NONINVASIVE DIAGNOSTIC SYSTEM

Abstract

A device for acquiring data and diagnosing a musculoskeletal injury. The device includes a semi-flexible housing, at least one ultrasonic transducer, a positional localizer, and a transmission system. The semi-flexible housing is positioned proximate a portion of the musculoskeletal system of a patient and supports the at least one ultrasonic transducer and the positional localizer. The at least one ultrasonic transducer is configured to acquire an ultrasonic data indicative of a bone surface. The positional localizer is positioned at a select location relative to the at least one ultrasonic transducer and tracks movement of the housing. The transmission system transmits the ultrasonic data of the at least one ultrasonic transducer and the movement data of the positional localizer to a data analyzer for analysis and diagnosis.

Description

RELATED APPLICATIONS

The present application claims the filing benefit of co-pending PCT Patent Application No. PCT/US2010/022939, filed on Feb. 2, 2010, and is a Continuation-In-Part of co-pending U.S. patent application Ser. No. 12/364,267, filed on Feb. 2, 2009, the disclosures of both applications are hereby incorporated by reference herein in their entirety.

FIELD OF THE INVENTION

The present invention relates to devices and methods for evaluating a physiological condition of a musculoskeletal system and, more particularly, to evaluating the physiological condition of bodily joints.

BACKGROUND OF THE INVENTION

In humans, the knee joint 50, as shown in FIGS. 1, 2, and 3, is functionally controlled by a mechanical system governed by three unique types of forces: (1) active forces resulting from motion, such as those resulting from a muscle flexing or relaxing; (2) constraining forces that constrain motion, such as those resulting from ligaments being in tension; and (3) interaction forces that resist motion, such as those acting upon bones. In addition to these three types of forces, the soft tissue in the knee joint 50 (e.g., cartilage and the meniscus) produce a dampening effect distributing the compressive loads acting on the knee joint 50.

Knee joint motions are stabilized primarily by five ligaments which restrict and regulate the relative motion between the femur 52, the tibia 54, and the patella 56. These ligaments are the anterior cruciate ligament (“ACL”) 58, the posterior cruciate ligament (“PCL”) 60, the medial collateral ligament (“MCL”) 62, the lateral collateral ligament (“LCL”) 64, and the patellar ligament 66. An injury to any one of these ligaments 58-66 or other soft-tissue structures may cause detectable changes in knee kinematics and the creation of detectable vibrations, each of which may be representative of the type of knee joint injury and/or the severity of the injury. These visual (knee kinematics) and auditory (vibrations) changes are produced as the bones 52, 54, 56 move in a distorted kinematic pattern and differ significantly from the look and sound of a properly balanced knee joint 50 moving through the same range and types of motion.

Conventionally, knee vibration has been detected using microphones with or without stethoscope equipment and correlated with clinical data regarding various joint problems. However, microphones and stethoscopes cannot reliably detect frequencies, especially those experiencing strong interference from noise. Also the signal clearance can be substantially be influenced by skin friction. It is desirable, therefore, to provide a diagnostic tool that compares patient specific data with kinematic data while providing visual feedback to clinicians.

SUMMARY OF THE INVENTION

While the present invention will be described in connection with certain embodiments, it will be understood that the present invention is not limited to these embodiments. To the contrary, this invention includes all alternatives, modifications, and equivalents as may be included within the spirit and scope of the present invention.

A device for acquiring data and diagnosing a musculoskeletal injury in accordance with one embodiment of the present invention includes a semi-flexible housing, at least one ultrasonic transducer, a positional localizer, and a transmission system. The semi-flexible housing is positioned proximate a portion of the musculoskeletal system of a patient and supports the at least one ultrasonic transducer and the positional localizer. The at least one ultrasonic transducer is configured to acquire an ultrasonic data indicative of a bone surface. The positional localizer is positioned at a select location relative to the at least one ultrasonic transducer and tracks movement of the housing. The transmission system transmits the ultrasonic data of the at least one ultrasonic transducer and the movement data of the positional localizer to a data analyzer for analysis and diagnosis.

Another embodiment of the present invention is directed to a method of diagnosing a musculoskeletal injury. The method includes creates a 3D model of a portion of the musculoskeletal system of a patient. A feature data is acquires by a sensor that is positioned proximate the portion of the musculoskeletal injury. The feature data is compared, by a neural network, to a database of feature data. A dataset within the database of feature data is representative of the musculoskeletal injury. Then, based on the comparing, a diagnosis is returned.

Still another embodiment of the present invention is directed to a diagnostic system for diagnosing a musculoskeletal injury. The system includes a 3D model reconstruction module that acquires a structural data indicative of a bone surface. The bone is within a portion of the musculoskeletal system of a patient. The 3D model reconstruction module constructs a patient-specific model from the structural data. The system further includes a kinematic tracking module that acquires movement data while the portion of the musculoskeletal system is articulated. A vibroarthography model acquires vibration data generated by the articulation. The structural data, the movement data, and the vibration data are received and analyzed by an intelligent diagnosis module in order to determine injury type.

The above and other objects and advantages of the present invention shall be made apparent from the accompanying drawings and the description thereof.

BRIEF DESCRIPTION OF THE DRAWINGS

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the present invention and, together with a general description of the invention given above, and the detailed description of the embodiments given below, serve to explain the principles of the present invention.

FIG. 1 is a side elevational view of a posterior portion of a knee joint with a 90° flexion.

FIG. 2 is a side elevational view of the knee joint of FIG. 1 but with the knee joint fully extended.

FIG. 3 is a side elevational view of an anterior portion of the knee joint in FIG. 1.

FIG. 4 is a flow chart illustrating a method of determining a type of knee injury in accordance with one embodiment of the present invention.

FIG. 5 is a schematic diagram of a diagnostic system in accordance with one embodiment of the present invention.

FIG. 6 is another schematic diagram of the diagnostic system of FIG. 5.

FIG. 7 is a schematic view of a knee brace in accordance with one embodiment of the present invention.

FIG. 8 is a side elevational view of a vibration detection module in accordance with one embodiment of the present invention.

FIG. 9 is a side elevational view of an exemplary shoe having a sensor array, for a shoe module, in accordance with one embodiment of the present invention.

FIG. 9A is an exemplary wireless transmitter for use with the shoe module of FIG. 9.

FIG. 9B is an enlarged view of one exemplary positional sensor of the shoe module of FIG. 9.

FIG. 10 is a schematic view of an ultrasound transducer wand for use with the diagnostic system in accordance with one embodiment of the present invention.

FIG. 11 is a diagrammatic view of an ultra wide band transmitter in accordance with one embodiment of the present invention.

FIG. 12 is a diagrammatic view of an ultra wide band receiver in accordance with one embodiment of the present invention.

FIG. 13 is a Cartesian coordinate system depicting an ultra wide band positioning system in accordance with one embodiment of the present invention.

FIG. 14 is a diagrammatic view comparing one embodiment of an ultra wide band positioning system to a global positioning system.

FIG. 15 illustrates the error in detecting a position along each of the x-, y-, and z-axes and with respect to a sequentially acquired series of data points.

FIG. 16 is an exemplary screen capture of a user interface of the diagnostic system of FIG. 5.

FIG. 17 is a side elevational view of a leg with a knee brace in accordance with another embodiment of the present invention.

FIG. 18 is a side elevational view of a leg with a knee brace in accordance with another embodiment of the present invention.

FIG. 19 is an individual transducer tracking sub-brace for use with a knee brace in accordance with one embodiment of the present invention.

FIG. 20 is an inter-transducers mechanical link sub-brace for use with a knee brace in accordance with one embodiment of the present invention.

FIG. 21 is a rotating transducer sub-brace for use with a knee brace in accordance with one embodiment of the present invention.

FIG. 22 is a diagrammatic representation of an inertia based localizer circuit in accordance with one embodiment of the present invention.

FIG. 23 is a diagrammatic representation of an alternate individual transducer tracking sub-brace circuit architecture in accordance with one embodiment of the present invention.

FIG. 24 is a diagrammatic representation of a high voltage circuit for use with a knee brace in accordance with one embodiment of the present invention.

FIG. 25 is a diagrammatic representation of the circuit layout of the high voltage circuit of FIG. 24.

FIG. 26 is a diagrammatic representation of a high voltage multiplexer for use with a sub-brace of a knee brace in accordance with one embodiment of the present invention.

FIG. 27 a diagrammatic representation of a receiving circuit for use with a sub-brace of a knee brace in accordance with one embodiment of the present invention.

FIG. 28 is a diagrammatic representation of a diagnostic system in accordance with an embodiment of the present invention.

FIG. 29 is a flow chart illustrating one method of using the diagnostic system of FIG. 28.

FIGS. 30A-30C illustrate various kinematic feature vectors acquired from a knee joint moving through a range of motion.

FIGS. 31A-31C illustrate feature vectors of a femoral position with respect to the tibia.

FIG. 32 illustrates average medial and lateral femoral condyle positions during a deep knee bend of a patient having an anterior cruciate ligament deficit.

FIG. 33 is a diagrammatic representation of a neural network classifier in accordance with one embodiment of the present invention.

FIG. 34 is a diagrammatic representation of a construction of a neural network.

DETAILED DESCRIPTION OF THE INVENTION

The exemplary embodiments of the present invention are illustrated and described below to encompass diagnosis of bodily abnormalities and, more particularly, devices and methods for evaluating the physiological condition of the musculoskeletal system (such as joints) to discern whether abnormalities exist and the extent of any abnormalities. Of course, it will be apparent to those of ordinary skill in the art that the exemplary embodiments discussed below are merely examples and may be reconfigured without departing from the scope and spirit of the present invention. However, for clarity and precision, the exemplary embodiments, as discussed below, may include optional steps, methods, and features that one of ordinary skill should recognize as not being a requisite to fall within the scope of the present invention. By way of example, the exemplary embodiments disclosed herein are described with respect to diagnosing a knee joint injury. Nevertheless, the exemplary embodiments may be utilized to diagnose other injuries of the musculoskeletal system (such as a hip joint injury or a bone fracture), as the knee joint 50 (FIG. 1) is merely exemplary to facilitate an understanding of the embodiments disclosed.

Turning now to the figures and in particular to FIG. 4, with reference also to FIG. 1, a low level exemplary process flow for a method 70 of determining a type of knee joint injury in accordance with one embodiment of the present invention is described. Still more particularly, the method 70 includes constructing a 3D model of the knee joint 50 (Block 72), which may include the detection of motion sound (Block 74) as well as tracking the kinematics (Block 76). The detected sound and tracked kinematics are automatically analyzed (Block 78) and the knee injury recognized based upon the analysis (Block 80).

FIG. 5 illustrates a first exemplary diagnostic system 82 for implementing the method 70 of FIG. 3. The diagnostic system 82 includes four modules: (1) a pulse echo A-mode ultrasound based 3D model reconstruction (“PEAUMR”) module 84 (FIG. 6) for constructing a patient-specific 3D-model of the patient's knee joint 50 (FIG. 1); (2) a joint kinematics tracking (“JKT”) module 86 for tracking the kinematics of the knee joint 50 (FIG. 1) using the patient-specific 3D model of the knee joint 50 (FIG. 1) from the PEAUMR module 84; (3) vibroarthography (“VA”) module 88 for capturing sounds emanating from the knee joint 50 (FIG. 1) while in motion; and (4) an intelligent diagnosis (“ID”) module 90 for identifying a likely diagnosis of the knee joint 50 (FIG. 1) using the kinematic data and the vibration data. Each of these four modules 84-90 is described in further detail below. If desired, a foot module 92 (FIG. 6) may be included with the JKT module 86 for providing dynamic force data, also described in detail below.

It will be understood by those of skill in the art that the diagnosis system 82 is usable with or without the use of the VA module 88. For example, the present invention may be used to mathematically describe the relative motion of the bones 52, 54, 56 in the patient's knee joint 50 as such motion is tracked on a 3D-patient specific bone model. The bone model and motion may be compared with a database of mathematical descriptions of joint motion. The database could contain mathematical descriptions of healthy or clinically undesirable joint motion.

As will be discussed in more detail hereafter, the interaction between bodily tissue (e.g., bone against cartilage or bone against bone) in a dynamic environment creates certain vibrations that are indicative of the condition or state of health of the joint. Even the healthiest and youngest joints create vibrations. However, joints that exhibit degradation, whether through wear or injury, will exhibit vibrations that are much more pronounced and amplified as compared to those of a healthy joint. The VA module 88 with the diagnostic system 82 utilizes those sounds, such as vibrations, exhibited by the joint during a range of motion to diagnose the condition of the joint without requiring an invasive procedure or subjecting the patient to radiation.

FIG. 6 provides still further details of the diagnostic system 82. The modules 84-92 may output the acquired data to a computer 96 for data processing by way of, for example, a neural network 98. The data processing, as will be discussed in more detail below, may provide one or more of a visual output, an audible output, and a diagnosis by way of a visual display 100.

Referring still to FIGS. 4-6, and now also FIG. 7, the VA module 88 is shown and comprises a plurality of accelerometers (three are shown 120a, 120b, 120c) that are utilized to detect sound, specifically, vibrations occurring as a result of motion of the knee joint 50. In this exemplary VA module 88, the accelerometers 120a, 120b, 120c are mounted directly to the skin or external tissue surface of the patient, as skin-mounted sensor 119s, in order to detect sounds from bone and soft tissue interaction. An intervening adhesive may be utilized between the accelerometers 120a, 120b, 120c. In the context of the knee joint 50, the VA module 88 includes one accelerometer 120a mounted on the medial side of the knee joint 50, a second accelerometer 120b mounted on the lateral side of the knee joint 50, and a third accelerometer 120c mounted on the front side of the knee joint 50, proximate the patella 56 (FIG. 3). As illustrated, the accelerometers 120a, 120b, 120c are mounted to the patient so that each lies along a common plane 121, though this is not required. It should also be understood, however, that any number of accelerometers 120a, 120b, 120c may be utilized to detect sounds generated by the patient's knee joint 50.

Each accelerometer 120a, 120b, 120c is in communication with one or more signal conditioning circuits or electronics 122. The accelerometers 120a, 120b, 120c are operative to detect sound, specifically vibrations, and output the sound detected in the form of frequency data (measured in Hertz) to the conditioning circuits 122. This frequency data is processed by the conditioning circuits 122 and communicated to the computer 96 as digital frequency data. While the accelerometers 120a, 120b, 120c are generating frequency data, the conditioning circuits 122 may include a clock 123 to time stamp the frequency data generated. As will be discussed in more detail below, correlating the frequency data with the time stamp provides a constant against which all of the detected data can be compared on a relative scale.

The first accelerometer 120a on the medial side of the knee joint 50 detects vibrations generated primarily by the interactions between the medial condyle 110 (FIG. 1) of the femur 52 against the medial cartilage 112 (FIG. 1) on top of the medial portion of the tibia 54. Similarly, the second accelerometer 120b on the lateral side of the knee joint 50 detects vibrations generated primarily by the interactions between the lateral condyle 114 (FIG. 1) of the femur 52 against the lateral cartilage 116 (FIG. 1) on top of the lateral portion of the tibia 54. The third accelerometer 120c on the front of the knee joint 50, proximate the patella 56 (FIG. 3), detects vibrations generated primarily by the interactions between the femur 52 against the patella 56 (FIG. 3). The resulting data output by the accelerometers 120a, 120b, 120c may then be wirelessly transmitted to the computer 96 via a wireless transmitter 124, such as an ultra-wide band transmitter, and utilized in combination with data from the other modules to ascertain the appropriate diagnosis.

FIG. 8 illustrates one example of a plurality of thin film accelerometers (four are shown, 120a, 120b, 120c, 120d) that are suitable for detecting the vibrations produced by motion of the knee joint 50. Thin film accelerometers 120a, 120b, 120c, 120d may be used in lieu of sound sensors because of better performance and less noise susceptibility. The thin film accelerometers 120a, 120b, 120c, 120d may also be used as a localizer and include the same circuitry. The accelerometers 120a, 120b, 120c, 120d are attached to the patients so the outputs may be amplified, digitized, and sent wirelessly to the computer 96 as described below.

With reference now to FIGS. 4-6 and 9, the foot module 92 (also referred to as the contact force module (“CFM”)) is shown and includes a plurality of pressure sensors 130 that are utilized to detect pressure or a contact force occurring at the bottom of the foot (not shown) when the knee joint 50 (FIG. 1) is moved through a range of motion under a loaded condition. In other words, as the patient walks, jogs, runs, etc., the foot module 92 detects pressure data at the bottom of the foot when the foot is partially or fully in contact with the ground. In exemplary form, the pressure sensors 130 are incorporated into an insole 132 of a shoe 134 that conforms to the general shape of a patient's foot. Because humans have different sized feet, the insoles 132 may be incrementally sized to accommodate humans with differently sized feet or to accommodate a particular type of shoe 134 (or lack thereof) needed for a particular activity.

The pressure sensors 130 may be arranged in a grid-shaped pattern on the insole 132, which may include a series of rows and columns. The pressure sensors 130 are exposed to the underside of a patient's foot so that the location and amplitude (or amount) of the contact forces applied by the foot to the shoe 134, by way of the insole 132, may be measured. As will be discussed in more detail hereafter, the location of the pressures and the relative amount of pressures provides information relevant to diagnosis of injury. For example, the detected pressures of a patient with a limp caused by a knee joint injury would differ from the detected pressures of a patient with a healthy knee joint and a normal gait.

In one embodiment, each sensor 130 may include a capacitor having a deformable dielectric between two electrode plates. Changes in the pressure applied to the plates cause a strain, or deformation, of the dielectric medium. Thus, a pressure applied to the capacitive sensor 130 changes the spacing between the plates and the measured capacitance. The capacitive sensors 130 are arrayed across the area of pressure measurement to provide discrete pressure data points corresponding to strains/deformation at the various locations of the array. These strains/deformations are used to find the stresses and thus the compressive forces and to calculate the output of pressure data having units of force per unit area and time (i.e., N/m sec).

The sensors 130 in the grid-shape enable positioning of each detected pressure from each of the sensors 130 relative to another sensor 130. The resultant data, which includes a two-dimensional map of the pressure sensors 130, is either stored on the computer 96 or stored locally with the sensors 130. The resulting data may be wirelessly transmitted to the computer 96 via a wireless transmitter 136, such as an ultra-wide band transmitter. Using the 2D map of the sensors 130 stored on the computer 96 in combination with the received sensor pressure data, the computer 96 is operative to generate data tying detected pressure to position, specifically the position of one pressure sensor 130 with respect to another.

By tying amounts of compressive force to its applied position, the foot module 92 provides data reflecting precisely what pressures are exerted at what location. In addition, the computer 96 may include an internal clock 97 to associate a time of which the pressure is applied with the pressure data generated by the pressure sensors 130. Accordingly, the diagnostic system 82 not only knows how much pressure was exerted and the location where the pressure was applied, but also has time data indicating the duration of the applied pressures. Again, by tying the pressure data generated by the pressure sensors 130 to time, the pressure data can be correlated with the sound data generated by the VA module 88 using a common time scale. As a result, the diagnostic system 82 may evaluate how pressures exhibited at the bottom of the foot change as a function of time, along with how the vibrational data changes during the same time.

FIGS. 4-6 and 10 illustrate the details of the JKT module 86, which comprises an ultrasound creation and positioning submodule 140, an ultrasound registration submodule 142, and an ultrasound dynamic movement submodule 144. Specifically, each submodule 140, 142, 144 includes an A-mode ultrasound transducer to generate sound and to detect reflected sound, wherein the reflected sound is representative of the structure, position, and acoustical impedance of the knee joint 50 (FIG. 1). Commercially-available transducers may include, for example, an immersion unfocused 3.5 MHz transducer, such as those that are available from Olympus Corp. (Tokyo, Japan). Those skilled in the art are familiar with the operation of ultrasound transducers generally and, more specifically, an A-mode ultrasound transducer that generates sound pulses and detects sound that is reflected at tissue boundaries of tissues having different acoustic impedances. The magnitude of the reflected sound and the time delay are utilized to determine the distance between the ultrasound transducer and the tissue interface.

In the illustrated embodiment, the A-mode ultrasound transducers are utilized to detect the interface between bone and the surrounding soft tissue so that the location of the bone surface may be determined. Because the operation of ultrasound transducers (including the A-mode ultrasound transducers) is well known to those skilled in the art, a detailed discussion of the operation of ultrasound transducers in general, and A-mode ultrasound transducers specifically, has been omitted only for purposes of brevity.

The ultrasound creation and positioning submodule 140 as shown in FIG. 10 comprises one or more A-mode ultrasound transducers 150 fixedly mounted to a wand 152. The wand 152 further includes at least one positioning device 170. In this exemplary embodiment, the ultrasound creation and positioning submodule 140 is physically separate from the ultrasound registration submodule 142 (FIG. 6) and the ultrasound dynamic movement submodule 144 (FIG. 6), the latter two of which are mounted to a rigid knee brace 220 schematically illustrated in FIG. 17. In this fashion, the ultrasound creation and positioning submodule 140 is repositionable with respect to the rigid knee brace 220 (FIG. 17) and adapted to place one or more of its A-mode ultrasound transducers 150 in contact with the patient's epidermis, proximate the knee joint 50 (FIG. 1). It should be noted, however, that the knee brace 220 (FIG. 17) does not have to be rigid, other than the linkages between certain components. Moreover, the knee joint 50 (FIG. 1) may be scanned by the ultrasound wand 152 before positioning the brace 220 (FIG. 17) thereon.

One of the functions of the ultrasound creation and positioning submodule 140 is to generate an electrical signal that is representative of the ultrasonic wave detected by the transducers 150 as the wand 152 moves over the patient's epidermis, proximate the knee joint 50 (FIG. 1). The ultrasound transducers 150 receive the ultrasonic wave based upon the magnitude of the reflected ultrasonic wave from the bone-tissue interface. As discussed previously, the magnitude of the electrical signal and the delay between the generation of the ultrasonic wave by the ultrasound transducer 150 to detection of the reflected ultrasonic wave by the ultrasound transducer 150 is indicative of the distance to the bone underneath the transducer 150. But, distance data alone is not particularly useful; therefore, one or more positioning devices 170 are used to provide a 3D coordinate system, one example of which is shown in FIG. 11.

The positioning devices 170 of the ultrasound creation and positioning submodule 140 are fixedly mounted to the wand 152 and may include any of a number of positioning devices 170. For example, the wand 152 may include one or more optical devices (as the positioning devices 170) that are configured to generate, detect, and/or reflect pulses of light. These pulses of light interact with a corresponding detector or light generator to discern the position of the wand 152, in 3D space, and with respect to a fixed or reference position. One such device includes a light detector configured to detect pulses of light emitted from light emitters having known positions. The light detector detects the light and sends a representative signal to the computer 96 or otherwise a controller (not shown) of the light detector. The computer 96 is also provided the time at which the light pulses were emitted by the optical devices 170. In this matter, the computer 96 determines the position of the wand 152 relative to the known positions of the detectors. Because the ultrasound transducer 150 and the optical devices 170 are fixedly mounted to the wand 152, the position of the ultrasound transducers 150 with respect to the position of the optical devices 170 is known. Similarly, because the ultrasound transducers 150 are generating signals representative of the straight line distance between the transducers 150 and the bone-tissue interface, and the position of the transducers 150 with respect to the optical devices 170 is known, the position of the bone-tissue interface with respect to the optical devices 170 may be determined. In other words, as the wand 152 moves over the patient's epidermis, the optical devices 170 generate data that is determined, by the computer 96, to represent that the relative position of the optical devices 170 with respect to the light detectors has changed in the 3D coordinate system. This change in the position of the optical devices 170 may be easily correlated to the position of the bone-tissue interface, in 3D, because the position of the bone-tissue interface relative to the ultrasound transducers 150, as well as the position of the optical devices 170 with respect to the ultrasound transducers 150 are known. Accordingly, the 3D position data may be used in combination with the fixed position data (distance data for the position of the ultrasound transducers 150 with respect to the optical devices 170) for the ultrasound transducers 150 in combination with the distance data generated in response to the signals received from the ultrasound transducers 150 to generate composite data. The composite data may, in turn, be used to create a plurality of 3D points representing a plurality of distinct points on the surface of the bone, along the bone-tissue interface. As will be discussed in more detail below, these 3D points are utilized in conjunction with a default bone model to generate a virtual, 3D representation of the patient's bone.

Alternatively, the positioning devices 170 may comprise one or more inertial measurement units (“IMUs”). IMUs are known to those skilled in the art and include accelerometers, gyroscopes, and magnetometers that work together to determine the position of the IMUs in a 3D coordinate system. Because the A-mode ultrasound transducer 150 and the IMUs 170 are fixedly mounted to the wand 152, the position of the ultrasound transducers 150 with respect to the position of the IMUs 170 is known. Similarly, because the ultrasound transducers 150 are generating signals representative of the straight line distance between the transducers 150 and the bone-tissue interface and the position of the transducers 150 with respect to the IMUs 170 is known, the position of the bone-tissue interface with respect of the IMUs 170 may be determined. In other words, as the wand 152 moves over the patient's epidermis, the IMUs 170 generate data that is determined, by the computer 96, to represent that the relative position of the IMUs 170 has changed in the 3D coordinate system. This change in the position of the IMUs 170 may be easily correlated to the position of the bone-tissue interface in 3D because the position of the bone tissue interface relative to the ultrasound transducer 150 is known, as is also the position of the IMUs 170 with respect to the ultrasound transducers 150. Accordingly, the 3D position data may be used in combination with the fixed position data (distance data for the position of the ultrasound transducers 150 with respect to the IMUs 170) for the ultrasound transducers 150 in combination with the distance data generated in response to the signals received from the ultrasound transducers 150 to generate the composite data as described above.

Referring now also to FIGS. 11-12, the positioning devices 170 may still alternatively comprise one or more ultra-wide band (UWB) transmitters. UWB transmitters are known to those skilled in the art, but the use of UWB transmitters and receivers for millimeter resolution 3D positioning is novel. In that regard, one or more UWB transmitters 170 are fixedly mounted to the wand 152 and configured to sequentially transmit UWB signals to three or more UWB receivers 172 having known positions in a 3D coordinate system. This embodiment of the positioning device 170 is comprised of active tags or transmitters 170 that are tracked by the UWB receivers 172. The system architecture of the UWB transmitter 170 is shown in FIG. 11 where a low noise system clock (“crystal clock”) 174 triggers a baseband UWB pulse generator 176 (for instance a step recovery diode (“SRD”) pulse generator). The baseband pulse from the baseband UWB pulse generator 176 is upconverted by a local oscillator 178 via a double balanced wideband mixer (not shown). The upconverted signal is amplified and filtered (“bandpass filter” 180). Finally the signal is transmitted, via an omnidirectional antenna 182, to the computer 96 (FIG. 6). The UWB signal may travel through an indoor channel where significant multipath and pathless effects cause noticeable signal degradation.

The UWB receiver 172 architecture in accordance with one embodiment is shown in FIG. 12. The signal is received via a directional UWB antenna 184 and is filtered and amplified (“bandpass filter” 186), downconverted by a local oscillator 187, and low-pass filtered (“LPF”) 188. A sub-sampling mixer 190 triggered by a second low noise system clock (“crystal clock”) 192 is used to tune extend the pulse by about 1,000 to about 100,000 times. This effectively reduces the bandwidth of the UWB pulse and allows sampling by a conventional analog-to-digital converter (“ADC”) 194.

Each UWB transmitter 170 and receiver 172 is in communication with the computer 96. Accordingly, the computer 96 detects each time the UWB transmitter 170 transmits a UWB signal, as well as the time at which the UWB signal was transmitted. Similarly, the computer 96 detects the position of each of the UWB receivers 172 in the 3D coordinate system, as well as the time at which the UWB signal was received. The final time-difference-of-arrival (“TDOA”) calculation, via a UWB positioning system 183, is shown in FIG. 13.

Referring now to FIG. 13, for the TDOA calculation, at least four base receivers 172 (Rx1, Rx2, Rx3, Rx4) are needed to localize the 3D position of the UWB transmitter 170 (“Tag”). The geometry of the receivers Rx1, Rx2, Rx3, Rx4 has important ramifications on the achievable 3D accuracy through what is known as geometric position dilution of precision (“PDOP”). A combination of novel filtering techniques, high sample rates, robustness to multipath interference, accurate digital ranging algorithms, low phase noise local oscillators, and high integrity microwave hardware are needed to achieve millimeter range accuracy (e.g. ranging from about 5 mm to about 7 mm in 3D real-time). An analogy of the UWB positioning system 183 to a GPS system 185 is shown in FIG. 14.

FIG. 15 shows actual experimental errors in each of the x-, y-, and z-coordinates for detecting the position of the UWB transmitter 170 in 3D space and in real-time for over 1000 samples while the transmitter 170 is moving freely within the 3D space.

Because the A-mode ultrasound transducers 150 and the UWB transmitters 170 are fixedly mounted to the wand 152, the position of the ultrasound transducers 150 with respect to the position of the UWB transmitters 170 is known. Similarly, because the ultrasound transducers 150 are generating signals representative of the straight line distance between the ultrasound transducers 150 and the bone-tissue interface and the position of the ultrasound transducers 150 with respect to the UWB transmitters 170 is known, the position of the bone-tissue interface with respect to the UWB transmitter 170 may be determined. In other words, as wand 152 moves over the patient's epidermis, the UWB transmitters 170 transmit UWB signals that are correspondingly received by the UWB receivers 172. These UWB signals are processed by the computer 96 in order to discern whether the relative position of the UWB transmitters 170 has changed in the 3D coordinate systems, as well as the extent of such a change. This change in 3D position of the UWB transmitters 170 can be easily correlated to the position of the bone-tissue interface in 3D because the position of the bone relative to the ultrasound transducer 150 and the position of the UWB transmitters 170 with respect to the ultrasound transducer 150 are known. Accordingly, the UWB 3D position data may be used in combination with the fixed position data (distance data for the position of the ultrasound transducers 150) to generate the composite data as described above.

Regardless of the positioning device 170 utilized with the ultrasound creation and positioning submodule 140, the wand 152 is repositioned over the skin of the patient, proximate to the knee joint 50 (FIG. 1) while the knee joint 50 (FIG. 1) is bent. Bending the patient's knee joint 50 (FIG. 1) during data acquisition enables the creation of a 3D series of points for each of the bones of the knee joint 50 (FIG. 1) (the distal femur 52, the proximal tibia 54, and the patella 56). Thus, as the wand 152 is repositioned, the data from the transducer 150 is transmitted to a wireless transmitter 200 mounted to the wand 152. When the wireless transmitter 200 receives the data from the transducers 150, the transmitter 200 transmits the data via a wireless link to the computer 96.

In order to power the devices on-board the wand 152, an internal power supply (not shown) may be provided. In one embodiment, the internal power supply comprises one or more rechargeable batteries.

Transformation is needed for transforming the position data from a reference coordinate frame of reference to a world frame of reference. According to one embodiment of the present invention, a linear movement of the ultrasound transducer 150 may be described:

v(n+1)=v(n)+a(n)dt  Equation 1

s(n+t)=s(n)+v(n)dt=0.5a(n)dt²  Equation 2

where s(n+1) is the position of the ultrasound transducer 150 at a current state, s(n) is the position from a previous state, v(n+1) is the instantaneous velocity of the current state, v(n) is the velocity from previous state, a(n) is the detected acceleration, and dt is the sampling time interval. The previous equations describe the dynamic motion and positioning of a point in 3D Euclidean space. Additional information is needed to describe 3D orientation and motion.

The orientation of the ultrasound transducer 150 may be described by using a gravity-based accelerometer (for example ADXL-330, analog device) and extracting the tilting information from each of a pair of orthogonal axes. The acceleration output on each of the x-, y-, or z-axes is due to gravity and is equal to the following:

A_i=(V_outx−V_off)=S  Equation 3

where A_i is the acceleration of the ultrasound transducer 150 along each of the x-, y-, or z-axes, V_outx is the voltage output on each of the x-, y-, or z-axes, V_off is the offset voltage, and S is the sensitivity of the accelerometer. The yaw, pitch, and roll may be thus calculated as:

$\begin{array}{cc}\rho =\arctan \left(\frac{A_x}{\sqrt{A_y^2+A_z^2}}\right)& \mathrm{Equation}4\\ \varphi =\arctan \left(\frac{A_y}{\sqrt{A_x^2+A_z^2}}\right)& \mathrm{Equation}5\\ \theta =\arctan \left(\frac{\sqrt{A_y^2+A_x^2}}{A_z}\right)& \mathrm{Equation}6\end{array}$

where pitch is ρ (the x-axis relative to the ground), roll is φ (the y-axis relative to the ground) and roll is θ (the z-axis relative to the ground). Since the accelerometer is gravity-based, the orientation does not require information from the previous state once the accelerometer is calibrated. The static calibration requires the resultant sum of accelerations from each of the three axes to equal 1−g (where g is the nominal acceleration due to gravity at the Earth's surface at sea level, defined to be precisely 9.80665 m/s² (approximately 32,174 ft/s²)). Alternatively, an orientation sensor that provides yaw, pitch and roll information of the bodily tissue in question may be used. One such orientation sensor may be the commercially-available model IDG-300 from InvenSense (Sunnyvale, Calif.). The orientation of the ultrasound transducer 150 may then be resolved by using, for example, a direction cosine matrix transformation:

X₂CθCφCθCφS_p−SθC_pCθSφC_p−SθS_pX₁

Y₂=SθCφSθSφS_p−CθC_pSθSφC_p−CθS_pY₁  Equation 7

Z₂−SφCφS_pCθC_p

where C represents cosine and S represents sine.

Referring again to FIGS. 4-6, and now also to FIG. 16, the PEAUMR module 84 is described in greater detail. The PEAUMR module 84 constructs a 3D model of the patient's knee joint 50 (FIG. 6) by converting the transcutaneously acquired a set of 3D data points (using the tracked pulse echo A-mode ultrasound transducer 150), that, in total, are representative of the shape of the bone-tissue interface and therefore each bones' surface.

Before the patient data is acquired, software residing on the computer 96 may request a series of inputs from the user to adapt the diagnostic system 82 to equipment specific devices and the particular portion of the musculoskeletal anatomy to be modeled. For example, a menu 204 on a user interface 206 may be presented for the user to select the type of digitizer, which may include, without limitation, ultrasound. After the type of digitizer is selected, the user may actuate buttons 205a, 205b to connect to or disconnect from the digitizer, respectively.

As wand 152 moves over the patient's epidermis, the set of points is generated, numerically recorded, viewable in a data window 210, and ultimately utilized by the software to conform a selected bone model to the patient's actual bone shape. Consequently, the wand 152 is repositioned over the bones (the distal femur 52, the patella 56, the proximal tibia 54) for approximately 30 seconds so that the discrete points to typify the topography of the bone. Repositioning the wand 152 over the bone in question for a longer duration results in more 3D points being generated increases the resolution and improves the accuracy of the patient-specific bone model. A partial range of motion of the knee joint 50 (FIG. 1) while repositioning the wand 152 over the knee joint (FIG. 1) aids in scanning additional portions of the bone in question for new 3D points that may have been obscured by other bones in another range of motion position.

Before, during, or after the ultrasound data is acquired, the software provides various drop-down menus allowing the software to load a bone model 208 that is roughly the same shape as the patient's bone. The computer 96 receives the ultrasound data, the computer 96 includes software that interprets the A-mode ultrasound transducer data and constructs a 3D map having discrete 3D points corresponding to points on the surface of the scanned bone. That is, the shape of the patient's bone is reconstructed in virtual space, using a set of points outlining the surface of the patient's bone as acquired by the tracked ultrasound transducer 150 (FIG. 10). The set of points is applied to an atlas-based deformable model software to reconstruct the patient-specific 3-D model.

More specifically, the computer 96 may include a database having a plurality of bone models of various portions of the musculoskeletal system, for example, the femur 52, the tibia, 54, and the patella 56, that are classified and selectable in a menu 212, for example based upon ethnicity, gender, height ranges, the side of the body, and so forth. Each of these classifications is accounted for in a drop-down menu of the software so that the model initially chose by the software most closely approximates the body of the patient.

For mapping each bone, the computer 96 uses either a default bone model or the selected bone model as a starting point to construction of the ultimate patient-specific, virtual bone model. The default bone model may be a generalized average, as the morphing algorithms use statistical knowledge of a wide database population of bones for a very accurate model. The selected bone model expedites computation. For example, in the case of generating a patient-specific model of the femur 52 where the patient is a 53 year old, Caucasian male, who is six feet tall, a default femoral bone model is selected based upon the classification of Caucasian males having an age between 50-60, and a height ranging from 5′10″ to 6′2″. In this manner, selection of the appropriate default bone model more quickly achieves an accurate patient-specific, virtual bone model because of the number of iterations between the patient's actual bone (typified by the 3D map of bone points) and the default bone model are reduced. Nevertheless, in view of the model bones taking into account numerous traits of the patient (ethnicity, gender, bone modeled, and body side of the bone), it is quite possible to construct an accurate patient-specific 3D model with as few as 150 data points comprising the set which typically may be acquired by repositioning the wand 152 over the bone for 30 seconds for each bone. Ultrasound will not be affected whether the patient has a prosthetic implant.

After the appropriate bone model is selected, the computer 96 superimposes the 3D points onto the default bone model and, thereafter, carries out a deformation process so that the bone model exhibits the 3D bone points detected during the signal acquisition. The deformation process also makes use of statistical knowledge of the bone shape based upon reference bones of a wide population. After the deformation process is complete, the resulting bone model is a patient-specific, virtual 3D model of the patient's actual bone. The foregoing process is repeated for each bone comprising the specific joint to create patient-specific, virtual 3D models of the patient's anatomy.

Referring back to FIGS. 4-7, and now also FIG. 17, the JKT module 86 may be configured to track the kinematics of the knee joint 50 (FIG. 1) and display the kinematics on the patient-specific 3D bone model generated by the PEAUMR module 84 using, for example, one or more bone motion tracking braces 220. Generally, the bone motion tracking brace 220 includes pulse echo A-mode ultrasound transducers 222 to transcutaneously localize the bone-tissue interface and derive a set of points outlining each bone's surface.

Turning specifically to FIG. 17, the brace 220 includes a plurality of A-mode ultrasound transducers 222 fixedly mounted to the knee brace 220. Specifically, in the context of a knee joint 50, there are at least two A-mode ultrasound transducers 222 (i.e., “a transducer group” 222a, 222b) fixedly mounted to the knee brace 220 for tracking of the tibia 54 (FIG. 1) and the femur 52 (FIG. 1). In other words, the knee brace 220 includes at least six ultrasound transducers 222 in order to track the two primary bones 52, 54 (FIG. 1) of the knee joint 50. Each transducer group 222a, 222b includes a rigid, mechanical connection linking the transducers 222 and the positioning devices 224 to the knee brace 220. In this manner, the relative positions of the transducers 222 with respect to one another do not change. A first transducer group 222a at least partially circumscribes a distal portion of the femur 52 (FIG. 1); while a second transducer group 222b at least partially circumscribes a proximal portion of the tibia 54 (FIG. 1); and an optional third transducer group (not shown) overlies the patella 56 (FIG. 3) if patella kinematics are desired. The ultrasound registration submodule 142 is accordingly configured to provide a plurality of static reference points for each bone as the bone is moved through a range of motion.

Each ultrasound transducer 222 is tracked using an accelerometer or a sensor-specific localizer (or any other appropriate inertial sensor). The tracking may then be used to generate localized bone points from the outputs of the ultrasound transducers 222 and to virtually display bone movement on the 3D model while the knee joint 50 (FIG. 1) is taken through the range of motion.

Referring to FIGS. 6 and 17, the ultrasound dynamic movement submodule 144 comprises a plurality of positioning devices 224 that is configured to feed information to the computer 96 regarding the 3D position of each transducer group 222a, 222b of the ultrasound registration submodule 142. In exemplary form, the position devices 224 may comprise light detectors operative to detect pulses of light emitted from light emitters having known positions. The light detectors 224 detect the light and transmit representative signals to a control circuitry (not shown) associated with the knee brace 220. The knee brace 220 transmits this information to the computer 96, which also knows when the light pulses were emitted as a function of time and position. In this manner, the computer 96 may determine the position of the transducers 222 in the 3D coordinate system. Because the ultrasound transducers 222 and the optical devices 224 are fixedly mounted to the knee brace 220, the position of the ultrasound transducers 222 with respect to the position of the optical devices 224 is known. Similarly, because the ultrasound transducers 222 are generating signals representative of the straight line distance between each of the ultrasound transducers 222 and the bone-tissue interface beneath, and the position of the ultrasound transducers 222 with respect to the optical devices 224 is known, the position of the bone-tissue interface with respect to the optical devices 224 may be easily determined. In other words, as the knee joint 50 (FIG. 1) is moved, and correspondingly so too is the knee brace 220, the optical devices 224 generate data that is determined by the computer 96 that the relative position of the optical devices 224 has changed in the 3D coordinate system. This change in the position of the optical devices 224 may be easily correlated to the position of the bone in question in 3D because the position of the bone relative to the ultrasound transducer groups 222a, 222b is known, as is the position of the optical devices 224 with respect to the ultrasound transducer groups 222a, 222b. Accordingly, the optical devices 224 generate data that is used in combination with the fixed position data (distance data for the position of the ultrasound transducers 222) to generate the composite data. The composite data may, in turn, be used to create dynamically moving map of the bone on the patient-specific 3D model.

Alternatively, the positioning devices 224 may be comprised of one or more IMUs. Because the ultrasound transducers 222 and the IMUs 224 are fixedly mounted to the knee brace 220, the relative positions between the ultrasound transducers 222 and the IMUs 224 are known. Similarly, because the ultrasound transducers 222 are generating signals representative of the straight line distance between the transducer 222 and the bone-tissue interface, and the position of the transducers 222 with respect to the IMUs 224 is known, the position of the bone with respect to the IMUs 224 may be easily determined. In other words, as the knee joint 50 (FIG. 1) with the knee brace 220 moves, the IMUs 224 generate data that is determined, by the computer 96, as a change in the position of the IMUs 224. This change in the position of the IMUs 224 may be easily correlated to the position of the bone in 3D because the position of the bone relative to the ultrasound transducer groups 222a, 222b is known, as well as the position of the IMUs 224 with respect to the ultrasound transducer groups 222a, 222b. By way of example, because the ultrasound transducers 222 do not move with respect to the knee brace 220, any movement of the IMUs 224 in space means that the knee brace 220 has also moved in space, and by continuing to track the distance data provided by each IMU 224, the movement of the bone may be correspondingly tracked. IMU tracking of the bone movements requires a static registration between the IMUs 224 and an initial known body position (such as standing). The IMUs 224 enable measurement of the relative motion between different bones via their corresponding ultrasound transducer group data and the IMU data. The IMUs 224 may be used alone or in conjunction with other positioning devices 170 (FIG. 10), such as those described in detail above. In this scenario, the IMU position is updated at a certain interval with the absolute position provided by the additional positioning system to minimize error. Therefore the two positioning systems act together as one positioning system.

As was described previously with respect to the wand 152, the positioning devices 224 of the brace 220 may alternatively be comprised of one or more ultra wide band (UWB) transmitters. In that regard, one or more UWB transmitters 224 are fixedly mounted to the brace 220 and operable to transmit sequential UWB signals to three or more UWB receivers (not shown) having known positions in the 3D coordinate system. Each UWB transmitter 224 is in communication with the computer 96, as are the plurality of UWB receivers (not shown). Accordingly, the computer 96 detects each time the UWB transmitter transmits a UWB signal, as well as the time at which the UWB signal was transmitted. Similarly, the computer 96 detects the position of each of the UWB receivers (not shown) in the 3D coordinate system, as well as the time at which the UWB signal was received. The computer 96 may then use the custom digital signal processing algorithms to accurately locate the leading-edge of the received UWB pulse based on the position of each UWB receiver (not shown), the time when each UWB signal was received, and the time that the UWB signal was transmitted. The position may then be determined by the TDOA calculation as was described with reference to FIG. 11). Again, because the ultrasound transducers 222 do not move with respect to the knee brace 220, any movement of the transducers 222 in space means that the brace 220 has moved. The movement of the knee brace 220 is tracked using the computer 96 in combination with the UWB transmitters 224 and the UWB receivers (not shown). Similarly, because the fixed orientation between the UWB transmitters 224 and the ultrasound transducers 222 changes in position in the 3D coordinate system, the UWB transmitter 224 may correspondingly be used to track movement of each bone.

In order to communicate information from the submodules 142, 144 to the computer 96, the brace 220 may include a transmitter 228, such as a UWB transmitter, in communication with the ultrasound transducer 222 to facilitate wireless communication of data to the computer 96. It should be noted that if UWB transmitter 228 is also utilized as the positioning devices 224, a dedicated transmitter 228 is unnecessary as the UWB transmitters 224 could function to also send ultrasound data directly to the computer 96 over a wireless link.

It should be understood that use of the transmitter 228 and a field programmable gate array design enables the computations to be cammed out on a real-time basis. For example, as patient's knee joint 50 (FIG. 1) is bent while wearing the brace 220, the ultrasound data is immediately transmitted to the computer 96, which in real-time, calculates and displays the position and movement of each bone with the 3D patient-specific bone model.

FIG. 18 illustrates a knee brace 230 in accordance with another embodiment of the present invention. The knee brace 230 has a first sub-brace 232 positioned at the distal portion of the femur 52, a second sub-brace 234 positioned at the proximal end of the tibia 54, and a third sub-brace 236 positioned at the patella 56 (FIG. 3). The sub-braces 232, 234, 236 include a plurality of transducers mounted thereto. Each transducer is responsible for determining the location of a point on the surface of the bone during movement of the knee joint 50. The sub-braces 232, 234, 236 reduce the occurrence of problems of locating and tracking the bone using ultrasound data when the motion of the bone relative to the skin is small compared to the gross joint motion. There are at least three approaches disclosed herein for tracking the motion of the ultrasound transducers themselves.

FIG. 19 illustrates the first approach commonly referred to herein as an “ITT” (individual transducer tracking) approach. In FIG. 19, each transducer 238 in the sub-brace 232 has an associated tracking module 240 to individually track each transducer 238. Using the ITT approach, the transducers 238 may be supported by a flexible length of strap.

Referencing FIG. 20, a sub-brace 241 according to the second approach is shown. The second approach, commonly referred to herein as an “ITML” (Inter-Transducers Mechanical Links) approach, involves the transducers 242 being connected to each other by movable mechanical links 244. Each mechanical link 244 includes length and angle sensors 246 that allow for detection of the movement of the transducers 242 relative to one another and the relative translational motions of the links 244. Every two links 244 are connected by a pivot pin 248 that allows rotation and translation of the links 244 relative to each other. The length and angle sensors 246 are mounted to at least one link 244 and proximate to the pivot pin 248 to allow for detection of the angle between adjacent 244 links. The ITML approach features a fewer number of localizers than the ITT approach of FIG. 19.

Referring now to FIG. 21, a sub-brace 249 according to the third approach is shown. The third approach, commonly referred to herein as a “RT” (Rotating Transducer) approach, involves using a single ultrasound transducer 250 that is mounted to a carriage 252. The carriage 252 traverses along a track 254, located on the inner circumference 256 of the sub-brace 249. For example, the carriage 252 may be moved along the track 254 by a string loop 258 that is wrapped around the drive shaft (not shown) of a motor 260. When the transducer 250 reaches the motor 260, the rotation direction of the motor 260 is changed and the transducer 250 moves in the opposite direction.

A tracking module 262 such as an inertia-based localizer is mounted to the transducer 250 to track its motion. As the transducer 250 rotates within the inner circumference 256 of the sub-brace 249, it collects data as to the bone-tissue interface. By using a single transducer 250, the RT approach includes the advantage of lower cost than the stationary transducer designs and higher accuracy due to the greater number of localized bone surface points for each tracking step, while maintain a mechanical flexibility.

Referring to FIG. 22, a localizer 270 of tracking each ultrasound transducer 238 (FIG. 19) mounted to the sub-brace 232 (FIG. 19) is shown. The localizer 270 comprises a plurality of nodes 272 with each node 272 comprising a CMOS accelerometer and a temperature sensor (not shown) for thermal drive comparison. Each node 272 is integrated to minimize noise and distortion. The outputs of the accelerometers 272 regarding the x-, y-, and z-coordinates and the temperature sensors (not shown) are directed to a multiplexer 274 (“MUX”) that multiplexes the signals. Multiplexed outputs are amplified by an amplifier 276 (“AMP”), and then directed to an ADC 278. The digital conversion of the signal may be performed within or outside the accelerometers 272. Outputting digital signals may then be directed to a wireless transmitter 280 by way of a parallel input/serial output device 282.

In FIG. 23 a design alternative for the sub-brace 232 is shown. The electronic architecture includes a high voltage amplifier circuit 286 (“HV IX AMP”) feeding a voltage multiplex circuit (“HV MUX”) 288 to excite each ultrasound transducer 238 and thereby acts as an analog switch. The echo signals from each transducer 238 are multiplexed pursuant to a logic control directing the opening of the switches in the MUX 290 at precise intervals. An exemplary logic control is the MSP430, available from Texas Instruments, Inc. (Dallas, Tex.). The output from the MUX 290 is I amplified by a low noise AMP 292 (“LNA”) and the signal is conditioned using a conditioning circuit (for example, a time-gain-control (“TGC”) circuit 294 and a band-pass filter (“BPF”) 296, and digitized using an ACS 298. Electric power to the foregoing components is supplied by way of a battery 300, which also supplies power to a wireless transmitter module 302. In exemplary form, the wireless transmitter module 302 utilizes a universe asynchronous receiver/transmitter (“UART”) protocol. The wireless transmitter module 302 includes a wireless transmitter circuit 304 receiving the output from a first in-first out (“FIFO”) buffer (not shown) of the ADS 298 by way of a serial interface 306. An output from the wireless transmitter circuit 304 is conveyed using a serial link coupled to an antenna 308. Signals conveyed through the antenna 308 are broadcast for reception by a wireless receiver (not shown) coupled to a controller (not shown) or the computer 96 (FIG. 6).

Referring now to FIGS. 24 and 25, an exemplary high voltage circuit 310 is shown and may be used to trigger and generate the excitation energy for a piezoelectric crystal in the ultrasound transducer 238 (FIG. 19). Exemplary high voltage circuits 10 for use in this embodiment may include, without limitation, the pulsar integrated circuit (HV379) available from Supertex, Inc (Sunnyvale, Calif.).

Referencing FIG. 26, an exemplary high voltage multiplexer 312 is shown and may be used to trigger and excite multiple piezoelectric transducers 238 (FIG. 19) without increasing the number of high voltage circuits 310 (FIG. 24). Exemplary high voltage multiplexer 312 for use in this embodiment may include, without limitation, the high voltage multiplexer (HV2221) available from Supertex, Inc (Sunnyvale, Calif.). The advantage of using a high voltage multiplexer 312 is the ability to use CMOS level control circuitry, thereby making the control logic compatible with virtually any microcontroller or field programmable gate array that is commercially-available.

Referring to FIGS. 23 and 27, an exemplary receiving circuit 314, which comprises the MUX 290, the LNA 292, the TGC 294, the BPF 296, and the ADC 298 is shown and may be utilized to receive the echo signals from each transducer 238. Exemplary receiving circuits 314 for use in the this embodiment include, without limitation, the AD9271 8-channel ultrasound receiving integrated circuits, available from Analog Devices, Inc. (Norwood, Mass.).

With reference now to FIGS. 28 and 29, one method 316 of using X-ray fluoroscopy and in-vivo measurements of dynamic knee kinematics, as described above, for understanding the effects of joint injuries, diseases, and evaluating the outcome of surgical procedures is described. In the particular illustrated embodiment, and using the two aforementioned techniques, six degrees of freedom (“DOF”) are determined for the knee joint 50 (FIG. 1) and include the position and orientation of each bone comprising the knee joint 50 (FIG. 1). The accuracy of this method 316 is within 1° of rotation and 1 mm of translation (except for translations that are parallel to the viewing plane).

Implementation of the method 316 includes joint movement visualization via the 3D model reconstruction with A-mode ultrasound system, as described previously. The method 316 also measures the vibrations produced to accurately localize the vibrational center and to determine the cause of the vibrations' occurrence.

Interpretation of the vibration and kinematic data is a complicated task involving an in-depth understanding of data acquisition, training data sets, signal analysis, as well as the mechanical system characteristics. Vibrations generated through the interactions of implant components, bones, and/or soft tissues result from induced by driving force leading to a dynamic response. The driving force may be associated with knee-ligament instability, bone properties, and conditions. A normal intact knee joint 50 (FIG. 1) will have a distinct pattern of motion and vibrational characteristics. Once degeneration or damage occurs to the knee joint 50 (FIG. 1), both the kinematic and vibrational characteristics change. This altering, for each type of injury or degeneration, leads to distinct changes (or signature) that may be captured by the kinematic and vibration methods described herein.

FIGS. 28-34 illustrate a diagnostic system 320 configured to perform the method 316 in accordance with one embodiment of the present invention. The diagnostic system 320 includes the ID module 90 configured to diagnose soft tissue and bone injuries. For example, a first patient having a normal knee joint and a second patient having an anterior cruciate ligament deficit (“ACLD”) may exhibit a similar pattern of posterior femoral translation during progressive knee flexion; however, the first and second patients exhibit different axial rotation patterns of 30° of knee flexion. Accordingly, the ID module 90 includes three stages: (1) a first stage that involves data analysis, (2) a second stage that includes sending the data to a neural network for detecting an injury, and (3) a third stage that classifies or determines severity of a detected injury.

The first stage includes acquisition of kinematic feature vectors, using multiple physiological measurements taken from the patient while the patient moves the knee joint 50 (FIG. 1) through a range of motion. Exemplary measurements may include, without limitation, medial condyle anteroposterior (“MAP”) motion and lateral condyle anteroposterior (“LAP”) motion. The LAP motion pertains to the anterior-posterior (“AP”) distance of the medial and lateral condyle points 110, 114 (FIG. 1) relative to a tibia geometric center. Other exemplary measurements may include lateral shear interferometer (“LSI”) measurement of the distance between the lateral femoral condyle 114 (FIG. 1) and the lateral tibial plateau 321 (FIG. 3), and medial shear interferometer (“MSI”) measurement of the distance between the medial, femoral condyle 310 (FIG. 1) and the medial tibial plateau 321 (FIG. 3) which includes the superior/inferior (“S/I”) distance of the lateral and medial condoyle points 114, 110 (FIG. 1) to a tibial plane, as shown in FIGS. 30A-30C.

Feature vectors may also include the femoral position with respect to the tibia which is defined by three Euler angles 340, three translation components with the vibrational signal 342, and force data 344. Examples of these vectors are shown in FIGS. 31A-31C, respectively. FIG. 32 is a graphical representation 346 showing the average medial and lateral condyle positions during a deep knee bend activity for the second patient having ACLD. The feature vectors that are extracted from the kinematic and vibration analyses are output to the neural network 98 (FIG. 6) for determining the injury, as described in greater detail below.

FIG. 33 illustrates one embodiment of a neural network classifier 322 having multiple binary outputs 323a, 323b, 323c, 323d, i.e., each output is either a “1” or “0,” wherein the “1” corresponding to “yes” and the “0” corresponding to “no.” In this neural network classifier 322, each output 323a, 323b, 323c, 323d represents the response of the neural network 98 (FIG. 1) to a particular injury type. For example, one output 323b may represent the response for ACLD, wherein its state will be “1” if an ACL injury is detected, and “0” otherwise. Obviously, the neural network 98 (FIG. 1) and the classifier 322 may be significantly more or less sophisticated, depending on the underlying model of the joint in question.

FIG. 34 illustrates one embodiment of a construction 325 of the neural network 98 (FIG. 6). The construction 325 includes formulating a supervised classifier using a training set 324 of the kinematic and vibration data corresponding to a dataset 326 of normal and injured knee joints. The neural network 98 (FIG. 6) is trained with the training set 324 of vectors, wherein each vector consists of data (sound 328, kinematic 330, and force 332) collected from the knee joint 50 (FIG. 1).

Fluoroscopy data 333 may be used to calculate the kinematics. While fluoroscopy data 333 is highly accurate, it requires the patient to remain within the small working volume of the fluoroscope unit and subjects the patient to ionizing radiation for a prolonged period of time. For most dynamic activities where the joints are loaded, such as running, jumping, or other dynamic activities, fluoroscopy is an unacceptable alternative. Therefore, use of fluoroscopy data 333 is not required.

It should further be noted that electromyography (“EMG”) electrodes 337 (FIG. 6) may also be utilized as a data input for the computer 96 (FIG. 6) and the neural network 98 (FIG. 6). In this fashion, one or more EMG electrodes 337 (FIG. 6) are mounted to the surface of the skin proximate the muscles adjacent the knee joint 50 (FIG. 1) to monitor the electrical signal transmitted to the muscles in order to provide relevant data of a muscle injury or disorder.

Once the neural network 98 (FIG. 6) is trained, it may be used to classify new cases and categorize an injury type using these kinematic 330, vibration 328, and force 332 data. Those skilled in the art will readily understand that the types and classifications desired to be accommodated by the neural network 98 (FIG. 6) necessarily include training the neural network 98 (FIG. 6) on these very types of classifications. Exemplary types and classifications of injuries to mammalian knee joints include, without limitation, osteoarthritic conditions, soft tissue damage, and abnormal growths. Likewise, the neural network 98 (FIG. 6) needs to be trained to differentiate between and normal and abnormal knee conditions.

Referring again to FIG. 29, for a new patient 326, acquired vibrational, kinematic, and force features 328, 330, 332 the knee joint 50 (FIG. 1) are compiled and input as a testing set 327 to the trained neural network 334. The trained neural network 334 then diagnoses the condition of the knee joint 50 (FIG. 1), and returns one of the outputs 323a, 323b, 323c, 323d.

Although now shown, some embodiments of the method may be adapted so that the testing set 327 is acquired outside of a clinical setting. For example, a knee brace in accordance with an embodiment of the present invention may be worn by a patient for an extended period of time while performing normal activities. For example, the patient may wear a device incorporating components of at least one of the JKT module 86 (FIG. 6), the VA module 88 (FIG. 6), and the foot module 92 (FIG. 6) during activities that are not reproducible in the office (for example, weight lifting, racquet ball, etc.) and that elicit the pain or patient's symptoms. In some embodiments, the patient may turn the device on immediately prior to the activity and/or the patient may mark onset of the pain or symptoms when it occurs. This enables analysis of the data range from few seconds before the marked time to see what abnormal sounds or joint kinematic were occurring.

Data may be stored on a portable hard drive (or any other portable storage device) and then may be downloaded to exemplary systems for analysis. The data can be wirelessly transmitted and stored in a computer. It can also be stored with a miniature memory drive if field data is desired. If the occurrence of the pain is more random, some embodiments of the devices may continuously acquire data. Although, continuously monitoring devices may require a larger data storage capacity.

It is understood that while the exemplary embodiments have been described herein with respect to the knee joint 50 (FIG. 1), those skilled in the art will readily understand that the aforementioned embodiments may be easily adapted to other joints of the musculoskeletal system of a mammalian animal. For example, embodiments may be adapted for use on hips, ankles, toes, spines, shoulders, elbows, wrists, fingers, and temporomandibular joints.

While the present invention has been illustrated by a description of various embodiments, and while these embodiments have been described in some detail, they are not intended to restrict or in any way limit the scope of the disclosed invention. Additional advantages and modifications will readily appear to those skilled in the art. The various features of the present invention may be used alone or in any combination depending on the needs and preferences of the user. This has been a description of the present invention, along with methods of practicing the present invention as currently known.

Claims

1. A device for acquiring data and diagnosing a musculoskeletal injury, the device comprising:

a semi-flexible housing configured to be positioned proximate a portion of the musculoskeletal system of a patient;

at least one ultrasonic transducer operably coupled to the housing and configured to acquire an ultrasonic data indicative of a bone surface;

a positional localizer operably coupled to the housing at a select location relative to the at least one ultrasonic transducer, the positional localizer configured to track movement of the housing; and

a transmission system operably coupled to the housing and configured to transmit the ultrasonic data from the at least one ultrasonic transducer and the movement data from the positional localizer to a data analyzer for analysis and diagnosis.

2. The device of claim 1, wherein the device is a brace configured to surround the portion of the musculoskeletal system for diagnosis.

3. The device of claim 2, wherein the brace is a knee brace for diagnosing a knee injury, the knee brace further comprising:

a first ultrasonic transducer positioned proximate the distal femur; and

a second ultrasonic transducer positioned proximate the proximal tibia.

4. The device of claim 3, wherein the first ultrasonic transducer, the second ultrasonic transducer, or both is comprised of an individual transducer tracking unit.

5. The device of claim 3, wherein the first ultrasonic transducer, the second ultrasonic transducer, or both is comprised of an inter-transducer mechanical link unit.

6. The device of claim 3, wherein the first ultrasonic transducer, the second ultrasonic transducer, or both is comprised of a rotating transducer unit.

7. The device of claim 1, wherein the positional localizer is an optical sensor device, an inertial measurement unit device, an ultra-wide band sensor device, or a combination thereof.

8. The device of claim 1, further comprising:

a vibrational sensor operably coupled to the housing and configured to acquire a vibration signal generated during movement of the portion of the musculoskeletal system.

9. The device of claim 8, wherein the vibrational sensor comprises at least one accelerometer.

10. A method of diagnosing a musculoskeletal injury, the method comprising:

creating a 3D model of a portion of the musculoskeletal system of a patient;

acquiring a feature data with a sensor positioned proximate the portion of the musculoskeletal system while the portion is articulated;

comparing, with a neural network, the acquired feature data with a database of feature data, wherein the database of feature data includes a dataset representative of the musculoskeletal injury; and

returning a diagnosis based on the comparing.

11. The method of claim 10, further comprising:

positioning a sensor proximate the portion of the musculoskeletal system;

operating the sensor to acquire the feature data; and

transferring the acquired feature data to the neural network.

12. The method of claim 11, wherein the sensor is an ultrasound transducer and the feature data includes an ultrasonic signal indicative of a bone surface, the method further comprising:

tracking a position of the ultrasound transducer relative to the portion of the musculoskeletal system.

13. The method of claim 10, where creating a 3D model further comprises:

acquiring structural data indicative of a surface of a bone within the portion of the musculoskeletal system; and

morphing a general bone model in accordance with the structural data.

14. The method of claim 13, wherein the structural data includes an ultrasonic signal, a computerized tomography data, a fluoroscopy data, or a combination thereof.

15. The method of claim 10, wherein the feature data includes a vibrational data, a kinematic data, a contact force data, or a combination thereof.

16. The method of claim 15, wherein the feature data comprises the vibrational data and the kinematic data, the vibrational data being time-synchronized with the kinematic data.

17. The method of claim 10, wherein comparing the acquired feature data further comprises:

training the neural network with a plurality of datasets, wherein at least one of the plurality of datasets is the dataset representative of the musculoskeletal injury.

18. The method of claim 10, wherein the feature data includes a shear measurement, at least one Euler angle, a translational component, a force data, or a combination thereof.

19. The method of claim 10, further comprising:

displaying the returned diagnosis, the 3D model, the acquired feature data, or a combination thereof on a user interface.

20. A diagnostic system for diagnosing a musculoskeletal injury, the diagnostic system comprising:

a 3D model reconstruction module configured to acquire a structural data indicative of a bone surface within a portion of the musculoskeletal system of a patient and to construct a patient-specific model from the structural data;

a kinematics tracking module configured to acquire a movement data while the portion of the musculoskeletal system is articulated;

a vibroarthography module configured to acquire a vibration data generated during the articulation; and

an intelligent diagnosis module configured to receive and analyze the structural data, the movement data, and the vibration data and to determine an injury type from the analysis.

21. The diagnostic system of claim 20, wherein the 3D model reconstruction module further comprises:

an ultrasound transducer configured to acquire an ultrasonic signal indicative of the bone surface;

a position sensor having a select location relative to the ultrasound transducer, the position sensor configured to track movement of the portion of the musculoskeletal system; and

a statistical bone atlas comprising a plurality of bone models, wherein at least one of the plurality of bone models is morphed in accordance with the ultrasonic signal.

22. The diagnostic system of claim 20, wherein the kinematics tracking module further comprises:

a brace configured to be positioned proximate the portion of the musculoskeletal system;

at least one ultrasonic transducer operably coupled to the brace and configured to acquire an ultrasonic data indicative of the bone surface;

a positional localizer operably coupled to the brace at a select location relative to the at least one ultrasonic transducer, the positional localizer configured to track movement of the brace; and

a transmission system operably coupled to the brace and configured to transmit the ultrasonic data from the at least one ultrasonic transducer and the movement data from the positional localizer to the intelligent diagnosis module.

23. The diagnostic system of claim 20, wherein the vibroarthography module further comprises:

at least one vibrational sensor positioned proximate the portion of the musculoskeletal system; and

a transmission system configured to transmit the vibration data from the at least one vibrational sensor to the intelligent diagnosis module.

24. The diagnostic system of claim 20, wherein the intelligent diagnosis module further comprises:

a neural network configured to compare the movement data, the vibration data, or both to a database comprising of movement, vibrational, and injury data, wherein the database includes the movement data or the vibration data and an associated musculoskeletal injury type;

at least one transformation configured to transfer an acquired data to a virtual data; and

a statistical atlas comprising a plurality of bone models, wherein at least one of the plurality of bone models is morphed in accordance with the structural data to construct the patient-specific model.

25. The diagnostic system of claim 20, further comprising:

a contact force module configured to acquire a pressure data while the portion of the musculoskeletal system is articulated.

26. The diagnostic system of claim 25, wherein the contract for module comprises:

a shoe insole configured to be positioned on a foot of a patient;

a plurality of pressure sensors operably coupled to the shoe insole and arranged in a pattern; and

a transmission system operably coupled to the shoe insole and configured to transmit the pressure data from the plurality of pressure sensors to the intelligent diagnosis module.