Apparatus and methods for bone condition change detection to predict intraoperative periprosthetic femoral fractures during total hip arthroplasty

Abstract

Apparatus and methods for preventing intraoperative periprosthetic femoral fracture in total hip arthroplasty (THA), by determining the implant position and femur bone condition changes during the THA by sensing changes in real time. The apparatus operates with a broach or stem, which is inserted into the femur; an auto-impactor; and a diversity of sensors on the interface adapter between the auto-impactor and the broach / stem. The apparatus and methods indicate the quality of implant insertion and enable early detection and warning of approaching the propensity to the intraoperative periprosthetic femoral fracture point by using quantifiable sensor data and trends and feeding real-time information into a decision support algorithm.

Description

[0001] APPARATUS AND METHODS FOR BONE CONDITION CHANGE DETECTION TO PREDICT INTRAOPERATIVE PERIPROSTHETIC FEMORAL FRACTURES DURING TOTAL HIP ARTHROPLASTY

[0002] FIELD OF THE TECHNOLOGY

[0003]

[0001] The present invention is in the field of prevention of intraoperative periprosthetic femoral fractures during total hip arthroplasty (THA) surgery.

[0004] BACKGROUND OF THE TECHNOLOGY

[0005]

[0002] A Total Hip Replacement (THR / THA) offers excellent results in patients of all ages. During total and partial hip replacement, a surgeon implants a femoral stem into the femur. It is very important to be able to estimate the broach / stem position, penetration progress, and the stress condition of the femur during the procedure.

[0006]

[0003] US 2022 / 354594 refers to an Al system to teach good, medium, and bad broach insertion, based on qualitative observations, and reported outcomes.

[0007] SUMMARY

[0008]

[0004] The present invention relates to an apparatus and methods for the sensing of bone condition changes in real-time during broach and stem insertion into the femur during total hip arthroplasty (THA) to indicate the risk level for intraoperative periprosthetic femoral fracture.

[0009]

[0005] The present invention is different from the prior art in that it arises from a deep understanding of microcrack and fracture mechanics as well as dynamics. The invention senses and processes real-time data from dedicated sensors to detect early signs of microcrack generation and propagation that are unobservable qualitatively by the surgeon.

[0006] The invention addresses sensing microcrack generation changes in the femur during the broach / stem impaction insertion process and provides real-time information on bone condition changes after each individual impact: including broach position; fixation level; degree of microcrack generation; and propensity for fracture occurrence. The invention uses several sensors, their signal processing by a processing unit, with apparatus hardware and software, and algorithms, integrated with robotic platforms, and implemented as a wired or as a wireless sensor interface with an auto-impactor.

[0010]

[0007] One of the required sensors is a position sensor to measure the broach and stem insertion into the femur. Another sensor senses the impact force. Other sensors are an acoustic emission sensor; a spectral response sensor; and a sensor measuring torsional stiffness / freedom of the inserted implant, replicating a manual procedure performed by the surgeon during total hip arthroplasty (THA). The method includes a real time Finite Element Analysis (FEA) boundary condition analysis and frequency response modeling of the femur, broach, and auto-impactor as a whole. This modeling is based on diagnostic images of the proximal femur geometry from a hospital’s pre-operative planning system, the implant geometry, and the implant insertion level determined by the position sensor. These are used to assess under specific femur anatomy, the femur dynamic interaction with the broach or stem during the THA and to compare it to the measured frequency response, as an additional tool to assess femur conditions. All of the sensors are deployed non-invasively and are remote from the sterile surgical zone.

[0011]

[0008] In the current THA state of the art, the surgeon depends heavily on his qualitative sensing of audio, visual, and tactile feedback to determine progress and effectiveness of the broach / stem insertion. Audible tone is discerned by the surgeon after each mallet strike to give an indication of implant initial stability and position. This qualitative feedback assessment process has a high degree of variability due to the diversity of patient anatomy and the experience level of the surgeon. The present apparatus / method provides real-time quantitative feedback to assist in a successful and fracture-free THA. The frequency response is determined by the acoustic sensor, including sensitivity beyond the audible range, io determine points of contact and stress to help determine the bone condition.

[0012] BRIEF DESCRIPTION OF THE DRAWINGS

[0013]

[0009] The invention may be more clearly understood upon reading the following detailed description of non-limiting exemplary embodiments thereof, with reference to the accompanying drawings.

[0014]

[0010] Fig. 1 is a side view of a real-time bone condition change monitoring apparatus, in accordance with the presently disclosed subject matter.

[0015]

[0011] Fig. 2 is a side view of a portion of the apparatus, in accordance with the presently disclosed subject matter.

[0016]

[0012] Fig. 3 is another side view of a portion of the apparatus, in accordance with the presently disclosed subject matter.

[0017]

[0013] Figs. 4A-4C are graphs of a quasi-static load-displacement curve of a broach of the apparatus inserted into the femur, after removing the head of the femur (osteotomy), in accordance with the presently disclosed subject matter.

[0018]

[0014] Fig. 5 is a graph of frequency response after impacts into a sheep bone, in accordance with the presently disclosed subject matter.

[0019]

[0015] Fig. 6 is a graph of an acoustic emission event outliers versus penetration into the bone, in accordance with the presently disclosed subject matter.

[0020]

[0016] Figs. 6A-6D are graphs displaying various results indicating the bone condition versus penetration, in accordance with the presently disclosed subject matter.

[0021]

[0017] Fig. 7A is a graph showing impact force versus penetration, in accordance with the presently disclosed subject matter.

[0018] Fig. 7B is a graph showing the broach-bone frequency response and the stabilization of the responses at the proper insertion position, in accordance with the presently disclosed subject matter.

[0022]

[0019] Fig. 8 is an algorithmic flow chart of an example of a methodology of the insertion, in accordance with the presently disclosed subject matter.

[0023]

[0020] Fig. 9A and 9B show additional examples of the apparatus, in accordance with the presently disclosed subject matter.

[0024] DETAILED DESCRIPTION

[0025]

[0021] The following detailed description of examples of the invention refers to the accompanying drawings. Dimensions of components and features shown in the figures are chosen for convenience or clarity of presentation and are not necessarily shown to scale. Wherever possible, the same reference numbers will be used throughout the drawings and the following description to refer to the same and like parts. In the interest of clarity, not all features / components of an actual implementation are necessarily described.

[0026]

[0022] The present subject matter relates to examples of an apparatus and methods for determining broach / stem propagation in the femur during each impact on the broach interface by the auto-impactor, while monitoring femur stress level and the generation of microcracks during the broach / stem insertion. For each example, a sensing solution is presented.

[0027]

[0023] Fig. 1 relates to one example of an apparatus to monitor broach insertion into a femur 10. The apparatus is used in conjunction with a broach 20 or stem for femoral insertion, an interface adapter 30, an auto-impactor 40; and two sensors, an accelerometer sensor 71 on the interface adapter of the autoimpactor, and a distal femur accelerometer sensor 72 attached on the leg / knee area. Double signal integration and subtraction of the two accelerometers yields the amount of insertion and compensates for any movement of the patient. Vibrational and velocity information are also available from the accelerometer sensors 71 , 72, via their original acceleration signals and their single integration and double integration (the velocity and propagation respectively). In some cases where the leg is fixed, the interface adapter 30 accelerometer sensor 71 can suffice.

[0028]

[0024] Interface adapter accelerometer sensor 71 is attached directly to the auto-impactor interface adapter 30 between the broach or stem 20 and the auto-impactor 40. Alternatively it can be within the internal impacting part (moving mass) of the auto-impactor 40. There are other methods to measure position, by other types of position sensors - resistive, capacitive, optical and more as known in the art. They are all applicable to measure the insertion of the broach 20 into the femur 10 in the invention.

[0029]

[0025] Fig. 2 illustrates using an acoustic piezo transducer / sensor 73. The interface between the broach / stem 20 and the femur 10 according to the osteotomy area in the femur and the level of insertion, will yield a change in the spectral response due to the load change of the broach after each impact. It is further explained in Fig. 5. The acoustic piezo transducer / sensor 73 can also detect the first propagating shockwave resulting from impact on the broach / stem 20; and afterwards, a smaller reflection is detected from a boundary 11 (osteotomy cutting area) between the broach / stem (orthopedic implant) and the femur. The timing of these reflections, and the change in timing between impacts, indicates the amount of insertion according to the time difference multiplied by the speed of sound in the broach or stem 20.

[0030]

[0026] Fig. 3 shows an example of a method in accordance with the subject matter which relates to sensing torsional stiffness or mechanical freedom. Such a method is performed today manually by the orthopedic surgeon to assess fixation stability of the broach / stem 20 in the femur 10. In the current state of the art, the surgeon manually attempts to rotate the broach / stem to qualitatively assess if mechanical freedom is present, or if the implant is secured. The torsional motion could be realized as part of the auto-impactor function. After each or several impacts of the auto-impactor 40, a slight torsional torque is applied by a rotary actuator, and the residual rotation of the broach / stem 20 is measured by an angular sensor.

[0031]

[0027] Fig. 3 shows an example of the apparatus where the auto-impactor 40 includes a torsional torque generating actuator 80 that automatically generates a minute torsional torque. The actuator is a torque actuator 80 inside the autoimpactor 40 producing a rotational movement after impacts, symmetrical around the impact axis.

[0032]

[0028] While actuator 80 is activated, the torsional stiffness or freedom assessment is detected via one of several types of angle sensors 81 , as known in the art, for example a magnetic Hall effect sensor.

[0033]

[0029] The apparatus uses acoustic transducer / sensor 73 to detect frequency response and / or to generate a controlled excitation sweep. The acoustic transducer / sensor 73 can be of many types to generate vibration in audible and ultrasonic (US) frequencies. Most common are piezo-ceramic disks but also a polymer acoustic transducer is applicable, such as a PVDF polymer transducer.

[0034]

[0030] Piezo transducers are usually thin PZT discs. They are brittle and can break from the impacts, especially if the piezo disc is large. One example uses an electret polymer with piezo properties, such as Polyvinylidene Fluoride (PVDF). The benefits are no risk of breakage and a large sensor area in which it is possible to optimize dimensions (increase) for better matching to a lower frequency range. The maximal response to surface waves will be when the dimension of the sensor is equivalent to half of the wavelength. PVDF is flexible and can be extended in dimensions, without the PVDF breaking, to increase sensitivity and matching to the surface waves frequencies.

[0031] An important aspect of the invention is the capability to detect intraoperative fracture risk in real-time regardless of the implant type or patient specific bone characteristics. This is accomplished by detecting pre-defined trends of bone condition change during implant insertion to determine impaction control criteria for all THA cases, rather than by detecting predefined fixed values, which are bone specific. Further explanation is as follows:

[0035]

[0032] A major challenge in the detection of femoral microcrack trends during THA is the varying characteristics of patients’ femurs, with different bone mass density, shape, geometry, and size of the medullary canal. The design of broaches and stems 20 for cementless THA varies between implant product lines from different brands. The broach designs support different methods of femoral cavity preparation, including compaction, blunt extraction, and sharp extraction broaches. Different broach and stem 20 designs, exhibit different friction coefficients during insertion, which has a different effect on intraoperative periprosthetic fracture risk. A key aspect of the invention is to follow pre-defined trends of microcrack development and propagation as understood from material science and fracture theory. This trend is observed in the femur 10 insertion process, during three phases of the broach / stem 20 implant insertion process.

[0036]

[0033] Phase 1 of the insertion is of asperity removal, low stress (minimal friction). Phase 2 of the insertion is of microcrack generation, increasing stress (friction). Phase 3 of the insertion is of higher microcrack density, higher stress (higher friction).

[0037]

[0034] The actual value of bone stress level during the implant insertion process varies due to specific patient femur characteristics and the type of implant used. Hence, a predefined absolute value of implant progression cannot be used to reliably predict when it is safe to continue impacting the implant or when to stop impaction to avoid a fracture. Monitoring progression through the three phases, in combination with real-time analysis of other sensor signals, allows impaction decisions to be effective for a broad range of individual sets of circumstances. No matter the diversity and combination of factors, each implant insertion process follows the same microcrack generation pattern. Regardless of the implant design or the patient’s bone mass density, or shape, geometry, and size of the medullary canal, the analysis of the three phases indicates periprosthetic fracture risk and when initial implant stability is achieved. The apparatus can thus decide on the appropriate and safe implant fixation for a broad range of a femur’s mechanical and material properties. The femur 10 in effect becomes its own sensor, following a known trend, and exhibits varying feedback characteristics in the three phases during the insertion impaction process. The sensor feedback during the progression through those three phases yields a trend to indicate when to stop impaction, when fixation is achieved and before an intraoperative femoral fracture is likely to occur. This methodology was verified using a combination of sensing methods A, B, and C, as follows:

[0038]

[0035] Method A generates a P-Delta load-displacement curve using signals from the insertion position sensor, implemented by an accelerometer as known in the art, 71 (or 71 and 72) or other insertion (position) sensors, and the impact force sensor 41 , and then monitoring the slopes of the P-Delta load-displacement curves throughout the broach and stem insertion process.

[0039]

[0036] Fig. 4A shows a quasi-static load-displacement curve of the broach 20 inserted into the femur 10 after removing the head of the femur 10 (Osteotomy) - femoral head osteotomy (FHO). Initially the broach 20 removes the bone asperities in the femur medullary canal, indicated by the low slope of the P-Delta load-displacement curve, representing the low friction phase 1. Starting with the phase 2, the broach 20 is in contact with the femur’s hard cortical bone to prepare the femur 10 by broaching into the cortical surface. Higher friction is indicated by an increasing slope of the P-Delta load-displacement curve. This phase 2 during the femur 10 preparation process also represents a stage when microcrack generation occurs. The broach insertion process then progresses to phase 3 that is indicated by a much steeper slope of the P-Delta load-displacement curve. During phase 3 the slope increase indicates that the fixation to the bone is tighter and microcrack density has increased. During phase 2, the broach will be in contact with the bone surface in the proximal metaphysis area of the femur where the femoral head osteotomy was cut to prepare the medullary canal for the broach and stem insertion. This contact creates a concentration of stress which promotes microcrack initiation in that area. In phase 3 of the broach insertion process, the microcrack generation rate is accelerated, increasing the microcrack density and risk of fracture. Note that the relatively high P-Delta slope accompanied by low strain energy, Ue, in the apparatus are strong indicators prior to a fracture.

[0040]

[0037] Fig. 4B shows a significant change in a second example, in that the osteotomy of the femoral neck is lower in the proximal metaphysis area, changing the contact location between the broach and the bone. Like in the first example, the P-Delta load-displacement curve’s slope is still low during the initial low friction asperities removal of phase 1 . However, the curve then has a slightly lower slope. Thereafter, the slope increases again, due to higher friction with the cortical bone area of the femur. At the peak load, the fracture initiates at a deeper part of the femur 10 than the level of the osteotomy and propagates in a pseudo-ductile behavior as seen from the significantly higher energy, Ue +Up, compared with the first example.

[0041]

[0038] Fig. 4C shows a third load-displacement curve, where a small broach was inserted in a natural sheep femur, showing a similar two slope behavior, with a limited quasi-ductile fracture at the peak load. Note the similar behavior of the natural and synthetic bone load-displacement curves. A natural sheep bone is much weaker than the synthetic human bone that was tested. Results are shown for a synthetic human femur and a sheep femur.

[0042]

[0039] The P-Delta method shows two modes of fracture, depending on the location of the osteotomy. When the osteotomy is made in a higher location, the broach 20 is most likely in contact along the free surface exposed by the cut, applying high local stress concentration. In addition to being thin, the bone in this area is more prone to microcrack and fracture initiation.

[0040] On the other hand, when the osteotomy is made in a lower location, the broach 20 is in contact with the thicker inner area of the femur 10, while the exposed osteotomy surface is relatively stress-free. A fracture could initiate due to natural defects of the inner surface of the medullary canal bone, and progress stably at increased deformation, to generate plastic-like high energy absorption mechanisms. The pre-defined bone behavior trend applies to both.

[0043]

[0041] Monitoring the insertion of the broach / stem 20 and sensing the impact force are required to dynamically derive the P-Delta or load-displacement curve. While the sensors 71 (or 71 and 72) will determine the propagation of the broach / stem 20 with each impact, another dynamic impact force sensor 41 (Fig. 1 and 3) is needed to measure the dynamic impact force of the auto-impactor 40. The force sensor 41 can be located between the auto-impactor 40 and the broach / stem 20 or internally within the auto-impactor 40, measuring the impact force applied by the auto-impactor. Sensors, like but not limited to, load cell, piezo load cell or impedance head, known in the art, can be used for the force sensing.

[0044]

[0042] Another feature related to the P-Delta load-displacement curve is where the auto-impactor 40 and force sensor 41 are bi-directional, the autoimpactor is capable to hit in two opposing directions, insertion and extraction; the measured force is also in both directions to allow (1 ) replicating the surgical process; and (2) assessment of residual stress when impacting occurs in the opposite direction-and movement of the broach / stem 20 is outwards from the femur 10, yielding hysteresis.

[0045]

[0043] Fig. 5 shows a major feature of the acoustic transducer, namely determining the stabilization of the frequency response of the bone-broach as an indication of the fixation stability level. Each impact yields a deeper insertion. There are multiple impacts and correspondingly multiple insertion steps. In Fig. 5 the frequency response is shown for clarity only for a sub-set of the steps. Step 1 is the position after the first impact, step 2 is the position after the second impact and so on with the consecutive insertion steps. One can see a stabilization of the frequency response in Fig. 5 as the implant is firmly held in the femur as seen in the frequency response of steps 10 and 11 of the corresponding example.

[0046]

[0044] Method B, as shown in Fig 5, refers to a frequency response measured by the acoustic transducers / sensors 73 (shown in Fig 2). Two frequency responses can be measured: (1 ) a Fourier Transform of the impact as sensed by the acoustic transducer / sensor; and (2) use a pair of acoustic transducers (sensors) 73, one generating a sweep frequency and the second detecting the bone-implant-impactor response to the sweep. The two responses may differ due to different boundary conditions. Stabilization of the frequency response shows fixation of the broach / stem 20 in the femur 10. Measurement of this stabilization showing fixation of the femur 10 and the broach / stem 20 during the sweep can be executed between impacts after the post-impact transient effects have settled down, using method 2 above. Such a measurement may take a few seconds to achieve high resolution and hence it is reasonable to execute it every few impacts. Fig. 5 shows a frequency spectrum after each impact for a sheep bone. One sees the stabilization of the frequency spectrum and the corresponding insertion. The frequency spectrum helps overcome the potential ambiguity of P-Delta due to non-linearities as shown in Fig, 7A and Fig. 7B.

[0047]

[0045] Method C, as shown in Fig. 6A-6D and Fig 9A, shows that the acoustic emission (AE) sensor 51 detects a multitude of signals. AE signals are produced during all phases of the implant insertion by asperities removal, the friction between surfaces of the femur and broach, and microcrack generation. When the insertion has entered phase 3 (higher stress / friction), the microcrack signal intensity increases as more energy is released by the microcracks in this phase. To determine if the femur 10 is approaching a critical fracture point, the AE outlier signals in the intensity distribution are focused-on, determining the propensity for fracture and confirming the phase the femur 10 is in.

[0048]

[0046] To determine what an outlier signal is for a specific bone, a histogram of events intensity is captured in the early stages of insertion, which sets the threshold on the tail of the intensity distribution, to determine which intensity level would be defined as an outlier signal. The threshold level is indicated in Fig. 6A and Fig. 6C. The threshold levels are different and were set per bone as explained above. These outlier signals will then indicate the severity level of microcracks, as explained above. Method C observes the outlier signal trends in number of events and energy levels and monitors their occurrence within the 3 phases explained above.

[0049]

[0047] Figs. 6A-6D show the different intensities of the acoustic emission (AE) signals in three insertion phases: (phase 1 ) initial low friction and removal of asperities; (phase 2) generation of microcracks; and (phase 3) increasing microcrack density and fracture risk propensity. These phases differ in energy and rate of occurrence as shown in Figs. 6A-6D. Fig. 6A shows the number of outlier signal events (acoustic emission signals above a determined threshold of 25) in a sheep bone experiment penetration cycle of load and unload. Fig 6B shows the total energy of outlier signals in a natural sheep bone per penetration cycle. Fig. 6C shows the number of outlier signal events (acoustic emission signals above a determined threshold of 31 ) in a synthetic human bone per penetration cycle. Fig. 6D shows the total energy of outlier signals in a synthetic human bone per penetration cycle. The significant increase in slope is an indication of approaching the fracture risk. The AE signal can also be analyzed in the frequency domain by taking the Fourier T ransform of the outlier signal and seeing a growing amount of high frequency spectrum as the bone approaches fracture.

[0050]

[0048] Figs. 7A and Fig. 7B show information by methods A and B. Fig. 7A shows the sensing of the P-Delta curve. Fig. 7B shows the broach-bone frequency response and the stabilization of the responses at the proper insertion position, with good primary stability.

[0051]

[0049] Fig. 7A shows a sheep bone P-Delta curve in 16 consecutive insertion steps, in load and unload cycles represented by the vertical lines, each resulting from an impact. The natural bone is nonhomogeneous, and the slope is non-linear. The P-Delta curve yields two candidate implant insertion stop points where the slope increases (insertion step 7 and insertion step 13). The way to differentiate them is shown in Fig 7B. The frequency response of the bone-- broach / stem is measured. Fig 7B shows a subset of the relevant insertion steps, it is noted that the frequency is stabilized starting from spectral curves related to steps 13-14. Insertions 6 and 7 show a P- Delta slope increase but no stabilization of the spectral response. This is a combined criterion indicating that the broach / stem 20 is firmly fixed to the bone. Figs. 7A and 7B confirm the broach / stem insertion step that has sufficient fixation and stability, with no risk of fracture.

[0052]

[0050] As an equivalent indication to P-Delta and the curve of method A, the invention can alternatively analyze the relationship between impact energy / force and incremental penetration. This is achieved by examining the rate of change in penetration, as the broach / stem advances, and by measuring the dynamic impact force that was applied. A marked decrease in the penetration rate, below a threshold, signifies that the broach has reached a steady state stage of implantation, and no further impacts of the same force are needed. The decision whether to increase the force or stop depends on the additional sensors. The insertion level as a function of energy / force is then used in the same manner explained for P-Delta and cross-referenced with the spectral analysis data, providing a comprehensive and robust indication of the optimum broaching insertion and sufficient fixation under impacts. This procedure also offers criteria on when to change a broach to a larger size. If a broach is inserted to its full length but the fixation criteria as described in methods A, B, and C are not fulfilled a bigger broach should be used.

[0053]

[0051] The three methods (A, B, and C) accompanied by the measured insertion per impact, are the criteria used to determine when to stop impacting the implant on both synthetic human bones, natural sheep bones, and any femur bones. The criteria are according to moving between the three phases as understood in Figs. 4, 5, and 6. The trend of the P-Delta; frequency response; and AE outlier signals, always follow the same pattern although the values will change depending on the bone’s mechanical and material properties, and the type of implant used. The fracture prediction algorithm result is therefore unique and specific to the condition changes of each individual bone even under broad diversity in properties of the bone. The bone in effect becomes its own sensor, following a known trend pattern and its behavior is monitored by the three methods A, B, and C.

[0054]

[0052] The signals of sensors 41 , 51 , 52, 73, 71 and 72, are processed by the processing unit 60 (Fig. 8) and analyzed by HW and SW and algorithms of the apparatus, per sensor and per the multiplicity of sensors listed above. The algorithms are explicit and are based on fracture mechanics and dynamic modeling. The algorithm can be further enhanced by unsupervised Al for part of, or the entire sensor array. Recording sensor data from prior THA surgeries and applying Al enhances the prediction of fracture risk levels. In the present apparatus, Al can be used to further improve the impact decision process.

[0055]

[0053] Fig. 8 shows an algorithm flow chart of the insertion method and feedback from the sensors. After each impact of the impactor 40 on the implant (broach / stem) 20, inserting it into the femur 10, the position sensors (such as by accelerometer 71 or both 71 and 72); force sensor 41 ; frequency response transducer / sensor 73; and AE sensor 51 (or 51 and 52), are analyzed by the processing unit 60. First the position and force readings are used to generate an additional P-Delta data point. The P-Delta curve is analyzed by the processing unit 60 to derive the slope change of the P-Delta curve 31 . If the P-Delta slope did not increase, the processing unit 60 checks sensor 71 and optionally 72 to determine the rate of insertion. If the rate of insertion decreased, the impact force is increased, if the rate of insertion did not increase, the impact force used remains unchanged.

[0056]

[0054] If the slope of the P-Delta curve 31 increases above a set %, the processing unit 60 checks the frequency response of sensor 73 for change. If the frequency response is still changing and the AE sensor 51 signal is checked and does not show an apparent increase in the amount of outlier signals, the impaction continues. If the outlier signals of the AE sensor 51 , checked for rate of occurrence, indicate that the slope of the amount of outlier signals versus impact increases above a certain set %, the processing unit 60 stops impaction. If the frequency response is not changing, the impaction is stopped, as the implant has reached a good fixation.

[0057]

[0055] Fig. 8 also shows another condition in the algorithm pertaining to an implant change. If the implant 20 is fully inserted and stabilization criteria according to methods A-C, are not achieved, a bigger implant should be used.

[0058]

[0056] AE sensor signal extraction provisions of the invention are described as follows: Some of the signals, such as the AE of method C, are measured during the impact, as microcracks are generated shortly thereafter. The provisions below improve the detection of microcrack signals that can be masked by the much more intense impact signals.

[0059]

[0057] Provision # 1 - The microcrack duration is shorter than the impact pulse. Thus, the AE sensors 51 , 52 are set to respond to higher frequencies (inversely relating to the geometrical dimensions of the AE sensor, typically above 300kHz, and reduced in sensitivity to avoid sensor saturation.

[0060]

[0058] Provision # 2 - Apply a HPF (High Pass Filter) to the amplifier of the AE signal as the microcrack signal is shorter in duration than the impact response and the HPF will attenuate more of the impact than the microcrack signal. This attenuates the impact signal as much as possible and increases the microcrack-to-impaction signal ratio.

[0061]

[0059] Provision #3 - Event trigger for high frequency sampling. An acquisition trigger is applied only upon the first pass of the signal threshold level, to start acquiring the data of the sensors 41 , 51 (Fig. 9A), 52 (Fig. 9B), 71 , 72 and 73. This will confine the amount of data acquired at a high sampling rate for rapid processing.

[0062]

[0060] Provision #4 - Based on the signal propagation time from an impact point 32 on the broach 20 and the distance between the femur 10 and the acoustic emission sensors 51 , a deadtime window is set to segregate between the main impact and microcrack signals, as understood from Fig. 9A where the time of arrival of the shock wave, resulting from the impact, depends on the location of the sensors.

[0063]

[0061] Fig. 9A shows how the timing sequence can be understood, where the distances D1 and D2 are converted to timing according to the speed of sound (c) in the elements:

[0064]

[0062] - An initial impact signal is detected at time D1 / c after the impaction, where D1 is the distance between the impact point 32 and the sensor 51 . The computed deadtime is set as the propagation time between the impact point 32 and the sensor 51 . The process starts with an indication signal from the autoimpactor 40 that an impact is generated.

[0065]

[0063] - Initial microcrack signals occur at time 2xD2 / c after the first event, where D2 is the distance between the AE sensor and the proximal metaphysis zone 12 of the femur 10 where microcracks will most likely develop,

[0066]

[0064] - After these time intervals between the initial impact signal and the initial microcrack signal, there are reflections of the primary impact, which do not contribute to the analysis.

[0067]

[0065] Numerical example: D1 = 1 .0 cm, D2 = 1 .5 cm, and the speed of sound, c = 5,000 m / sec. The impact signal arrives after 2 microseconds (D1 / c) from an indication from the impactor 40 of a forthcoming impact; the microcracks signals arrive after an additional 6 microseconds (2xD2 / c); aftershocks occur a few tens of a microsecond later. A deadtime of 3-4 microseconds is set to eliminate the signal of the main impact. This timing is longer than the 2 microseconds mentioned above and shorter than the 6 microseconds when the microcracks signals begin to appear.

[0068]

[0066] Provision #5 - accumulate typical microcrack signals in the three phases of the acoustic emission signals, under static loading (no impact) to allow devetoping a matched filter and a robust pattern recognition of typical microcrack signals when under impact.

[0069]

[0067] Provision # 6 - the two AE sensors 51 , 52, are used, as shown in Fig. 9B, symmetrical with respect to the propagation time from the impact surface 32 but not symmetrical in distance to the bone. A differential signal of the two AE sensors 51 , 52 will mostly cancel the impact signal as it appears simultaneously on both sensors due to the equal distances from the impaction surface 32 in Fig. 9B, and will leave only the microcrack’s emission signal from the bone. There will be two time zones of detection of the microcracks in the differential signal as the propagation from the bone microcracks to the sensors have different distances.

[0070]

[0068] The position of AE sensors 51 , 52 with respect to the impact surface is shown in Fig. 9B. The microcrack signals are not symmetrical to the acoustic emission sensors 51 , 52 in respect to propagation time while the impact is symmetrical.

[0071]

[0069] By the above provisions, the attenuation of the main impact signal has been maximized in favor of the microcrack signals. Matched filters will be applied in the time domain or Fourier domain to extract the typical microcrack signals to build a histogram of the three phases.

[0072]

[0070] Provision # 7 - operating in the frequency domain can alleviate some of the detection challenges and enable monitor trends via the AE spectrum.

[0073]

[0071] Integration to Robotic Platforms: All sensing and analysis methods use hardware and software, and algorithms, and can be implemented on robotic platforms (not shown).

[0074]

[0072] The invention algorithms for sensing and analyzing can be integrated within the robotic operating platforms and application software of a robotic platform (not shown) and a robotic arm (not shown).

[0073] Hardware for the auto-impactor 40 can be integrated as part of a robotic platform and robot arm (not shown).

[0075]

[0074] Use of Al: In the invention, unsupervised Al is used as follows:

[0076]

[0075] 1 . To supplement decision analysis of signals from the sensors 41 ,

[0077] 51 , 52, 71 , 72 and 73 as explained in methods A, B, and C.

[0078]

[0076] 2. Unsupervised recording signals of the sensors from THA operations and cadaver bone experiments with intentional fractures, where Al is used to detect extremely minute deviations and provide supporting input to the algorithm on fracture risk.

[0079]

[0077] Another embodiment is where the apparatus is wireless with a battery powered auto-impactor 40 and sensors 41 , 51 , 52, 71 , 72 and 73, and communication among components via Wi-Fi or other protocols that are approved in operating rooms.

[0080]

[0078] The above description is merely exemplary and various embodiments of the present invention may be devised, mutatis mutandis, and that the features described in the above-described embodiments, and those not described herein, may be used separately or in any suitable combination; and the invention can be devised in accordance with embodiments not necessarily described above.

Claims

CLAIMS1 . A real-time bone condition change monitoring apparatus for use with an auto-impactor that is operable to insert an orthopedic implant into a femur during a total hip arthroplasty (THA) procedure, the monitoring apparatus comprising: sensors to generate an output in real-time in response to progressive change in femur condition during the femoral preparation in a THA procedure from an asperities-removal phase to a microcrack-generation phase and further to an increasing-microcrack-density-and-fracture-risk-propensity phase; and a processing unit to perform signal analysis upon the output of the sensors to monitor the change in the femur condition.

2. The apparatus of claim 1 , wherein the signal analysis performed by the processing unit includes P-Delta analysis.

3. The apparatus of claim 2, wherein the sensors include a position sensor to measure insertion progress of the orthopedic implant into the femur, and a dynamic impact force sensor to measure the impact force applied by the auto-impactor while the orthopedic implant is being driven into the femur, and wherein the P-Delta analysis is based on the output of the position sensor and the dynamic force impact sensor.

4. The apparatus of claim 3, wherein the P-Delta analysis involves monitoring the slope of a load-displacement curve constructed from the output of the position sensor and the dynamic force impact sensor.

5. The apparatus of claim 3, wherein the position sensor includes a first accelerometer to measure acceleration of the orthopedic implant.

6. The apparatus of claim 5, wherein the position sensor further includes a second accelerometer sensor to measure acceleration at the distal femur area near the knee joint, to account for any body movement, the processing unit performing signal analysis on the output of the first accelerometer sensor and the second accelerometer sensor to determine insertion depth of the orthopedic implant into the femur.

7. The apparatus of claim 3, wherein the position sensor includes an acoustic transducer to detect a shockwave resulting from an impact on the implant, and a reflection wave at a boundary between the orthopedic implant and the femur, the processing unit performing signal analysis on the output of the acoustic transducer to determine insertion depth of the orthopedic implant into the femur.

8. The apparatus of claim 7, wherein the processing unit performs signal analysis on the output of the acoustic transducer to further determine fixation stability level of the orthopedic implant relative to the femur.

9. The apparatus of claim 1 , wherein the signal analysis performed by the processing unit includes spectral response analysis.

10. The apparatus of claim 9, wherein the sensors include an acoustic transducer to detect spectral response due to load change of the orthopedic implant on the femur while the orthopedic implant is being driven into the femur by the auto-impactor, wherein the spectral response analysis is based on the output of the acoustic transducer.11 . The apparatus of claim 1 , wherein the sensors include a first acoustic transducer to transmit a sweep frequency between each impact; and a second acoustic transducer to detect femur-implant-impactor combined response to the sweep frequency, the processing unit performing signal analysis on the output of the second transducer to determine fixation stability level of the orthopedic implant relative to the femur.

12. The apparatus of claim 11 , wherein the processing unit determines if the orthopedic implant is fixed in the femur when a frequency response of the output of the second acoustic transducer stabilizes.

13. The apparatus of claim 1 , wherein the sensors include an acoustic emission sensor, the output of which is indicative of acoustic waves emitted from microcracks and stress within the femur.

14. The apparatus of claim 13, wherein the processing unit performs acoustic emission signal analysis on the output of the acoustic emission sensor to determine if the femur is approaching a critical fracture point.

15. The apparatus of claim 13, wherein the processing unit performs a Fourier Transform on outlier signals of the output of the acoustic emission sensor to determine if the femur is approaching a critical fracture point.

16. The apparatus of claim 13, wherein the processing unit further determines the femur condition phase with respect to fracture risk based on acoustic emission signal analysis.

17. The apparatus of claim 13, wherein the processing unit processes the output of the acoustic emission sensor including high pass filtering to increase the microcrack-to-impaction signal ratio.

18. The apparatus of claim 13, wherein the acoustic emission sensor includes a first acoustic emission sensor within the auto-impactor and a second acoustic emission sensor at the interface of the auto-impactor and the orthopedic implant, the first acoustic emission sensor and the second acoustic emission sensor being symmetrical with respect to impact signal propagation time and are asymmetrical with respect to the femur microcrack signal propagation time to the first acoustic emission sensor and the second acoustic emission sensor.

19. The apparatus of claim 1 , wherein the sensors include an acoustic transducer sensor, and the processing unit performs frequency responseanalysis on the output of the acoustic emission sensor to provide an indication of approaching fracture state of the femur.

20. The apparatus of claim 1 , wherein the processing unit includes an unsupervised artificial intelligence system to analyze the output of the sensors from prior THA procedures and to learn from the recorded output to enhance prediction of fracture risk levels.21 . The apparatus of claim 1 , wherein the sensors are wireless sensors.

22. An auto-impactor operable to insert an orthopedic implant to a femur during a total hip arthroplasty (THA) procedure, the auto-impactor comprising:Sensors to generate an output in real-time in response to progressive change in femur condition during the THA procedure from an asperities- removal phase to a microcrack-generation phase and further to an increasing- microcrack-density-and-fracture-risk-propensity phase; and a processing unit to perform signal analysis upon the output of the sensors to monitor the change in the femur condition, the processing unit providing closed-loop control feedback to control frequency and energy of impacts of the auto-impactor based on monitored changes in the femur condition.

23. The auto-impactor of claim 22, further comprising a torsional torque generating actuator to apply a torsional torque onto the orthopedic implant after impacts to determine torsional stiffness or freedom of the orthopedic implant, wherein the sensors further include a torsional stiffness sensor to measure rotational movement of the orthopedic implant resulting from the applied torsional torque.

24. The auto-impactor of claim 22, wherein the processing unit increases the impact force of the auto-impactor when, based on the output of the sensors, the rate of change in penetration depth of the orthopedic implant due to application of the impact force decreases.

25. A robotic platform comprising: a robot arm; an auto-impactor assembled on the robot arm to insert an orthopedic implant into a femur during a total hip arthroplasty (THA) procedure; sensors to generate an output in real-time in response to progressive change in femur condition during the THA procedure from an asperities- removal phase to a microcrack-generation phase and further to an increasing- microcrack-density-and-fracture-risk-propensity phase; and a robotic platform processing unit with integrated apparatus hardware, software, and algorithms, to perform signal analysis upon the output of the sensors to monitor the change in the femur condition, the processing unit providing closed-loop control feedback to control frequency and energy of impacts of the auto-impactor based on the monitored change in the femur condition.

26. A real-time change monitoring apparatus for use with an auto-impactor that is operable to insert an orthopedic implant into a femur during a total hip arthroplasty (THA) procedure, the monitoring apparatus comprising: sensors to generate an output in real-time in response to change in bone condition relative to the femur during the THA procedure from an implant-insert phase to an implant-stop phase; and a processing unit to perform signal analysis upon the output of the sensors to monitor the change in the bone condition.

27. A method for real-time bone condition change detection, the method comprising: sensing the response to an impact on a broach or a stem during a THA to predict propensity to a femoral fracture using an impact force sensor; the sensing comprising using an acceleration sensor for measuring implant insertion into a femur; an acoustic transducer to detect a frequency response and / or to generate a controlled excitation sweep; and an acoustic emission sensor for detecting the generation of microcracks in a femur.

Images