voice control or remote control can usually be used, although such voice control or remote control may increase the complexity of the system. In the present invention, the same flexibility advantage can be obtained without such remote control or increased complexity.

[0071] In other embodiments, as described above, in order to obtain greater flexibility in positioning the instrument, the activation of the device can be controlled by an appropriate contact force between the object and the sleeve portion located at the open end of the housing. As described below, the appropriate contact force can also add other desired features to the system. The sleeve portion can be open at its free end, wherein the object rests, squeezes or contacts the portion for resting, pressing or contacting at least a portion of the object during measurement. The contact assistance of the sleeve portion stabilizes the device on the object. During measurement, unlike the impact force of the energy application tool, the force applied by the sleeve portion on the object is controlled by the operator, which can be controlled by various factors of the above-mentioned system, and the appropriate contact force applied to the object may be important and may need to be monitored, because, for example, insufficient or excessive force applied by the operator will complicate the measurement and may even produce less accurate results. There can be a sensor arranged inside the housing that is not physically or mechanically coupled to the energy application tool to ensure that even different operators can apply appropriate contact forces based on the contact portion of the sleeve portion by the operator to achieve better reproducibility.

[0072] In some embodiments, once the appropriate contact force is applied to the object by the object contact portion of the sleeve, as indicated by a visual or audible signal, the instrument can be turned on instantaneously. In other embodiments, once the appropriate contact force is applied to the object by the object contact portion of the sleeve, as indicated by a visual or audible signal, there may be a delay before turning on the instrument. In another embodiment, once a certain thrust between the object contact portion of the sleeve and the object is detected and maintained for a period of time, such as about 1 second, and also for example about 0.5 seconds, the instrument can be turned on to start measuring. In this embodiment, a green light illuminates the tip, and the impact begins after maintaining a force within the correct range for about 1 second, and also for example about 0.5 seconds.

[0073] The operator, for example, acts as a switch for the system by applying an appropriate force to the object through the sleeve portion. When the system is not turned on, it may be necessary to know whether it is faulty, not applying enough force, or applying too much force. In some embodiments, the force measurement can be connected to a visual output, such as a light. The light can be installed in any convenient position of the device or instrument, for example, one or more LEDs can be installed on the front of the device or instrument. In one aspect, a multi-light system can be included. For example, two LEDs can be used. When the force is within the correct range, a green light can be lit. If too much force is detected, the LED can turn red, and the instrument will not work unless the thrust is reduced. In some embodiments, if the user pushes the object too much, the light can first turn amber and then red. If the thrust is enough to turn the light red, the knocking can not be started, or it will be interrupted if it has been started. In addition, there can be an amber LED state to warn when the user is close to too much thrust. At this stage, when the LED lights amber, the instrument can still be running. On the other hand, no light can indicate that the force is too small, green light can indicate the right amount of force, and red light can indicate that the force is too large. On the other hand, a single light system can be included. For example, no light can signal that the force is too little and a red light can signal that the force is too much. On the other hand, a flashing red light can indicate that the force is too much and no light can indicate that the force is too little.

[0074] In other embodiments, the force measurement can be connected to an audible output. In one aspect, the audible output can include a beep to indicate too little force and multiple beeps to indicate too much force. In another aspect, the audible output can include a beep to indicate too little force and a beep with a flashing red light to indicate too much force. In yet another aspect, the force measurement can be connected to a voice alarm system to alert of too much force or too little force. In yet another aspect, the force measurement can be connected to a voice alarm system to alert of too little force and a voice alarm and a flashing red light to alert of too much force.

[0075] During the measurement process, as described above, the system can measure the impact response, such as energy reflected from the object by energy application such as striking or applying energy, or deceleration information of the energy application tool, i.e., energy return, over a period of time. The response can be fed to a computer, and the information is recorded or compiled for analysis by the system, which can include creating an impact response distribution map, and / or evaluating the impact response distribution map to assist in determining structural characteristics of the object, such as vibration damping capacity; acoustic damping capacity; defects, including, for example, inherent defects in the bone or materials constituting the object; cracks, microcracks, fractures, microcracks; glue seals; glue failures; adhesive failures; microleakage; lesions; decay; overall structural integrity or overall structural stability.

[0076] The loss coefficient is generally an indicator of the overall damping capacity of the object or structure being tested. During an impact, it is based on the maximum energy return or square of the impact force, which is measured with a measuring or sensing mechanism coupled to an energy application tool such as a strike bar, as discussed in the context. The normal fit error (NFE) or damage or instability is the total error (difference) between the ideal curve (produced by a defect-free object) and the actual test data. These results can be calculated based on the ERG, FRG or other physical return values ​​or metrics. Before determining the NFE, all response curves are normalized to a maximum value of 1, so it has no direct relationship to the loss coefficient.

[0077] As described above, the system and method of the present invention are non-destructive and non-invasive, and can include a device capable of operating in a substantially horizontal position by keeping the device at different angles to the horizontal during measurement and adjusting the energy application process. The system may include or not include disposable components and / or features for assisting repositionability. The present system and method for measuring structural characteristics can minimize the impact, even slight impact, on the object being measured without compromising the sensitivity of the measurement or operation of the system. When the energy application tool is a knock rod, the amount of impact energy can also vary depending on, for example, the length of the rod, the diameter of the rod, the weight of the rod, or the speed of the rod before the collision. In some embodiments, the system includes an energy application tool that is light in weight and / or can move at a slower speed, thereby minimizing the impact force on the object during measurement while demonstrating, maintaining or providing equivalent or better measurement sensitivity. On the one hand, energy application tools such as knock rods can be made of lighter materials to minimize the weight of the handpiece, thereby minimizing the impact on the object being measured. In other embodiments, energy application tools such as knock rods can be made shorter and / or have a smaller diameter, so that the size of the handpiece can also be minimized, and thus the impact on the object being measured can be minimized. In another embodiment, the system may include a drive mechanism that can reduce the acceleration of the energy application tool and thus minimize the impact on the object being measured. For example, the drive mechanism may include a separate drive coil to reduce the acceleration of the energy application tool, regardless of whether it is lighter in weight and / or smaller in length or diameter, and minimize the impact force on the object during operation while maintaining the sensitivity of the measurement. These embodiments can be combined with one or more embodiments described in the context, including a lighter handpiece housing. The speed at which the measurement is performed may also be preferred without increasing the initial impact velocity to minimize the impact on the object during the measurement process. The system may or may not have disposable components and / or features for assisting the above or below repositionability.

[0078] The system may include a drive mechanism that can change the travel distance of the energy application tool while maintaining the initial velocity of the energy application tool's impact on the object. For example, when the energy application tool includes a striking tool, the distance can vary within a range of about 2 mm to about 4 mm. The travel distance of the energy application tool is reduced by, for example, about 4 mm to about 2 mm while maintaining the same initial velocity when impacting or contacting, which can achieve faster measurements without compromising the operation of the system. The system may or may not include various exemplary embodiments described in the context. For example, the system may or may not have disposable components and / or features for assisting repositionability and / or mitigating impact using features mentioned in the context.

[0079] For any of the above exemplary embodiments, the system and method for measuring structural characteristics using an energy application tool may also include a disposable feature to assist in eliminating or minimizing contamination of the measured object through transfer from the system or cross-contamination from a previously measured object without interfering with the measurement or the system's capabilities. The device includes a housing having a hollow interior with an open end and an energy application tool, such as a tapping rod or impact rod mounted in the housing for movement within the housing. The system provides a non-destructive measurement method in which there is some contact with the object undergoing such measurement, but without the need to wipe or autoclave the energy application tool, and also without the need to handle the energy application tool and / or the housing and anything that may be contained within the housing of the device.

[0080] Disposable features may include a cover for covering or surrounding a portion of the system that may be in proximity or contact with the object being measured without interfering to any substantial extent with sensitivity, reproducibility (if desired), or general operation of the instrument.

[0081] The disposable features may include any of the features described below or disclosed in US Pat. No. 9,869,606 or WO 2011 / 160102 A9, entitled “System and Method for Determining Structural Features of an Object,” the entire contents of which are incorporated herein by reference.

[0082] The disposable feature may include a sleeve portion extending from the open end of the housing and / or surrounding the open end of the housing. In one example, for a mechanical energy application tool, the sleeve portion includes a hollow interior and an open free end, with an object resting or contacting portion, for resting, pressing or contacting an object at its open end during measurement. Features such as contact features, which may or may not be movable, have a certain length, are arranged toward the open end of the sleeve portion and may include a closed end for substantially closing the free end of the sleeve portion, such as being tightly embedded in the sleeve portion by friction. The contact feature may be, for example, a short tubular portion or a ring, and may include a closed end for substantially closing the free end of the sleeve portion. The contact feature may be located between the tip of the energy application tool and the surface of the object being measured. The above-mentioned contact feature may include a film, which may be attached or integrally formed to form a closed end as described in the context. The film may be thick or thin, as long as they are selected to have minimal impact on the operation of the energy application tool. For example, the closed end, whether closed by a membrane or other structure, can have a certain elasticity or deformability, and can adjust itself to adapt to various surface structures of the object being measured, so as to achieve close contact with the object during the impact process.

[0083] The closed end may include a thin polymer film, which may or may not be the same material as the rest of the contact feature, or may be a material having substantially the same properties as the rest of the contact feature. The polymer may include any polymer material that can be molded, cast or stretched into a film, thereby substantially not adversely affecting the measurement. In other embodiments, the closed end may include an insert molded metal foil film. The metal may be any metal material that can be stretched, cast or molded into a film, thereby substantially not adversely affecting the measurement. The film may also be formed to conform to the shape of the energy application tool, or vice versa, to achieve the best transfer of force / energy. In some embodiments, the film may be made of stainless steel foil or sheet, and may be, for example, stamped and / or molded. In other embodiments, the closed end may be integrated with the contact feature. For example, the contact feature may be formed by a tubular or annular structure of a closed end that may be formed into a desired thickness by stamping a metal (e.g., stainless steel, aluminum, copper or other suitable metal).

[0084] Additionally, the sleeve portion, contact features and tabs and / or the sleeve, tabs and components may be made from recyclable, compostable or biodegradable materials, which are particularly useful in embodiments where the materials are disposed of after a single use.

[0085] The device itself may be connected to an external power source or powered by a power source included within the housing, such as a battery, capacitor, transducer, solar cell, external power source, and / or any other suitable power source.

[0086] As previously mentioned, the system and method can be applied to test various mechanical objects such as those mentioned above. For mechanical objects, which may include but are not limited to polymer composite structures, including honeycomb or layered honeycomb or metal composite structures; aircraft structures, automobiles, ships, bridges, tunnels, trains, buildings, industrial structures, including but not limited to power generation facilities, arch structures or other similar physical structures, it is also possible to test stationary objects or moving objects in motion. Therefore, mechanical objects can also be tested when they are stationary or moving, which can provide a special understanding of the object under actual working conditions. For moving objects such as trains, the test can be performed at a large number of different locations. This operation can be performed at multiple points on the object using an energy application tool to obtain an average condition of the object as a whole, or multiple independent tools or devices can be used to perform this operation at the same point to obtain an average result at the same point. In order to perform measurements at the same point using multiple energy application tools, the device or tool can be positioned, for example, along the path of the moving object for a continuous distance, for example, an array of knocking rods that strike the object, and by controlling the intervals between the tools or devices, one can be able to match the speed of a moving object such as a train with the intervals of applying energy at the same point of the object to obtain the average value of the point. In this example, the measurement can be performed under actual operating conditions. In some embodiments, the array of devices can be a linear array, a vertical or horizontal array, or a curved array. On the other hand, the array can be arranged in a two-dimensional array, a plane or a curve.

[0087] In embodiments where the object is larger, measurements at different locations on the object, for example, impacts at multiple portions of the object, may enable a better assessment of the structural properties that are better representative of the object.

[0088] In general, structural properties as defined herein may include: vibration damping capacity; acoustic damping capacity; defects, including inherent defects in, for example, bone or the material forming the object; cracks, microcracks, fissures, microcracks; glue seals; glue failure; adhesive failure; microleakage; lesions; decay; overall structural integrity or overall structural stability. For anatomical objects such as tooth structures, natural teeth, natural teeth with cracks due to wear or trauma, natural teeth that have become at least partially abscessed, or natural teeth that have undergone bone strengthening surgery, artificial dental implant structures, dental structures, orthopedic structures or orthopedic implants, such properties may indicate the health of the object, or the health of the underlying substrate to which the object may be anchored or attached. The health of the object and / or the underlying substrate may also be related to the following: density or bone density or level of osseointegration; any inherent or other defects; or cracks, cracks, microcracks, microcracks; glue seals; glue failure; adhesive failure; microleakage; lesions; or decay. For general objects such as polymer composite structures, including honeycomb structures or layered honeycomb or metal composite structures; aircraft structures, automobiles, ships, bridges, buildings, industrial structures, including but not limited to power generation facilities, arch structures or other similar physical structures; such measurements can also be related to any structural integrity, or structural stability, such as defects or cracks, even very fine cracks or microcracks, etc.

[0089] Additionally, changes in the structure of any substrate to which a tooth or mechanical structure is attached or anchored that reduce its ability to dissipate mechanical energy associated with impact forces and thereby reduce overall structural stability of, for example, a tooth can be detected by evaluation of energy return data compared to an ideal undamaged specimen. Additionally, as described above, the present invention also aids in detecting defects, cracks, microcracks, fissures, microcracks, leaks, lesions, localization accuracy of glued seals; microleakage; decay; structural integrity at glue failure; adhesive failure; overall or structural stability in general.

[0090] In some exemplary embodiments, the present invention includes:

[0091] A method for providing a structural property analysis system trained by machine learning, comprising:

[0092] providing or generating a data set comprising a plurality of signals from a plurality of groups of different subjects, the signals being generated from a series of tap measurements on the different subjects and being grouped based on a common characteristic of one of the groups of different subjects;

[0093] Generate an optimized signal set for each of the signals, each of the optimized signal sets being generated by the following operations:

[0094] performing a guess decomposition of each of said signals to generate a signal set including at least one sub-signal for each signal;

[0095] performing an optimization operation to minimize the difference between the signal set and each of the signals to generate an optimized signal set;

[0096] Identifying and resolving potential errors or deficiencies in each of the optimized signal sets;

[0097] After eliminating the potential errors or defects, repeating the guess decomposition and the optimization operation to regenerate an optimized signal set;

[0098] selecting at least one desired signal set for each signal from the optimized signal set to add to the optimized signal set; and

[0099] Incorporating the set of optimized signal sets and associated methods for deriving the set of optimized signal sets into a machine learning trained analysis system (MLTA); and

[0100] The MLTA is connected to a measuring device, which is suitable for generating clinical signal data by tapping a target object and transmitting the clinical signal data to the MLDA, so that the MLMA can process the clinical signal data to generate a clinically optimized signal set, compare the characteristics of the clinically optimized signal set with the characteristics of the optimized signal set, and present the comparison results in a human-readable form.

[0101] In some exemplary embodiments, the present invention includes:

[0102] A structural property analysis system trained by machine learning, comprising:

[0103] Percussion measurement equipment, including:

[0104] a housing having an open front end and a longitudinal axis;

[0105] an energy application tool mounted inside the housing, the energy application tool having a rest configuration and an activated configuration;

[0106] a drive mechanism supported within the housing, the drive mechanism adapted to activate the energy application tool between the rest configuration and the activated configuration to apply a set amount of energy; and

[0107] a control mechanism connected to provide instructions to the drive mechanism;

[0108] wherein the drive mechanism varies an amount of energy applied to activate the energy application tool between the rest configuration and the activated configuration based on input from the control mechanism;

[0109] A program logic module is connected to the control mechanism, and the program logic module is provided by the following operations:

[0110] providing or generating a data set comprising a plurality of signals from a plurality of groups of different subjects, the signals being generated from a series of tap measurements on the different subjects and being grouped based on a common characteristic of one of the groups of different subjects;

[0111] Generate an optimized sub-signal set for each of the signals, each of the optimized sub-signal sets being generated by the following operations:

[0112] performing a guess decomposition of each of the signals to generate a set of sub-signals including at least one sub-signal for each signal;

[0113] performing an optimization operation to minimize the difference between the sub-signal set and each of the signals to generate an optimized sub-signal set;

[0114] Identifying and resolving potential errors or defects in each of the optimized sub-signal sets; after resolving the potential errors or defects, repeating the guess decomposition and the optimization operation to regenerate the optimized sub-signal set;

[0115] selecting at least one desired sub-signal set for each signal from the optimized sub-signal sets to add to the optimized sub-signal set; and

[0116] incorporating the set of optimized sub-signal sets and associated methods for deriving the set of optimized sub-signal sets into a machine learning trained analysis system (MLTA); and connecting the MLTA to the tapping measurement device, the tapping measurement device being adapted to generate clinical signal data by tapping a target object and transmitting the clinical signal data to the MLTA so that the MLTA can process the clinical signal data to produce a clinically optimized sub-signal set and compare characteristics of the clinically optimized sub-signal set with characteristics of the set of optimized sub-signal sets, and determine physical parameters associated with the comparison;

[0117] A control regulator is connected to the program logic module and the control mechanism, the control regulator being adapted to output a change to the instruction in response to the MLTA output suggesting a change in the physical parameter.

[0118] In some exemplary embodiments, the present invention includes:

[0119] A method for providing a structural property analysis system, comprising:

[0120] providing a program logic module (PLM) configured to receive an input of a signal to generate a signal from a tap measurement performed by a tap measurement device (PMD);

[0121] Connecting the PLM to the PMD;

[0122] performing a tap measurement on a tooth-shaped object using the PMD to generate the signal using the PLM;

[0123] performing a guessing of a prominent sub-signal of the signal by fitting the signal to a basis function;

[0124] subtracting the primary sub-signal from the signal to form a remainder;

[0125] performing a guessed sinusoidal decomposition on the remaining portion to generate a secondary sub-signal, the secondary sub-signal being added to the primary sub-signal to form an approximation of the signal;

[0126] performing an optimization operation to minimize the difference between the approximation of the signal and the signal to generate a set of optimized sub-signals;

[0127] Identifying and resolving potential errors or defects in each of the optimized sub-signal sets;

[0128] Repeating the guessing, guessed sinusoidal decomposition and optimization operation on the main sub-signal to regenerate the optimized sub-signal set after eliminating the potential errors or defects;

[0129] selecting at least one desired sub-signal set from the optimized signal set; and

[0130] The desired set of sub-signals is presented in a human-readable form.

[0131] The present invention and the above and other advantages may be best understood in conjunction with the following detailed description of aspects, embodiments and examples of the present invention and as shown in the accompanying drawings. The following description, while indicating various aspects, embodiments and examples of the present invention and numerous specific details thereof, is given by way of illustration and not limitation. Numerous substitutions, modifications, additions or rearrangements may be made within the scope of the present invention, and the present invention includes all such substitutions, modifications, additions or rearrangements. BRIEF DESCRIPTION OF THE DRAWINGS

[0132] Figure 1 shows the connection arrangement of the components of the system of the present invention;

[0133] Figure 2 , 2a, 2b, 2c, 2d, 2e and 2f show embodiments of the knock measurement device or its components of the present invention;

[0134] Figure 3 and 3a An example distribution diagram of a signal in the present invention is shown;

[0135] Figure 4 and 4a An example of a missing portion of a signal is shown;

[0136] Figure 4b An example of sub-signal decomposition of a signal is shown;

[0137] Figure 4c Examples of the main sub-signals in the signal are shown;

[0138] Figure 4d A finite element analysis model is shown;

[0139] Figure 4e The periodontal ligament attached to the tooth structure is shown; and

[0140] Figure 5 , 5a 5a and 5b show examples of heat maps of the present invention. DETAILED DESCRIPTION

[0141] The detailed description set forth below is intended as a description of some exemplary systems, devices and methods of the present invention provided according to aspects of the present invention, but is not intended to represent the only form in which the present invention can be prepared or utilized. Moreover, it should be understood that the same or equivalent functions and components can be implemented by different embodiments, which are also intended to be encompassed within the spirit and scope of the present invention.

[0142] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the invention belongs. Although any systems, methods, devices, and materials similar or equivalent to those described herein can be used in the practice or testing of the present invention, some exemplary systems, methods, devices, and materials are now described.

[0143] All publications mentioned herein are incorporated herein by reference for the purpose of describing and disclosing, for example, the designs and methods described in these publications, which can be used in conjunction with the invention described herein. The publications listed or discussed above are provided separately below and throughout the text due to their disclosure prior to the filing date of the present application. Nothing herein should be construed as an admission that the inventor is not entitled to precede such disclosure by virtue of prior invention.

[0144] The present invention relates to a system and method for measuring and evaluating structural properties of an object, whether anatomical or non-anatomical, in a non-invasive manner and / or using non-destructive measurement methods. The damping capacity of an object, including anatomical or non-anatomical objects, is an important parameter in a wide variety of applications. For example, in the field of dentistry, when a healthy tooth is subjected to an impact force, the mechanical energy associated with the impact is primarily dissipated by the periodontal ligament. Structural changes in the periodontal ligament that reduce its ability to dissipate the mechanical energy associated with the impact force and thereby reduce overall tooth stability can be detected by measuring the loss coefficient of the tooth.

[0145] Objects, whether anatomical or non-anatomical, such as dental systems, whether natural teeth or implants, may also develop defects over time. Some defects require restorative dental surgery. Such surgery can be invasive and expensive and require a long recovery time, especially if the defects are not easily discernible before they develop into more discernible severe defects. Technology that can quickly identify and pinpoint the type of problem and its location before the existing problem becomes severe and / or before destructive surgery is required is highly desirable to reduce the risk of ineffective or unnecessary surgery.

[0146] As used herein, the following professional definitions apply generally unless otherwise stated:

[0147] As used herein, "signal" refers to the energy, force, displacement or other physical change value returned from the object over time during measurement after the percussion measurement device is struck, specifically manifested as an electrical response generated by a sensing mechanism such as a piezoelectric sensor element, a strain gauge, a displacement sensor, etc., by recording the force and time data generated by this electrical response, or a graphical representation of force and time (force return graph or FRG), energy and time data (energy return graph or ERG), displacement and time data (displacement return graph or DRG) or other graphical representations of applicable data. Unless otherwise specified, a signal can generally be understood to be in a waveform distribution graph, which essentially represents the relationship data between the total force, energy, displacement or other physical change value and time recorded by the percussion measurement device and received by the system during the measurement, wherein the measurement has not been adjusted or manipulated by the system. "Signal" may also refer to the artificially generated simulated version described above. Different physical change values ​​can also be derived or calculated from the actual measured physical values.

[0148] As used herein, a "sub-signal" refers to a signal component having a waveform profile where the sum of all sub-signals produces the original signal or an approximation thereof.

[0149] As used herein, "sinusoidal" refers to a waveform that takes the approximate shape of a sine wave, including sine waves whose amplitude decays over time, such as exponentially decaying or otherwise damped sinusoids ("decayed sine waves").

[0150] As used herein, a "clinical signal" refers to a signal captured by an end user (ie, a "clinician") when using a tap measurement device in a clinical, industrial, or other non-training or non-testing environment.

[0151] As used herein, a “basis function” refers to a function or waveform that fits a signal or sub-signal to a Gaussian distribution curve, a sinusoidal curve (including an attenuated or damped sinusoidal curve as described above), an exponential function, a quasi-sinusoidal curve, and / or any other suitable fitting function or waveform.

[0152] Structural properties of an object can be identified based on measurements of the same or other objects previously made using the system and captured by the system. The system may include a device (i.e., a tapping measurement device) capable of applying energy to an object, which may be an anatomical or mechanical object; measuring a response over a time interval, such as energy reflected from the object due to energy application (e.g., tapping the object), or measuring a response, such as deceleration information of an energy application tool such as a tapping rod, such as energy return; recording or compiling these measurements for analysis by a computing system; creating a response distribution map, such as a signal, return curve, or return map (e.g., a time distribution map or a frequency distribution map) to evaluate the characteristics of the object being measured (i.e., generating a signal). The signal can typically be generated from measurements of force / energy / displacement, etc. returned to the tapping measurement device, such as by measuring the force, energy, displacement, or other physical return value on a sensing mechanism over a period of time; or deceleration information from the energy application tool after the energy application process. Context Figure 2 , 2a , 2b, 2c, 2d and 2e describe several embodiments of the tap measurement device.

[0153] Figure 1 An embodiment of a system architecture using an endpoint device (i.e., a percussion measurement device referenced throughout the text) is shown, for example, for dental measurements. The percussion measurement device may typically be located, for example, in a dentist's office or other location where measurements of a patient's objects (e.g., teeth, implants, etc.) may be performed. The percussion measurement device may typically include or be connected to a computing device, such as a PC workstation, a laptop, a tablet computer, or other general computing device that may be connected to a larger network (e.g., the Internet) or a dedicated network (e.g., a cloud service). At least one device may be connected to the percussion measurement device via a wired data transmission technology (e.g., USB or FireWire) or a wireless data transmission technology (e.g., Bluetooth). The system may also include a base station, as shown in the figure, for interfacing with the percussion measurement device and the computing device.

[0154] A knock measurement device suitable for use with a test object may include a housing having a longitudinal axis, the housing having an open end and an energy application tool, such as a knock rod or impact rod, mounted within the housing for axial movement along the longitudinal axis of the housing, such as Figure 2a As shown, this figure shows an embodiment of the above-mentioned knock measurement device. In some embodiments, the system may include a handpiece 104 in the form of a knocking instrument. The handpiece 104 may have a cylindrical shell 132, which has an open end 132a and a closed end 132b. As illustrated here, the open end 132a is conical, but other configurations may also be considered. As described above, an energy application tool 120 such as a knock rod 110 may be installed in the shell 132 for axial movement. The handpiece also includes a drive mechanism 160 installed in the shell 132 for axially driving the knock rod 120 in the shell 132 between the retracted position and the extended position during operation. The drive mechanism 160 may include an electromagnetic coil 156, which will be discussed further below. The knock rod 110 may have a permanent magnetic ensemble 157 installed at the end away from the free end. The electromagnetic coil 156 of the drive mechanism 160 may be located behind the other end of the knock rod 110, so that the outer diameter of the handpiece 104 is relatively small.

[0155] The mounting mechanism of the energy application tool 110, such as the tap rod 110, can be formed by bearings 1003 and 1004, such as Figure 2a and 2b 1000. As shown, the drive coil 156 is used to receive or support the knock rod 110 in a substantially frictionless manner. A magnetic or propulsion coil 156 can be located in the housing 132, adjacent to the permanent magnet 157, and axially behind the permanent magnet 157. The magnetic coil 156 and the permanent magnet 157 form a drive for the forward and return movement of the knock rod 110. The drive coil 156 can be an integral part of the housing 130 and can be connected to the supply hose or pipeline 1000.

[0156] The two bearings 1003 and 1004 may be substantially frictionless and may include Figure 2a and 2b A plurality of radially inwardly extending ridges are shown separated by axial openings 1400. The axial openings 1400 of the bearing 1003 allow air to move between chamber 1500 and chamber 1600, which is separated from chamber 1600 by the bearing 1003 and formed between the inner wall surface of the housing 132 and the knock rod 110. Thus, the movement of air between these chambers 1500 and 1600 can compensate for the movement of the knock rod 120.

[0157] Refer again Figure 2f, sleeve 108 is positioned toward and extends beyond end 132a. Sleeve 108 wraps around the end of housing 132a and is flattened at end 116 thereof for easy positioning on the surface of an object during operation. The sleeve helps position handpiece 104 on the object to stabilize the handpiece during operation. Figure 2f As shown, the sleeve 108 may also include a fin 118 protruding from a portion of its end 116, so that when the open end 116 of the sleeve 108 is in contact with the surface of the object being measured, the fin 118 can rest on a portion of the top of the object. The fin 118 and the sleeve 108 both contribute to the stable and repeatable positioning of the handpiece 104 relative to the object, and the fin 118 can be placed at a substantially the same distance from the top of the object each time. As described above, the object may include an anatomical structure or a physical structure.

[0158] Figure 2 1 and 2. The system may include a handpiece 100 having a housing 102 for housing an energy application tool and a sensing mechanism, such as a handpiece 100 having an energy application tool 110 and a sensing mechanism 111. Figure 2As shown in the block diagram of , the sensing mechanism 111 is generally placed near the end of the energy application tool 110 to receive force or energy from the target. Generally, the handpiece can refer to a handheld device, but can also include, but is not limited to, any other suitable form for the desired application, such as a mounted device or a tool / machine / robot articulated device. The handpiece 100 can also be interchangeably referred to herein as, for example, a device or an instrument. In some embodiments, as shown in the figure, the energy application tool 110 can be mounted in the housing 102 for axial movement in a direction A toward the object, and this axial movement can be completed by a drive mechanism 140. The drive mechanism 140 can generally be a linear motor or actuator, such as by generating a magnetic field to affect the axial position of the energy application tool 110, and the magnetic field interacts with at least a portion of the energy application tool 110 through magnetic interaction to control the position, velocity and / or acceleration of the energy application tool 110. For example, an electromagnetic coil at least partially arranged around the energy application device 110 can be energized to propel the energy application tool 110 forward toward the object to be measured, as shown by the electromagnetic coil 140. For example, alternatively, the electromagnetic coil may also be energized to push the energy application tool 110 backward in preparation for a subsequent impact. Other elements such as a resilient magnetic element may also be included to assist in repositioning the energy application tool 110, for example, after being propelled by the electromagnetic coil. As shown by power supply 146, the drive mechanism 140 and / or other portions of the instrument may generally be powered by a power source, which may be a battery, a capacitor, a solar cell, a transducer, a connection to an external power source, and / or any suitable combination / multiples of the above. An external connection to a power source may be provided to power the handpiece 100 or to charge an internal power source (e.g., power supply 146), such as Figure 2 The power interface 147 in the device may include, for example, power contacts for direct conductive charging, or the power interface 147 may utilize wireless charging such as inductive charging.

[0159] In other embodiments, the energy application tool 110 may be configured to move generally along a direction A that may be perpendicular or substantially perpendicular to the longitudinal axis of the housing 102, such as Figure 2c100 is shown in a block diagram of the handpiece 100 in FIG. As shown, the energy application tool 110 can be, for example, generally L-shaped to accommodate interaction with the drive mechanism 140 and protrude in direction A, substantially perpendicular to the axis of the housing 102. As shown in the example, the drive mechanism 140 can act on the energy application tool 110 to cause it to rock on the pivot 110a, causing its tip to move in direction A. The drive mechanism 140 can utilize, for example, an alternating magnetic element that can act on the energy application tool 110 to cause it to move alternately in two directions, such as up and down. In another example, the bend of the L-shaped energy application tool 110 (for example, shown as bend 110b) can include a flexure and / or a deformable structure, so that the linear force applied by the drive mechanism 140 can push the energy application tool along direction A at the tip by transmitting forward motion around the bend 110b. For example, the bend 110b can include braided, segmented, spring-like and / or otherwise bendable portions, which can also transmit motion and / or force around the bend. Typically, the shape of the L-shaped energy-application tool 110 can generally include angles other than 90 degrees, for example, between approximately + / - 45 degrees from the rearward portion 110d. In some embodiments, the energy-application tool 110 can also include multiple separable parts, such as parts 110c and 110d, so that, for example, part 110c can be removed and placed between users or patients to assist in preventing cross contamination. Typically, as described below, the separable parts can include an interface that couples them for measurement so that they essentially act as a single energy-application tool 110.

[0160] In some embodiments, Figure 2d and 2e As shown, the L-shaped energy application tool 110 can be rocked on the pivot 110a by, for example, an external force applied from the drive mechanism 140. Figure 2d As shown, the drive mechanism 140 can apply an alternating force to the energy application tool 110 to cause it to swing about the pivot 110a, for example, by applying a force D from the portion 140d to the rearward portion 110d, thereby causing it to swing in the direction A' away from the target object, or as shown in FIG. Figure 2e As shown, the energy application tool 110 is driven in the direction A by applying a force E from the portion 140c to the rearward portion 110d, thereby causing the energy application tool 110 to swing in the direction A" toward the target object. The forces D and E may be applied by any suitable means, such as by applying a magnetic force to the energy application tool 110, which may include a magnetic or metallic element that can respond to the application of force from the drive mechanism 140. In general, the shape and curvature of the swinging motions A' and A" may be designed so that the energy application tool 110 strikes the target object in a direction substantially perpendicular to the surface of the target object, such as Figure 2eAs shown, the swing A" is about the bend 110b in a direction substantially perpendicular to the bend 110c. Figure 2d As shown, to reset the device 100 for subsequent measurements, portion 140d can apply a return force D so that the swing motion A' returns the energy application tool 110 to a withdrawn or resting state. Generally, the interior of the device 100 can be adapted to allow the rocking motions A' and A" without interfering with the energy application tool 110.

[0161] Other examples of endpoint devices may include, but are not limited to, those described in, for example, U.S. Patent Nos. 6,120,466, 7,008,385, 6,997,887, 9,358,089, 9,869,606, US10,488,312, PCT / US17 / 69164, PCT Patent Application Serial No. PCT / US20 / 40386, U.S. Patent Publication No. 20190331573, PCT / US2018 / 068083, and / or PCT Publication WO2019133946, the entire contents of which are incorporated herein by reference.

[0162] In general, the system of the present invention includes a program logic module, as described above, which is generally used to process the signal received from the knock measurement device behind the measurement object. The program logic module can generally apply an algorithm to form an initial guess that at least partially decomposes the signal into at least one component sub-signal. The result can generally be an initial guess decomposition. Then, the program logic module can perform optimization to minimize the difference between the initial guess decomposition and the original signal, for example, by an optimization algorithm. For example, this can include optimizing (i.e. minimizing) the error, absolute error or square error between the signal and the guess decomposition. For example, gradient descent (also known as steepest descent) is a first-order iterative optimization algorithm for finding the local minimum of a differentiable function, which repeats steps in the opposite direction of the function gradient (or approximate gradient) at the current point because this is the direction of the steepest descent. Gradient descent optimization can be performed using commercially available or open source artificial intelligence or high-resource computing tools, such as Google TensorFlow, etc. Optimization algorithms can also be used to guess the decomposition of the missing bottom or negative amplitude of auxiliary signals or sub-signals.

[0163] After optimizing the decomposition, the program logic module can then (e.g., automatically or in conjunction with an expert operator) identify and resolve potential errors or defects in the decomposition and perform subsequent rounds of guesswork decomposition and optimization to form an optimized decomposition. In some examples, the potential errors or defects can represent implausible or impossible physical situations, the result of obvious or possible mathematical errors, overly complex solutions, and / or other results that indicate improper decomposition or optimization. Corrections for potential errors or defects in the decomposition can be further incorporated into the program logic module so that it can better identify situations where such potential errors or defects may occur due to the characteristics of the signal, thereby being able to perform more efficient decompositions without generating solutions with such potential errors or defects.

[0164] After eliminating potential errors or defects, it is preferred that the decomposition guessing and optimization steps can be repeated to obtain a different, modified or new optimized decomposition that does not include or reduces the eliminated potential errors or defects. The process can be re-iterated as necessary to eliminate or reduce any potential errors or defects in the decomposition to a desired level.

[0165] The optimized decomposition may also be compared to other prior decompositions (eg, as a whole, by numerical metrics, by common characteristics, by measured physical parameters, etc.) to assist the clinician in making decisions or interpreting information about the structural characteristics of the object.

[0166] In some embodiments, the object can be real, artificial or simulated oral tissue, such as teeth, oral prostheses, appliances, implants or splints, and / or related tissue or orthopedic implants. In general, a percussion measurement device can apply mechanical energy to an object by percussion, and measure the energy returned to the device, for example by measuring the force / energy / displacement, etc. returned to the percussion measurement device, such as by measuring the force, energy, displacement or other physical return value on a sensing mechanism over a period of time.

[0167] In other embodiments, the object can be a machine, an industrial structure or a complex, which may include but is not limited to a polymer composite structure, which includes a honeycomb or layered honeycomb or a metal composite structure; an aircraft structure, a car, a ship, a bridge, a tunnel, a train, a building, an industrial structure, which includes but is not limited to a power generation facility, an arched structure or other similar physical structures.

[0168] In general, as described above, the system of the present invention may include a program logic module that incorporates a machine learning algorithm trained on a large signal data set to obtain an optimized decomposition of a sub-signal or part thereof. The training of the program logic module may be performed in a controlled or prefabricated environment, such as in a laboratory, manufacturing or development environment, before an end user (such as a dentist or other clinical / industrial clinician) uses the system in a non-training or non-testing environment. Of course, the signal (i.e., clinical signal) collected from the actual operating environment may be included in a pre-existing data set or expanded, as described below, to continue to improve the system by performing additional training on the expanded data set, making it a living system. The data set may be typically based on an object of a single type or a group of related types, so that training may produce an applicable program logic module for a specific field or application (such as the dental field). In certain embodiments, the object may be a real, artificial or simulated oral tissue, such as a tooth, an oral prosthesis, an appliance, an implant or a splint, and / or related tissue or an orthopedic implant. Generally speaking, a knock measurement device can apply mechanical energy to an object by knocking (if not simulated on a computer), and measure the energy returned to the device after impact with the object or the deceleration of the impactor by, for example, measuring the force / energy / displacement, etc. returned to the knock measurement device, for example, by measuring the force, energy, displacement or other physical return value on the sensing mechanism over a period of time to form a signal. For an object simulated on a computer, this can be done, for example, by using, for example, Figure 4d The energy return may be simulated using a finite element analysis (FEA) model as shown, or a constructed signal or sub-signal may be created on a computer or by manipulating / altering a pre-existing signal or sub-signal.

[0169] The present invention may also include a simulation model component that can be used to train program logic modules. FIG. 9 shows an example of a FEA model used as a physical simulation model. The analysis method may involve using a numerical model to simulate actual testing using the apparatus described herein. In general, modeling and simulation may be preferred for training the system and its predictive capabilities using a simulation model that may specifically represent a test object that has not yet been tested and is not readily available for actual physical testing, etc.

[0170] In one example of modeling, a physiologically accurate 3D model of a mandibular second molar was created using a solid modeling computer-aided design program using 3D X-ray computed tomography tooth data, but the same process can be applied to other teeth and other solid objects. The model includes both enamel and dentin as well as the pulp chamber, periodontal ligament (PDL), and surrounding bone, an example of which is shown in FIG. Figure 4d shown.

[0171] The solid model is then exported to a computer-aided engineering program to mesh the solid. Nonlinear finite element solvers can be adapted to model nonlinear material behavior, such as that reported for PDL and using transient environmental conditions including knocking. It is necessary to include a knock bar in the current simulation model in order to fully analyze the knock event using comparisons with experimental data. The elastic modulus of the knock bar, its mass and initial velocity are input into the program. The knock force generated is measured by a piezoelectric sensor in the bar.

[0172] The FEA model may include a large number of elements, for example, each including about 500,000 to about 1,000,000 elements. Second-order isoparametric three-dimensional 4-node tetrahedrons may be used for the PDL, 8-node isoparametric arbitrary hexahedrons may be used for the percussion probe, and linear isoparametric three-dimensional tetrahedrons may be used for the remainder of the model. Boundary conditions may be defined to minimize or prevent free-body motion, so that elements on the outer surface of the object, such as bones, may be constrained. The model is run in time increments of, for example, 4 μs.

[0173] The solution to the equations of motion for the model can be obtained using direct integration methods. In addition, viscous damping can be included in the analysis using classical Rayleigh damping (RD), which facilitates the numerical solution in an incremental manner. The damping matrix D is defined as a linear combination of the mass matrix and the stiffness matrix of the system, and the damping coefficients are specified on an element-by-element basis. Rayleigh damping uses coefficients on the element matrix, given by the equation It is represented by, where D is the global damping matrix, M i is the mass matrix multiplier of the ith element, K i is the stiffness matrix multiplier of the i-th element, α i is the mass damping coefficient of the ith element, β i is the common stiffness damping coefficient of the ith element, γ i is the numerical damping coefficient of the ith element, and Δt is the time increment. In a given model, the same damping coefficient can be used throughout the PDL. For example, the mechanical properties of hard tooth tissue (i.e., tissue outside the PDL) are assumed to have linear elastic and isotropic behavior, following σ ij =C ijkl ε kl , where C ijkl The non-zero components of are functions of the elastic modulus E and Poisson's ratio.

[0174] For example, but not limited to, a particular data set may include data derived from measurements of various types of teeth, dental implants / appliances, and / or oral tissues in the dental field (and may also include simulations of these objects or simulations of data derived therefrom), and preferably have a large number of diverse subjects or conditions so that the trained program logic module can be faced with numerous possibilities for its optimized decomposition. The data set may also be updated or augmented through continued use of the system by end users (i.e., clinicians) in a clinical / industrial setting so that the program logic module can be trained with the augmented data set over time as it is used, thereby improving its performance. The improved program logic module may then be disseminated for use by various clinicians through updates (e.g., through updates to the cloud storage / operation portion of the system).

[0175] In an exemplary aspect of the invention, a signal may be measured from a percussion measurement device that strikes an object or a simulation of such a signal, and may generally be grouped together based on at least one common characteristic in the data set. Examples of such characteristics may include object type, object size, location in a given space or environment, physical condition (e.g., amount / type / location of damage, physical recovery, etc.), location of measurement using a percussion measurement device, age of the object, treatment or surgery administered to the object, and / or any other applicable physical condition or simulation thereof. These groupings are not limited to and need not be exclusive, and multiple, different, overlapping, self-organizing, and / or composite groupings may be used.

[0176] In an exemplary embodiment of the present invention, the program logic module may generally apply an algorithm, such as a machine learning training algorithm as described above, to form an initial guess of at least partially decomposing the signal into at least one of its component sub-signals. Other algorithms of the program logic module, such as a machine learning training algorithm, an FFT or Fourier-like operation, a gradient descent or the like, may also be used in combination to form the remainder of the initial guess decomposition, as applicable. The result may generally be an initial guess decomposition, such as Figure 4b , where the complete signal is decomposed into 8 component sub-signals and 1 residual sub-signal (e.g., which may represent noise, a small or non-significant portion of the signal). The program logic module may then perform an optimization, as described above, to minimize the difference between the initial guess decomposition and the original signal, such as by an optimization algorithm, such as gradient descent or a similar algorithm (e.g., implemented using commercial or open source artificial intelligence or high-resource computing tools, such as Google TensorFlow, etc.). The optimization algorithm may also be used for guessed decompositions where the bottom or negative amplitude of the auxiliary signal or sub-signal is missing.

[0177] As generally described above, after the decomposition is optimized, the program logic module can then (e.g., automatically or in conjunction with an expert operator) identify and resolve potential errors or defects in the decomposition and perform subsequent rounds of guesswork decomposition and optimization to form an optimized decomposition. In some examples, the potential errors or defects may represent implausible or impossible physical situations, the result of obvious or possible mathematical errors, overly complex solutions, and / or other results that indicate improper decomposition or optimization. Corrections for potential errors or defects in the decomposition can be further incorporated into the program logic module so that it can better identify situations where such potential errors or faults may occur due to the characteristics of the signal, thereby being able to perform a more efficient decomposition without generating solutions with such potential errors and faults.

[0178] After resolving potential errors or defects, it may be preferable to repeat the decomposition guessing (e.g., using an algorithm trained by machine learning) and optimization steps to obtain a different, modified, or new optimized decomposition that does not include or reduces the resolved potential errors or defects. The process may be re-iterated as necessary to eliminate or reduce any potential errors or defects in the decomposition to a desired level.

[0179] The optimized decomposition, its properties, and the method used to derive it can then be incorporated into the system by encoding it in an algorithm (such as a machine learning or deep learning algorithm), enabling the system to more efficiently and accurately decompose new signals encountered, such as signals obtained from a tap measurement device that generates signals from a physical object. The system is also able to apply the encoded algorithm to new simulated data sets, such as data sets that simulate hypothetical or novel physical properties / scenarios.

[0180] In some embodiments, the system of the present invention can not only measure and analyze the structural characteristics of the measured object, but can also be trained to detect whether the correct energy application parameters are used in a clinical setting. For example, using machine learning, the system can notice that the detected response indicates that too much force is applied to a physical tool such as a tapping rod, the duration of each application is too short or too long, the tapping may not be applied in the correct location, etc., or a combination of the above situations, and can automatically adjust the device settings to compensate or prompt the clinician to tap the object in a different location to produce a more optimal response. The system can also identify these situations as the result of high uncertainty or signal loss detected during the decomposition process.

[0181] In general, and without being limited or constrained by any particular theory, the energy or force returned to the knock measurement device may only form a signal with a positive amplitude relative to a baseline (e.g., the x-axis of the signal), and may be a missing portion of the signal with a negative amplitude relative to the baseline. In general, some sensing mechanisms may be inherently limited to detecting response signals in a single direction, such as only the compression direction of a piezoelectric force sensor (i.e., because piezoelectric elements generally only generate signals in response to compression). This typically results in the loss of at least some of the signal in a signal whose amplitude contains a negative portion relative to a certain baseline or other energy distribution diagram, such as the application of energy to an object due to the knock, because any response from the object will cause the object to vibrate, move away from the sensing mechanism, or lose mechanical contact or connection, and the sensing mechanism may not be able to detect a portion of the signal due to insufficient compression in the appropriate direction detected by the sensing mechanism. Therefore, the signal generated by the knock measurement device may be partially incomplete. In addition, such sensing mechanisms are typically one-dimensional and cannot inherently filter or separate individual sub-signals that may be present in the entire received signal because they are typically superimposed on each other. While methods exist to decompose such signals into their component sub-signals, the "missing" portions of the signal can present challenges, as standard decomposition methods are not adapted to signals with missing portions. Furthermore, in some tap measurements, the length of time over which the signal can be measured presents additional challenges, as the time may be too short to capture enough cycles (or partial cycles) of any signal or its component sub-signals for accurate decomposition.

[0182] In one exemplary aspect of the invention, the system may be adapted to adapt to or analyze signals generated from a tap measurement on an object, wherein the sensing mechanism results in at least a portion of the return signal being limited or lost. In general, some sensing mechanisms may be inherently limited to detecting a response signal in a single direction, for example, only in a compression direction, such as Figure 2 , 2a , 2c, 2d and 2e for the arrangement shown for the sensing mechanism 111 (eg a piezoelectric force sensor). As described above, this will typically result in at least some signal loss, resulting in the signal generated by the tap measurement device being potentially partially incomplete.

[0183] Generally speaking, and without being limited or bound by any particular theory, in the dental field, recognizing that the PDL typically forms the majority of the damping from a strike, the energy / force return measured in the signal should appear as a sinusoid (e.g., like a sound wave), but typically only forms a single-peaked, approximately Gaussian form, and it can be inferred that the PDL response may appear in the signal as the first half cycle of the sinusoid, whose amplitude is "negative." Therefore, the PDL sub-signal can generally be inferred to be the main sub-signal component of the signal, which is then removed, treated as the carrier of the remaining sub-signal, and able to be decomposed using typically more standard decomposition methods (such as fast Fourier transforms or similar Fourier-like operations). This can be particularly useful because the remaining sub-signal (if present) may typically be a sinusoid of shorter period and / or amplitude (e.g., a decaying sinusoid) that can be captured in the measurement, but is obscured or incorrectly characterized as other Gaussian forms due to the presence of the larger single half-cycle PDL sub-signal. In some cases, it may be recognized that the PDL sub-signal (if present) would not form the primary sub-signal (e.g., the PDL is damaged, weakened, non-existent, etc.), and it may be determined or assumed that another physical characteristic (e.g., a crack or other damage in the tooth or tooth-like structure) rather than the PDL may form the primary sub-signal.

[0184] In some exemplary embodiments, the system may utilize machine learning methods to process and / or analyze signals that inherently lack negative amplitude portions and attempt to reconstruct or treat the signal as the missing portion or portion rather than as a complete response measured. In the dental field, machine learning algorithms and methods may be trained on a large set of collected percussion data (e.g., ERG) from a variety of teeth with different characteristics, such as different types of teeth (e.g., incisors, bicuspids, canines, molars, etc.), sizes, number of roots, varying degrees of physical damage (e.g., fractures, cavities, etc.), degrees or types of restorations (e.g., artificial crowns, fillings, etc.), ages, etc., to train the algorithms and methods to enable them to decompose newly encountered signals into component sub-signals, such as a collection of sinusoidal sub-signals that form a signal or its approximation.

[0185] Figure 3 It is shown that by using Figure 2 , 2a , 2c, 2d and 2e are examples of signals (shown as ERG or FRG) generated by the devices in Figures 2c, 2d and 2e striking an object (e.g., a tooth) and recording the force returned to a piezoelectric force sensor (e.g., sensing mechanism 111, as shown below). As shown, since a portion of the original signal generated by the object being struck is lost, the signal does not record any signal above a given threshold (i.e., the X-axis). In general, the signal returned from the object being struck can generally form a more sinusoidal shape (e.g., similar to Figure 4), which can be rectified or "cut off" by a sensing mechanism (e.g., a piezoelectric force sensor) to produce only the portion of the signal above a threshold (e.g., Figure 4a truncated signal in ).

[0186] Data sets used to train machine learning algorithms or compare to clinical signals, such as those discussed in the context, may also be grouped in other ways, such as by the location of the taps on the object (e.g., for teeth that are buccal or mesial, distal or proximal, etc.), the location of the object relative to other reference points (e.g., mandibular vs. maxillary in the mouth), the amount / frequency / number of taps, or any other appropriate type of grouping. In such embodiments, the signal may be analyzed and processed as at least one signal in which a portion is missing from the measurement (e.g., the lower half of a sinusoid below a given threshold is missing from the signal, such as the second half cycle of a sinusoid that crosses below the threshold, forming a Gaussian-like shape in the first half cycle). Figure 3a This can lead to better interpretation of the data because the overall response can be considered (or at least an approximation of it to explain the missing portion of the signal) rather than treating the signal as only the portion that is present in the signal (i.e., assuming the signal is intact). Additionally, some data analysis operations may not correctly interpret data from missing portions of the signal (e.g., standard Fast Fourier Transforms or other similar Fourier-like operations).

[0187] In one exemplary aspect of the invention, a signal may be identified, analyzed, and / or processed as a sum or aggregation of multiple different sub-signals resulting from the interaction of energy applied to the tooth with structural features of the tooth and / or surrounding tissue / structure. Without being bound by any particular theory, a signal may generally represent multiple different sub-signals, each of which is generated by a separate physical structure or feature of the tooth or surrounding the tooth, and generally each of which may take the form of an approximation of a sinusoid, such as Figure 4b , such as, but not limited to, a decaying sinusoid (e.g., a sinusoid that decays exponentially in response to dissipation of the material / structure / motion of the object), as shown in Figure 4 and 4b The summation or aggregation of the sinusoids to form the overall shape of the signal (or the detected portion, i.e., there is no missing or "bottom" portion below a given threshold for detection or measurement, as shown in FIG. Figure 3 , 4a 4b (full signal). In some embodiments, the signal can be decomposed into a "sparse" or finite number of sinusoids (i.e., a bounded number of sinusoids to give the original signal or an approximation thereof) to identify, without being bound by any particular theory, that each sinusoid is caused by at least one element of the tooth, its restoration (if any), and / or surrounding tissue. Examples of sparse decomposition are shown in Figure 4b , where the incompletely decomposed parts are shown as residuals.

[0188] In some embodiments, it is generally recognized that the dominant Gaussian form in the signal from a tooth, particularly a signal from an original or undamaged tooth, is primarily caused by the response of the PDL (periodontal ligament), such as Figure 4e As shown, the PDL absorbs kinetic energy from the energy application tool, which typically looks similar to Figure 3a . For an undamaged healthy tooth, the impact energy from chewing is attenuated by the PDL at the healthy bone-natural tooth interface. Even in cases where the PDL may be missing or defective, the portion of the signal referred to as the PDL portion refers to the portion of the signal that is due to the anchoring of the tooth or implant directly or indirectly in the bone, roughly producing a pendulum-like response. Often, the PDL portion of the signal can constitute the majority of the amplitude of the signal, such as Figure 4c The single peak signal in the PDL portion of the image is shown and can generally be interpreted as a carrier wave that can be used to isolate and / or separate the response from other elements of the tooth or surrounding tissue, for example by subtracting the PDL portion from the signal.

[0189] In some exemplary embodiments, sinusoidal decomposition of a signal using machine learning may generally include the steps described above with respect to more general signals and their sub-signals, where the sinusoids have some specificities for processing dental objects such as teeth and implants. For example, these steps can typically include de-tapering the signal (e.g., removing signal deformations near the x-axis of the signal, such as deformations caused by the energy application tool sticking to the tooth), finding an initial guess for the larger and / or more pronounced sinusoidal components of the signal (e.g., typically sinusoids generated by the PDL or other major sub-signals), finding an initial guess for the remaining sinusoidal components of the signal (e.g., sinusoids from cracks, damage, interlaminar separations, or other features), including guesses for the frequencies (e.g., by performing a Fourier transform-like operation such as a fast Fourier transform (FFT) on the signal after subtracting the initial PDL sinusoidal guess, by machine learning or artificial intelligence methods, gradient descent-based methods, etc.), performing optimization to minimize the difference between the original signal of the signal and the composite sum of the initial guess sinusoids to produce candidate sinusoidal decompositions, identifying / repairing decomposition defects or errors (e.g., untrustworthy or negatively indicated decomposition results) to rerun the above decomposition steps as needed to remove them, and selecting the best or desired candidate from the resulting decomposition. The resulting decomposition and related data / results / visualizations can then be displayed or output in a human-readable form so that they can be used or interpreted by a human physician or other user, e.g., for clinical diagnosis, monitoring, and / or treatment planning. The system-generated decomposition can also typically include determining or calculating uncertainty measures at various steps of the decomposition, such as calculating error values ​​at various steps or specific calculated values ​​in the decomposition.

[0190] In some embodiments, the composite sum of the sub-signals may not approximate the original signal well or at least in some parts. For example, since the original signal usually has only positive amplitudes above its baseline, it is conceivable that the possible composite sum of the sub-signals may contain negative amplitude parts. During the optimization process, it can be selected or preferred as needed to take the maximum value of zero and the composite sum value (i.e., max(0, sum)) in those parts of the negative amplitude to assist in the optimization. In other examples, a sensing mechanism or other physical arrangement of the knock measurement device can be used, which can capture the negative amplitude part that the directional polarization setting (i.e., the piezoelectric sensing element as described above) cannot capture, such as using a strain gauge-based sensing mechanism or other non-rectifying sensor, or using a design of the knock measurement device, in which the energy application tool or other part that transmits energy / force to the sensing mechanism can follow or "stick" to the target object during the knocking process to capture forward and reverse motion / force waves, etc. In this case, the max(0, sum) operation may not be preferred in general because the resulting signal may not be limited to positive amplitudes.

[0191] In some exemplary embodiments, the program logic module may be trained using machine learning or artificial intelligence methods as described above with respect to training to perform initial and / or subsequent guesses (i.e., after an optimization step) on a primary sub-signal in a signal to allow the system to generate a remaining portion of the signal without the primary sub-signal, wherein a sinusoidal decomposition of the remaining portion is performed using other resources of the program logic module or a computing device of the system (e.g., a local computer or a cloud service) to decompose the complete signal into sub-signals. In general, without being bound by any particular theory, a plurality of signals may contain a primary sub-signal that, when determined and removed from the signal, produces a remaining portion of the sub-signal that can be determined without complex machine learning or artificial intelligence methods, such as by utilizing FFT or Fourier-like operations, gradient descent-based methods, etc., thereby saving intensive resources for machine learning or artificial intelligence operations. For example, removing a primary sub-signal (e.g., a PDL sub-signal) may produce a more easily discernible sub-signal (e.g., a sinusoidal curve, etc.), such as by not leaving out important portions of the sub-signal, "negative" amplitude portions, or other features that may require or benefit from machine learning or artificial intelligence methods, which may be used as needed.

[0192] In some embodiments, a program logic module or another component of the system (e.g., a local computer or a cloud service) can perform initial and / or subsequent guesses on the primary sub-signal (i.e., after an optimization step), for example by fitting at least a portion of the signal, such as to a basis function (e.g., a Gaussian curve, a sinusoidal curve, a quasi-sinusoidal curve, etc.).

[0193] For example, the program logic module can generally apply an algorithm to fit the signal to a basis function, such as by utilizing gradient descent optimization, Levenberg-Marquardt optimization, and / or any other similar or appropriate method or combination / multiples of the above methods. The system can then generate a remaining portion of the signal without the primary sub-signal, and perform a sinusoidal decomposition of the remaining portion using other resources of the program logic module or a computing device of the system (e.g., a local computer or a cloud service) to decompose the complete signal into sub-signals. In general, and without being bound by any particular theory, a large number of signals may contain a primary sub-signal, which when determined and deleted from the signal produces a remaining portion of the sub-signal, which can be determined without the need for complex machine learning or artificial intelligence methods, such as by utilizing FFT or Fourier-like operations, gradient descent-based methods, etc., thereby saving intensive resources for machine learning or artificial intelligence operations. For example, removing a primary sub-signal (e.g., a PDL sub-signal) can produce a sub-signal that is more easily distinguished, such as by not missing important portions of the sub-signal, "negative" amplitude portions, or other features that may require or benefit from machine learning or artificial intelligence methods, which can be used as needed. The result can generally be as follows Figure 4b An initial guess decomposition is shown, where the full signal is decomposed into 8 component sub-signals and 1 residual sub-signal (e.g., which may represent noise, a small or insignificant portion of the signal). The program logic module may then perform an optimization, as described above, to minimize the difference between the initial guess decomposition and the original signal, such as by an optimization algorithm, such as gradient descent or a similar algorithm (e.g., implemented using commercially available or open source artificial intelligence or high-resource computing tools, such as Google TensorFlow, etc.). Optimization algorithms may also be used to assist in guessing decompositions where the bottom or negative amplitude of a signal or sub-signal is missing.

[0194] As described above, after the decomposition is optimized, the program logic module can then (e.g., automatically or in conjunction with an expert operator) identify and resolve potential errors or defects in the decomposition and perform subsequent rounds of guesswork decomposition and optimization to form an optimized decomposition. In some examples, the potential errors or defects can represent implausible or impossible physical situations, the result of obvious or possible mathematical errors, overly complex solutions, and / or other results that indicate improper decomposition or optimization. Corrections for potential errors or defects in the decomposition can be further incorporated into the program logic module so that it can better identify situations where such potential errors or defects may occur due to the characteristics of the signal, thereby being able to perform a more efficient decomposition without generating solutions with such potential errors and defects.

[0195] The decomposition results of the obtained signals, such as clinical signals from measurements in a clinical environment, can be compared or analyzed in various ways with a dataset (e.g., a large dataset of signals or information / metrics derived from signals of objects similar to the target object in the clinical signal). The dataset can be similar in form or content to the dataset described above for training machine learning algorithms, or can be customized, abridged, expanded, or formed from different sources. In some embodiments, the objects contained in the signals of the dataset can be real, artificial, or simulated, such as oral tissues, such as teeth, oral prostheses, appliances, implants or splints, and / or related tissues or orthopedic implants. In general, as described above, the percussion measurement device can apply mechanical energy to the object by percussion (if not simulated on a computer), and measure the energy returned to the device after collision with the object or deceleration of the impactor, such as by measuring the force / energy / displacement, etc. returned to the percussion measurement device, for example, by measuring the force, energy, displacement or other physical return value on the sensing mechanism over a period of time to form a signal. For objects simulated on a computer, it can be, for example, by using a sensor such as Figure 4dThe energy return can be simulated by the FEA shown in the finite element analysis (FEA) model shown, or the constructed signal or sub-signal can be created on a computer by manipulating / changing the pre-existing signal or sub-signal. For example, but not limited to, a specific data set may also include or have available data derived from measurements of various types of teeth, dental implants / appliances and / or oral tissues in the dental field (and may also include simulations of these objects or simulations of data derived therefrom), and preferably have a large number of diverse subjects or conditions so that multiple possibilities for comparison or other analysis of the optimized decomposition of clinical signals can be performed. The data set can also be updated or expanded by the end user (i.e., clinician) continuing to use the system in a clinical / industrial environment to provide additional data for comparison or analysis. The improved data set can be disseminated for various clinicians to utilize through updates (e.g., by updating to the cloud storage / operation portion of the system).

[0196] In an exemplary aspect of the invention, signals can be measured from a percussion measurement device that strikes an object or a simulation of such a signal, and can generally be grouped together based on at least one common characteristic in the data set. Examples of such characteristics may include object type, object size, location in a given space or environment, physical condition (e.g., amount / type / location of damage, physical recovery, etc.), measurement location using a percussion measurement device, object age, treatment or surgery administered to the object, and / or any other applicable physical condition or simulation thereof. These groupings are not limited to and need not be exclusive, and multiple, different, overlapping, self-organizing, and / or complex groupings may be used.

[0197] In another exemplary aspect of the invention, the system can generate or calculate various numerical metrics based on the decomposition of the signal that show statistically significant differences in the data set, so that these numerical metrics can be used in probability distributions or heat maps to assist in the prediction or detection of different physical characteristics or attributes by comparison with numerical metrics derived from the decomposition in a clinical setting. For example, a numerical metric generated from a data set containing known physical characteristics (e.g., the type of lesions on a tooth) can be used for comparison using the same type of numerical metric derived from a clinical measurement, and the probability or degree of match can be determined by the system to output a likelihood of a match with that particular physical characteristic.

[0198] Types of numerical metrics may include, but are not limited to, normal fit error (NFE), which is the overall error (difference) between an ideal curve (e.g., generated by a defect-free object) and the actual test data. These results may be calculated from the signal. In the case of decomposing sinusoids or other sub-signals from the original full signal, the NFE may be calculated for the decomposition as the error between the decomposed primary sinusoid (e.g., the PDL sinusoid) and the full signal (i.e., the sinusoidal decomposition NFE or SDNFE). Other numerical metrics may include the PDL partial period or frequency, the frequency or period of other sinusoids, the decay rate of an exponentially decaying sinusoid, the amplitude of a sinusoid, the number of periods in a sinusoid, other statistically significant metrics, and / or weighted versions / combinations of any of the above (i.e., taking into account close value or tapered weighting for a distribution of values).

[0199] Figure 5 , 5a 5a and 5b show examples of heat maps showing the distribution of numerical metrics in a population. As shown, darker areas represent higher counts and lighter areas represent lower counts, with circles representing weighted averages calculated by standard deviations. Such heat maps may be preferred or useful when assessing the probability that a measurement matches certain physical characteristics or combinations of characteristics based on cohorts in a systematic data set that are consistent with those characteristics or combinations. For example and illustrative purposes, Figure 5 can show the general population in the data set, while Figure 5a A cohort of teeth identified as having a particular type of injury may be shown superimposed on a general population, while Figure 5b Another contemporaneous group may be shown (e.g. a group of "good" or healthy / undamaged teeth superimposed on a general population). For example, the placement of measurements applied to these heatmaps may then assist in determining whether the subject being measured is likely to fit neatly into one or the other population.

[0200] In general, heat maps can be of any appropriate dimension (e.g., one, two, three, etc.). Some higher dimensions may be difficult for humans to visualize or interpret, and thus computerized interpretation or reduction to numbers or simplified graphical representations may be utilized to assist human users in their interpretation.

[0201] In another aspect of the invention, the machine learning algorithms and methods of the system, such as may be generally separated or isolated by use, training or purpose from the machine learning algorithms discussed for other uses in relation to signal decomposition or elsewhere, may be trained on large, collected data sets of knocks that may be annotated with features (which may be determined by an "expert" or other trusted characterizer, or by a machine learning algorithm) to identify, guess with some probability, and / or associate a measured signal with a particular feature, or to select an appropriate manner for further analysis or algorithmic manipulation to produce a useful output for user use or interpretation, such as for selecting an appropriate plan for further diagnosis, monitoring, treatment, etc. For example, a signal from a data set may be associated with certain physical features determined by an expert, such as a dentist identifying the type or extent of a tooth lesion through physical examination, x-ray imaging, deconstruction, etc., which the system may use as additional correlation data for the signal. As described above, training on these groupings and annotated correlations may be further used for machine learning-assisted analysis of heat maps of numerical metrics, such as by using a machine learning algorithm trained on such to find correlations and probabilities that match the population. This may be preferred in heat maps or comparisons where the dimensionality or other complexity presents a challenge to human interpretation.

[0202] Machine learning algorithms can also be used for comparison to assist in improving efficiency and probabilistic matching with known data sets. For example, basic machine learning methods such as kernel density estimation and / or calibration curve-adjusted Bayesian networks can be used. After the machine learning system is fully developed and / or trained, deep learning methods can also be used to generate more advanced comparison methods.

[0203] In some embodiments, the system can detect numerical metrics in clinical measurements, which can indicate or suggest that the user change certain physical parameters of the clinical measurements, such as changing the parameters of the knock measurement device to generate better or more accurate data. For example, some digital metrics can indicate or imply the use of different knock forces, frequencies or positions used for knocking on the object to illustrate additional information or improve the quality of measurements using the knock measurement device. For example, the system can use information that is not directly derived from the signal in its data set, such as the knock position on the object during the measurement, the knock force, the knock measurement device settings, and / or other related data or factors. The system can then provide feedback for the control of the knock measurement device, such as suggesting or implementing repeated measurements with different settings, positions, timings, etc. This suggestion / indication detection of the system can be trained into a program logic module using a machine learning method similar to the above method. This can help minimize the subjective decisions made by clinicians, which would be impossible without the present invention.

[0204] A drive mechanism 140 supported within the housing, such as a housing 132 Figure 2aAs shown, instructions can generally be received from the system for activating the energy application tool 110 between the rest configuration and the activation configuration to apply a set amount of energy in a horizontal direction; having an inclinometer adapted to measure the inclination of the energy application tool 110 relative to the horizontal plane. For a given object, the drive mechanism 140 varies the energy applied to activate the energy application tool 110 between the rest configuration and the activation configuration based on the inclination to at least approximate the set amount of energy at an inclination other than horizontal. Thus, for the above-mentioned different types of objects, the same drive mechanism 140 for varying the amount of energy applied (e.g., varying the voltage, current, or both) can vary the coil drive time (changing the length of time the coil is energized or activated), can vary the coil delay time (changing the time between drive activities), can vary the number of coil energizations (i.e., varying the number of drive pulses applied), the polarity of the coil, and / or a combination thereof, and is adapted to modulate the energy application process to simulate a substantially horizontal position during measurement. These factors, including different powers, drive times, polarities, and delay times, can be managed by firmware settings that vary the power, drive time, number of drives, number of drive pulses, polarity, and drive delay of the coil to obtain the desired results. Without wishing to be bound by any particular theory, it is speculated that a variety of variations may be employed to achieve the desired results, and the firmware may be designed to select a particular solution or to select the optimal solution for certain situations. The system may suggest, adjust, or otherwise indicate to the user a variety of variations based on the probability of the program logic module detecting or calculating a change in the measured physical parameter, which may help improve accuracy or elucidate more information about the object. The system may perform some or all of these functions automatically, or may prompt or instruct the user to adjust the parameters.

[0205] Generally, when an implant replaces a natural tooth due to injury or disease, the ligament is often lost. However, as described above, the systems and methods of the present invention can also be used to evaluate the structural properties of implant structures using abutments. Some materials used for abutments, such as composites, gold, and zirconia, may produce some sub-signals similar to the PDL response.

[0206] In addition, the present system and method can be used to measure the dynamic response when force is applied to the abutment material, and can also be used to predict suitability or compatibility before implantation, or to select appropriate materials to protect natural teeth adjacent to the implant, and to better select materials to minimize the difference between the way implants and natural teeth respond to impact. This can improve the effectiveness of the abutment structure and increase the selection of potentially suitable materials or material combinations, thereby better caring for patients.

[0207] In some embodiments, a distributed computing environment (such as a cloud) can be used to store device measurements and / or expert annotations. For example, storage on the cloud can allow for the collection of multiple expert annotations simultaneously and reduce the time used to accumulate expert annotation data sets to improve prediction accuracy. In some embodiments, device measurements and / or expert annotations can be collected on multiple instances of the system and merged into one or more of these instances. In some embodiments, device measurements and / or expert annotation entries can be encrypted.

[0208] As previously described, machine learning techniques can include regression (e.g., logistic, linear), clustering (e.g., k-means), neural networks (e.g., deep learning), classifiers (e.g., support vector machines, decision trees, random forests), deep learning, and the like. The underlying machine learning techniques used can themselves be standardized techniques rather than unique; however, the inventors have discovered certain unique adaptations to the types of data stored in case files that make them useful for machine learning algorithms. For example, the signals generated by the systems and methods described above can be used to measure and assess structural properties of an object, whether anatomical or non-anatomical, in a non-invasive manner and / or using non-destructive measurement methods. The structural properties of the object can be determined based on previous work using, for example, Figure 2 , 2a , 2c, 2d and 2e and the devices illustrated in their respective descriptions are identified by measurements taken and captured by the system on the same or other objects, or according to U.S. Patent Nos. 6120466, 7008385, 6997887, 9358089, 9869606, US10488312, PCT / US17 / 69164, PCT Patent Application Serial No. PCT / US20 / 40386, U.S. Patent Publication No. 20190331573, PCT / US2018 / 068083 and / or PCT Publication No. WO2019133946, the entire contents of which are incorporated herein by reference, the measurements can be filtered and converted into a spectrogram for deep learning. The model can then be trained, versioned, and stored in a secure database running on a centralized cloud-based server group.

[0209] Example of Sine Decomposition Optimization Algorithm

[0210] An algorithm may be utilized to optimize the sinusoidal decomposition, for example by utilizing gradient descent optimization, which in some exemplary embodiments may utilize an optimization toolkit (e.g., Google TensorFlow, etc.). An example of a gradient optimization algorithm used by such an optimization toolkit is shown below:

[0211] Optimization Algorithm

[0212] Find {a that minimizes the squared error i},{p i},{o i},{d i}:

[0213]

[0214] in:

[0215]

[0216] and:

[0217] a i = amplitude of sinusoid i

[0218] p i = period of sinusoid i

[0219] o i = time offset of sinusoid i

[0220] d i = exponential decay rate of sinusoid i

[0221] Other constraints

[0222] There are n+1 sinusoids, and sinusoid i=0 is always assumed to be the PDL sinusoid.

[0223] We usually set the basic decay rate of all sine curves to d i ≥3000 and fixed d " =3000.

[0224] Example of detecting and resolving potential errors / defects

[0225] The following table shows examples of potential problems with decomposition performed by program logic modules, such as during training or when decomposing clinical / industrial measurement signals, the uncertainties such problems create, how these problems can potentially be detected, and how the system can resolve these problems.

[0226] Table 1:

[0227]

[0228] Although the invention has been described with respect to specific aspects, embodiments and examples thereof, these are merely illustrative and not intended to limit the invention. The description of the illustrative embodiments of the invention including the description in the abstract and the summary of the invention is not intended to be exhaustive or to limit the invention to the exact form disclosed herein (and in particular, the functions including any specific embodiment, feature or in the abstract or the summary of the invention are not intended to limit the scope of the invention to such embodiments, features or functions). Rather, the description is intended to describe the illustrative embodiments, features and functions so as to provide those of ordinary skill in the art to understand the invention without limiting the invention to any specifically described embodiments, features or functions, including any such embodiment features or functions described in the abstract or the summary of the invention. Although the specific embodiments and examples of the invention are described herein only for illustrative purposes, various equivalent modifications are possible within the spirit and scope of the invention, as will be recognized and understood by those skilled in the relevant art. As noted, these modifications may be made to the invention according to the aforementioned description of the exemplary embodiments of the invention, and will be included in the spirit and scope of the invention. Therefore, although the present invention has been described with reference to specific embodiments thereof, there is room for modification, various changes and substitutions in the above disclosure, and it will be understood that in some cases, some features of the embodiments of the present invention will be used without the corresponding use of other features without departing from the scope and spirit of the present invention as set forth. Therefore, many modifications may be made to adapt a particular situation or material to the basic scope and spirit of the present invention.

[0229] In general, references to "the cloud" may include both Internet-connected computing services and / or resources or services and resources that may exist on smaller or private networks.

[0230] In general, "program logic modules" and software elements can be configured, run, stored, processed and / or executed separately on different computer processors and / or memories, combined with each other on the same computer processor or memory, and / or, if applicable, configured, run and / or stored in different time schedules on any of the above. Nothing in this specification should be implied or interpreted as requiring that any program logic module and / or software element be run on any one or combination of computing processors and / or memories, and any suitable combination or single unit may be used.

[0231] References throughout this specification to "one embodiment," "an embodiment," or "a specific embodiment," or similar terms mean that a particular feature, structure, or characteristic described in conjunction with that embodiment is included in at least one embodiment and is not necessarily present in all embodiments. Thus, the phrases "in one embodiment," "in an embodiment," or "in a specific embodiment," or similar terms appearing throughout this specification do not necessarily refer to the same embodiment. In addition, the particular features, structures, or characteristics of any specific embodiment may be combined with one or more other embodiments in any suitable manner. It should be understood that other variations and modifications of the embodiments described and illustrated herein are possible, in accordance with the teachings herein, and are to be considered part of the spirit and scope of the present invention.

[0232] In the description herein, numerous specific details are provided, such as examples of components and / or methods, to provide a thorough understanding of embodiments of the present invention. However, those skilled in the relevant art will recognize that other devices, systems, components, methods, parts, materials, parts, etc. may be used without one or more specific details. In other examples, known structures, components, systems, materials, or operations are not specifically shown or described in detail to avoid confusing aspects of embodiments of the present invention. Although the present invention may be illustrated by using specific embodiments, this is not and does not limit the present invention to any specific embodiments, and those of ordinary skill in the art will recognize that other embodiments are easily understood and are a part of the present invention.

[0233] As used herein, the terms "comprises," "comprising," "including," "including," "having," "having," or any other variations thereof, are intended to cover a non-exclusive inclusion. For example, a process, product, article, or apparatus that includes a list of elements is not necessarily limited to only those elements but may include other elements not expressly listed or inherent to such process, product, article, or apparatus.

[0234] In addition, unless otherwise indicated, the term "or" as used herein is generally intended to mean "and / or". For example, any of the following satisfies condition A or B: A is true (or exists) and B is false (or does not exist), A is false (or does not exist) and B is true (or exists), and both A and B are true (or exist). As used herein, including the claims that follow, unless explicitly indicated, terms preceded by "a" or "an" (and "the" when the prior basis is "a" or "an") include both the singular and the plural of such terms, unless explicitly indicated within the claims (i.e., the introduction "a" or "an" explicitly indicates only the singular or only the plural). Moreover, as used in the description herein, the meaning of "in..." includes "in..." and "on..." unless the context explicitly indicates otherwise.

Claims

1. A method for providing a structural property analysis system trained by machine learning, comprising: providing or generating a data set comprising a plurality of signals from a plurality of groups of different subjects, the signals being generated from a series of tap measurements on the different subjects and being grouped based on a common characteristic of one of the groups of different subjects; Generate an optimized signal set for each of the signals, each of the optimized signal sets being generated by the following operations: performing a guess decomposition of each of said signals to generate a signal set including at least one sub-signal for each signal; performing an optimization operation to minimize the difference between the signal set and each of the signals to generate an optimized signal set; Identifying and resolving potential errors or deficiencies in each of the optimized signal sets; After eliminating the potential errors or defects, repeating the guess decomposition and the optimization operation to regenerate an optimized signal set; selecting at least one desired signal set for each signal from the optimized signal set to add to the optimized signal set; as well as Incorporating the set of optimized signal sets and the associated method for deriving the set of optimized signal sets into an analysis system MLTA trained by machine learning; as well as The MLTA is connected to a measuring device, which is suitable for generating clinical signal data by tapping a target object and transmitting the clinical signal data to the MLDA so that the MLMA can process the clinical signal data to generate a clinically optimized signal set, compare the characteristics of the clinically optimized signal set with the characteristics of the optimized signal set, and present the results of the comparison in a human-readable form.

2. A structural property analysis system trained by machine learning, comprising: Strike measurement equipment, including: a housing having an open front end and a longitudinal axis; an energy application tool mounted inside the housing, the energy application tool having a rest configuration and an activated configuration; a drive mechanism supported within the housing, the drive mechanism adapted to activate the energy application tool between the rest configuration and the activated configuration to apply a set amount of energy; and a control mechanism connected to provide instructions to the drive mechanism; wherein the drive mechanism varies an amount of energy applied to activate the energy application tool between the rest configuration and the activated configuration based on input from the control mechanism; A program logic module is connected to the control mechanism, and the program logic module is provided by the following operations: providing or generating a data set comprising a plurality of signals from a plurality of groups of different subjects, the signals being generated from a series of tap measurements on the different subjects and being grouped based on a common characteristic of one of the groups of different subjects; Generate an optimized sub-signal set for each of the signals, each of the optimized sub-signal sets being generated by the following operations: performing a guess decomposition of each of the signals to generate a set of sub-signals including at least one sub-signal for each signal; performing an optimization operation to minimize the difference between the sub-signal set and each of the signals to generate an optimized sub-signal set; Identifying and resolving potential errors or defects in each of the optimized sub-signal sets; After eliminating the potential errors or defects, repeating the guess decomposition and the optimization operation to regenerate an optimized sub-signal set; selecting at least one desired sub-signal set for each signal from the optimized sub-signal sets to add to the optimized sub-signal set; and Incorporating the set of optimized sub-signal sets and the associated method for deriving the set of optimized sub-signal sets into a machine learning trained analysis system MLTA; and connecting the MLTA to the tapping measurement device, the tapping measurement device being adapted to generate clinical signal data by tapping a target object and transmitting the clinical signal data to the MLTA to enable the MLTA to process the clinical signal data to produce a clinically optimized sub-signal set, and to compare characteristics of the clinically optimized sub-signal set with characteristics of the optimized sub-signal set, and to determine a physical parameter associated with the comparison; A control regulator is connected to the program logic module and the control mechanism, the control regulator being adapted to output a change to the instruction in response to the MLTA output suggesting a change in the physical parameter.

3. A method for providing a structural property analysis system, comprising: providing a program logic module PLM configured to receive an input of a signal to generate a signal from a tap measurement performed by a tap measurement device PMD; Connecting the PLM to the PMD; Performing tapping measurement on a tooth-shaped object with the PMD to generate the signal with the PLM; performing a guessing of a dominant sub-signal of the signal by fitting the signal to a basis function; subtracting the primary sub-signal from the signal to form a remainder; performing a guessed sinusoidal decomposition on the remaining portion to generate a secondary sub-signal, the secondary sub-signal being added to the primary sub-signal to form an approximation of the signal; performing an optimization operation to minimize the difference between the approximation of the signal and the signal to generate a set of optimized sub-signals; Identifying and resolving potential errors or defects in each of the optimized sub-signal sets; Repeating the guessing, guessed sinusoidal decomposition and optimization operation on the main sub-signal to regenerate the optimized sub-signal set after eliminating the potential errors or defects; selecting at least one desired sub-signal set from the optimized signal set; as well as The desired set of sub-signals is presented in a human-readable form.

4. The method according to any one of claims 1 and 3, wherein the performing of the optimization operation comprises: The difference between the maximum value of the sum of each of said sub-signal sets and zero and each of said signals is minimized to generate an optimized signal set.

5. The system of claim 2, wherein performing the optimization operation comprises: The difference between the maximum value of the sum of each of the sub-signal sets and zero and the signal is minimized to generate an optimized sub-signal set.

6. The method according to any one of claims 1 and 3, wherein the at least one sub-signal comprises a waveform from a periodontal ligament (PDL) damping response.

7. The method according to any one of claims 1 and 3, wherein: The guess decomposition is performed on at least one member of the set of sub-signals using a machine learning algorithm.

8. The method according to any one of claims 1 and 3, wherein the signal set comprises at least one sinusoidal sub-signal.

9. A method according to any one of claims 1 and 3, wherein the signal is generated by a tap measurement device, which only records return signals at the sensing element with positive amplitude relative to a threshold, and the at least one sinusoidal waveform is determined by the guess decomposition, which interprets the signal as a waveform that may contain missing signals with negative amplitudes from the signal.

10. The method according to any one of claims 1 and 3, wherein the guess decomposition further comprises performing frequency guessing on the at least one sub-signal.

11. The method according to any one of claims 1 and 3, wherein the frequency guessing comprises Fourier transform or Fourier-like operation.

12. The method according to any one of claims 1 and 3, wherein the guess decomposition further comprises performing frequency guessing on the at least one sub-signal using a Fourier transform or a Fourier-like operation, wherein the Fourier transform or the Fourier-like operation is not optimized for sinusoidal curves that are missing a portion of the sub-signal below the threshold.

13. A method according to any one of claims 1 and 3, wherein the common characteristics of the groupings of different objects are selected from the group including the following items: tooth type, tooth size, tooth age, degree or type of tooth restoration, degree or type of tooth damage, mandibular position, maxillary position, number of tooth roots, dental or orthodontic treatment, and the location of tap measurements on the objects.

14. The method according to any one of claims 1 and 3, wherein the clinically optimized signal set and the characteristics of the optimized signal set set are selected from the group consisting of: the period of the periodontal ligament damped response PDLP, the frequency of at least one of the sub-signals, the decay rate of an exponentially decaying sinusoid, the amplitude of the sinusoid, the number of periods in the sinusoid, and combinations thereof.

15. The method of any one of claims 1 and 3, wherein the at least one sinusoidal waveform comprises an exponentially decaying sinusoid.

16. The method according to any one of claims 1 and 3, wherein the optimization operation comprises a gradient descent or related based optimization.

17. The method according to any one of claims 1 and 3, wherein the data set is subjected to a signal filtering operation to remove signal deformation.

18. The method of any one of claims 1 and 3, wherein the guess decomposition comprises a sparse sinusoidal decomposition.

19. The method of any one of claims 1 and 3, wherein the guess decomposition further comprises calculating an uncertainty value for the signal set.

20. The method according to any one of claims 1 and 3, wherein the MLTA is adapted to use a machine learning algorithm to generate a result of the comparison.

21. The method of any one of claims 1 and 3, wherein the machine learning algorithm comprises a deep learning algorithm, kernel density estimation, or a calibrated Bayesian network.

22. The method according to any one of claims 1 and 3, wherein the different objects are selected from the group consisting of natural teeth, artificial or replicated teeth, simulated teeth or oral tissue, dental restorations, oral tissue, dental appliances or implants, and dental splints.

23. The method according to any one of claims 1 and 3, further comprising: The main sub-signal is guessed using a machine learning algorithm MLA on the PLM, and the MLA is trained on a data set including a plurality of signals with known main sub-signals and known sinusoidal decomposition.

24. The method of any one of claims 1 or 3, wherein the guessed sinusoidal decomposition is generated by gradient descent or related optimization, Fourier-like operations, machine learning algorithms, or a combination thereof.

25. The method according to any one of claims 1 and 3, wherein the basis function is selected from the group consisting of: a Gaussian curve, an exponential curve, a sinusoidal curve, an exponentially decaying sinusoidal curve, and a quasi-sinusoidal curve.

26. The system of any one of claims 2 and 5, wherein said performing an optimization operation comprises: The difference between the maximum value of the sum of each of the signal sets and zero and each of the signals is minimized to generate an optimized signal set.

27. The system of any one of claims 2 and 5, wherein the at least one sub-signal comprises a waveform from a periodontal ligament (PDL) damping response.

28. The system of any one of claims 2 or 5, wherein the guessed sinusoidal decomposition is generated by gradient descent or related optimization, Fourier-like operations, machine learning algorithms, or a combination thereof.

Images