Patent Application
20180020973
This application is related to prosthetic sockets for use on residual limbs of amputees. More specifically, the application is related to prosthetic sockets with sensors.
All publications and patent applications identified in this specification are hereby incorporated by reference to the same extent as if each such individual publication or patent application were specifically and individually indicated to be so incorporated by reference.
Prosthetic limbs for the upper and lower extremities typically include a residual limb socket, an alignment system, and a distal prosthetic component, such as a knee, foot, arm, or hand. The prosthetic socket is the interface between the user's body and the distal prosthetic components. To fulfill its role as an effective and practical interface, the socket needs to fit onto the residual limb and support it well, it needs to effectively suspend onto (or “hold onto”) the residual limb, and it needs to provide a stable and aligned support for the distal prosthetic components. All of these functions are important, but fitting the socket very well to the residual limb has been a long-standing challenge for prosthetic socket technologies.
Even with more modern techniques of socket construction, one continuing challenge with prosthetic sockets is that the volume of any patient's residual limb changes continually, for example with weight gain, recovery from injury, muscle loss or gain, amount of activity of the patient, etc. These continual volume changes pose a serious constraint on the ability of any fixed-form laminated plastic socket to provide a fit that is fully satisfactory over time. In many instances, it takes a significant amount of time to make a prosthetic socket for a patient. In those cases, the initial socket might not fit the patient well, even from the very beginning. Even when the socket does initially fit well, the patient's residual limb almost always undergoes size and shape changes over time, and the initial socket then no longer fits the patient. With most currently available sockets, whenever significant changes in the residual limb occur an entirely new socket must be made. Wearing an ill-fitting socket is generally extremely uncomfortable and can even do permanent damage to the residual limb.
Therefore, it would be desirable to have improved prosthetic sockets and methods for making and adjusting the fit of such sockets. It would be even more ideal to have a system for monitoring the fit of a socket on a patient, to enable the patient or a caregiver to adjust the socket as needed.
Prosthetic sockets act as an interface between the residual limb and more distal prosthetic components. As such, prosthetic sockets are critically important for in the overall functionality of limb prosthetic devices. As discussed above, for a prosthetic socket to function properly, it must fit extremely well, and the fit of the socket must be maintained through what are often very frequent and significant size and shape changes in the residual limb. Furthermore, the prosthetic socket must fit well during motion, for example with a lower limb socket during walking. This further complicates the prosthetic socket fit equation.
This application describes various embodiments of a prosthetic socket, a socket system, and methods for making and using the socket and the system, all of which involve one or more sensors on the socket for sensing various parameters, such as the forces placed on the prosthetic socket by the patient's residual limb and the socket. Sensed parameter data may be processed and analyzed and may be used by the patient, a caregiver and/or any other user to better understand the way the prosthetic socket fits and interacts with the patient's residual limb and in some cases to help modify the socket to improve the fit of the socket on the residual limb. One example of a sensed parameter is the force (or forces) placed on the socket by the residual limb while the patient is moving, stationary, etc.
In one aspect of the present disclosure, a system for profiling a distribution of forces applied to a prosthetic socket by a residual limb of a wearer of the socket may include a prosthetic socket, a sensor network and a processor. The prosthetic socket may include a proximal portion, a longitudinal portion including multiple struts, a distal portion including a distal base coupled with distal ends of the multiple struts, and an adjustment member coupled with the proximal portion, the longitudinal portion and/or the distal portion, configured to adjust the prosthetic socket to alter a force distribution profile within the prosthetic socket. The sensor network may include multiple sensors coupled with the prosthetic socket in a pattern defining multiple internal regions within the prosthetic socket. The processor is coupled with the sensor network (wirelessly or via a wired connection) and is configured to receive sensed data from the sensor network, divide the sensed data into groups corresponding to the multiple internal regions within the prosthetic socket, and process the sensed data to provide force distribution profile data corresponding to the force distribution profile.
In some embodiments, the processor is directly attached to the prosthetic socket. Alternatively, the processor may be separate from, and wirelessly coupled with, the prosthetic socket. In some embodiments, the processor is housed in a controller configured to allow the wearer of the prosthetic socket or another user to control at least one feature of the prosthetic socket. In one embodiment, the processor may include a microprocessor directly attached to the prosthetic socket and configured to receive the sensed data and wirelessly transmit the sensed data and an off-socket processor separate from the prosthetic socket and configured to receive the sensed data from the microprocessor, divide the sensed data into the groups corresponding to the multiple internal regions within the prosthetic socket, and process the sensed data to provide the force distribution profile data. In some embodiments, such a microprocessor may perform initial processing of the sensed data to provide processed sensed data before transmitting to the off-socket processor. In various embodiments, the off-socket processor may be positioned in a location such as but not limited to a computer application on a smart phone or other smart device, a tablet computer, a laptop computer, a desktop computer, a computer server or the cloud.
Any suitable sensor may be included in the sensor network, such as but not limited to a force sensor, a strain gauge, a Hall sensor, a flex sensor, a proximity sensor, a GPS, a flex sensor, a 3-axis accelerometer, a 3-axis gyroscope, and/or a 3-axis magnetometer. In some embodiments, the multiple internal regions include a proximal region corresponding to the proximal portion of the prosthetic socket, a longitudinal region corresponding to the longitudinal portion of the prosthetic socket, and a distal region corresponding to the distal portion of the prosthetic socket. In some of these embodiments, each of the proximal region, the longitudinal region and the distal region is further divided into four sub-regions: a mediolateral region; a medioposterior region; a lateroanterior region; and a lateroposterior region. In some embodiments, the multiple internal regions further include an ischial seat region. In some embodiments, the longitudinal region is further divided into an upper longitudinal region and a lower longitudinal region.
In some embodiments, the proximal portion of the prosthetic socket may include a brim member coupled with proximal ends of at least some of the struts. In some embodiments, the proximal portion of the prosthetic socket may include an ischial seat member. Optionally, the force distribution profile data may include multiple percentages of force applied by the residual limb to the prosthetic socket over the multiple internal regions. The force distribution profile may also describe force distribution through each of the multiple struts. In some embodiments, the force distribution profile describes force distribution in a central path through a distal end of the residual limb and a peripheral path through the longitudinal portion of prosthetic socket.
In some embodiments, the processor may be further configured to compare the force distribution profile data with a desired force distribution profile stored in a computer memory of the system and provide the wearer or another user with comparison data. For example, such comparison data may include an alert when the force distribution profile is outside of a predetermined range of desired force distribution profile data. In some embodiments, the adjustment member may be an adjustable hinge configured to be fixed at a desired angle. Alternatively, the adjustment member may be an adjustable height ischial seat member mounted on a proximal end of one of the multiple struts. In some embodiments, the adjustment member may include at least one tensioning band coupled with the multiple struts and at least one tension adjustment member attached with each of the at least one tensioning bands. In other embodiments, the adjustment member may include a motorized adjustable closure mechanism configured receive a command signal from the processor and automatically adjust the prosthetic socket based on the command signal. Optionally, the adjustable motorized closure mechanism may be further configured to maintain a tension in the prosthetic socket within a predetermined tension range. In some embodiments, the adjustment member includes a hinge mechanism configured receive a command signal from the control unit and automatically adjust the prosthetic socket based on the command signal.
Optionally, the sensor network may further include at least one off-socket sensor configured to be attached to the wearer of the prosthetic socket at a location separate from the prosthetic socket. The system may also optionally include a control unit coupled with the prosthetic socket, where the processor is housed within the control unit. In an alternative embodiment, the system may include a control unit that is physically separate from the prosthetic socket, where the processor is housed within the control unit and the sensors are wirelessly coupled with the control unit.
In another aspect of the present application, a method for generating force distribution profile data for a prosthetic socket on a residual limb of a wearer of the socket may first involve sensing forces applied to multiple predefined regions of an inner surface of the prosthetic socket by the residual limb, using a sensor network attached to the prosthetic socket. The sensor network generally includes multiple sensors disposed in the predefined regions. Next, the method involves transmitting sensed data from the sensor network to a processor coupled with the sensor network and processing the sensed data with the processor. The processing step includes grouping the sensed data into multiple groups corresponding to the predefined regions and generating force distribution profile data for the prosthetic socket, based on the sensed data and the multiple groups.
Optionally, the method may further involve providing the force distribution profile data to a user. Additionally or alternatively, the method may also include automatically adjusting the prosthetic socket, based on the force distribution profile data. In such an embodiment, the method may optionally also involve comparing the force distribution profile data to a desired force distribution profile, and the step of automatically adjusting the prosthetic socket is based at least in part on the comparing step. In some embodiments, the method also includes sensing an acceleration of at least a portion of the prosthetic socket, using the sensor network, and generating an acceleration distribution profile for the prosthetic socket, based on the sensed acceleration. Alternatively, the method may involve sensing a position of at least a portion of the prosthetic socket, using the sensor network, and generating a position distribution profile for the prosthetic socket, based on the sensed position.
In some embodiments, the prosthetic socket may include three or more struts, and generating the force distribution profile may involve comparing amounts of force in each of the struts. In some embodiments, the prosthetic socket is a transfemoral prosthetic socket including four struts, where one of the struts is a medial-posterior strut, and generating the force distribution profile involves comparing an amount of force delivered through the medial-posterior strut with amounts of forced delivered through the other three of the four struts.
In some embodiments, the prosthetic socket includes a microprocessor, and the transmitting step involves transmitting the sensed data from the sensor network to the microprocessor and transmitting the sensed data from the microprocessor to the processor, where the processor is located separately from the prosthetic socket. Such a method may optionally further involve conducting initial processing of the sensed data at the microprocessor before transmitting to the processor. The processor may be located off of the prosthetic socket and be coupled wirelessly with the sensors of the sensor network, according to some embodiments. In other embodiments, the processor may be directly attached to the socket and may be coupled wirelessly or via wire(s) with the sensors.
In some embodiments, transmitting the sensed data may involve transmitting a location identifier from each of the multiple sensors. Some embodiments of the method may also involve displaying the force distribution profile data on a display device. For example, the display device may be a controller that is separate from the prosthetic socket. The processor may be housed in the controller, for example. The method may also include providing an alert on the controller when the force distribution data falls at least partially outside of a predetermined range of desired force distribution data.
In some embodiments, the method also includes adjusting tension in the prosthetic socket, using a motorized tensioning mechanism attached to the socket, based at least in part on the force distribution profile data. In some embodiments, the tension is adjusted automatically. Some embodiments may involve automatically adjusting at least one characteristic of the prosthetic socket to adjust the force distribution profile data toward a desired force distribution profile. In some embodiments, adjusting the force distribution profile data may involve moving a force distribution profile longitudinally within a length of the prosthetic socket. In some embodiments, adjusting the force distribution profile data involves moving a force distribution profile within a cross sectional anterior-posterior/lateral-medial grid within the prosthetic socket. In some embodiments, adjusting the force distribution profile data involves moving a force distribution profile comprises actuating a hinge mechanism within the prosthetic socket.
Examples of force distribution profile data include but are not limited to a distribution of forces impinging on the prosthetic socket in relation to a central longitudinal axis of the socket, an absolute level of force applied by a distal end of the residual limb to the prosthetic socket, and a relative fraction of a total force applied by a distal end of the residual limb to the prosthetic socket. In some embodiments, the method may also include testing the prosthetic socket on the residual limb and determining a desired force distribution profile for the socket and the residual limb. Testing the prosthetic socket, for example, may include placing the prosthetic socket on the residual limb and making initial adjustments to one or more mechanical features on the prosthetic socket.
In any given embodiment, the method may involve repeating the sensing, transmitting and processing steps as many times as desired. For example, the steps may be repeated continuously over a period of time or at set time intervals.
In another aspect of the disclosure, a non-transitory computer readable medium for use on a computer system may contain computer-executable programming instructions for performing a method as described immediately above. In various embodiments, the method may include any or all of the features and/or steps described above.
These and other aspects and embodiments will be described in more detail below, in reference to the attached drawing figures.
As mentioned above, this application describes various embodiments of a prosthetic socket, a socket system, and methods for making and using the socket and the system, all of which involve one or more sensors on the socket for sensing various parameters. For example, two types of prosthetic socket sensors are force and motion sensors. Sensed force and motion data may be used to evaluate a patient's experience and clinical outcome with the socket and in some embodiments to make adjustments to the socket to improve fit. Adjustments may be made automatically, for example using microprocessors in communication with actuators, manually by the patient or prosthetist, or a combination of automatically and manually. In addition to sensing forces and motion, a host of other types of sensors may be included in a prosthetic socket, according to various alternative embodiments. For example, some sockets may sense temperature or humidity within a prosthetic socket, pulse, blood pressure, electrocardiogram (ECG or EKG) signals and/or electromyographic (EMG) signals.
In this application, sensor-enabled prosthetic socket embodiments and the hardware and software of the system used to support and interact with the socket may be referred to generally as embodiments of a “smart socket.” The totality of data reported by prosthetic socket sensors at a specific time on smart socket embodiments may be referred to as the physical “state” of the socket. The smart socket state may also be time stamped or tagged in conjunction with qualitative data, clinician data, or other data. The socket state as well as qualitative data, clinician data, or other data may be referred to as “total input data.”
Many, if not all, of the embodiments of prosthetic socket systems described herein may be incorporated into (or adapted for use in) any of a number of types of prosthetic sockets. Examples of the types of prosthetic sockets that may be especially amenable to the socket sensors and sensor systems described herein are the modular, strut-based prosthetic sockets that have been described previously by the assignee of the present application. Examples of such prosthetic sockets are described in detail in U.S. Pat. Nos. 8,978,224, 9,044,349 and 9,468,542, and U.S. Patent Application Pub. Nos. 2014/0277584 and US 2016/0058584, all of which are incorporated herein by reference. Other examples will be discussed below, and these examples are not meant to be limiting.
The above-referenced publications describe a prosthetic socket assembled from modular components that include a distal base, multiple longitudinal struts indirectly or directly connected to the distal base, and one or more brim elements, each of which is connected to at least one strut. In some prosthetic socket embodiments, a strut connector element intervenes between the distal end of a strut and a connection site on the distal base, or the struts connect indirectly. In addition to these basic structural components, embodiments of the prosthetic socket may also include a distal cup, one or more encircling bands and/or tensioning elements or closure mechanisms that are applied around the struts and or the brim elements, as well as soft good elements, such as sleeves that are placed over the struts. Ancillary components associated with the prosthetic socket may include socket liners that protect the residual limb and can remove accumulated moisture. Sensors can be integrated into the smart socket in various different ways, between or around various different parts, and within various different parts of the modular assembly.
Examples of sensors that may be used in various embodiments of the prosthetic sockets described herein include but are not limited to sensors that detect and quantify force, torque, load, and/or pressure in many different ways, including in-line load cells, pancake load cells, rotary shaft torque sensors, and flush threaded pressure sensors. Other examples of sensors include motion sensors (such as accelerometers, inclinometers, gyroscopic sensors), magnetic field sensors, global positioning sensors, altimeters, thermometers, moisture sensors, photo sensors, and cameras. Timing and/or clock functions may also be integrated into the sensors. In some embodiments, multiple positioning or gyroscopic sensors can be configured as RF transceivers that communicate with a single receiver, thereby enabling triangulation to add another dimension to positioning of the patient and/or the socket in space.
Other vital sign or biometric sensing devices may be included in various embodiments, such as a heart rate sensor, blood flow sensors, blood oxygen sensors, blood pressure sensors, EMG sensors, and other heath monitoring sensors integrated into modular components to monitor limb health and/or the general health of the patient. In some instances, some or all of the sensing devices used in a given prosthetic socket system may be off-the-shelf units and/or not substantially integrated within the prosthetic socket itself. Such sensors, however, may be included in the prosthetic socket system as a whole, and data acquired by these sensors may be integrated downstream in a data processing and analysis function, to build out a more complete clinical picture. The preceding enumerated sensors are intended merely as non-limiting examples, and various embodiments may include any sensor that can deliver clinically useful information regarding the status of the patient, the socket and/or the environment within or near the socket.
Other sensed data may come from third party or associated sources, such as distal components, a smart prosthetic gel liner, a smart prosthetic liner or sock, a third party smart phone application (such as a dietary application or sleep application), a third party data collecting device (such as a smart watch or fitness tracker) and/or the like. These third party or associated sources may also provide electrical power or other resources. For example, a smart socket may couple with the microprocessor prosthetic knee or power knee to collect data, share data, and share electrical power. This may improve accuracy and quantity of data, may allow for new automated response mechanisms, such as data from the smart socket regarding how the knee or other distal components should respond to the environment, and reduce the number of devices the user needs to plug in or replace batteries for.
The modular aspect of some of the sensor-enable prosthetic socket embodiments described herein refers to the fact that components of these sockets are available in a range of sizes and shapes and can be assembled into a socket and swapped out for other parts as necessary, while still fitting together via common connection elements. Of particular significance in embodiments of sensor-enabled sockets, as described herein, is the fact that each connection site is a site across which force (e.g., pressure, tension, torque) may be transmitted when the socket is worn by a patient engaged in the activities of daily living. Tensioning or closure elements that provide circumferential adjustments typically include two connection portions, such as a buckle, ladder-lock or loop-lock arrangement connection, or any suitable reversible or adjustable tensioning or closure element connection mechanism.
Accordingly, in some embodiments, force sensors may be advantageously positioned at these sites where prosthetic socket components connect to another component, or within a two-part component, such as a tensioning element. Force sensors and other sensors may also (or alternatively) be located at any advantageous position on the hard structure of the prosthetic socket, or at any suitable position within soft goods or fabric portions of the socket that are made to couple with any type of prosthetic socket. The sensors may also be integrated into many different types of prosthetic sockets, such as vacuum formed thermoset or thermoplastic prosthetic sockets that are non-modular, modular prosthetic sockets that are fabricated from one or more inventoried components, thermoset or thermoplastic modular prosthetic sockets, or prosthetic sockets made of other materials, such as elastomers, urethane, alloys and/or phase changing alloys. Non-modular laminated prosthetic sockets may require a dummy or spacer element that is integrated into the mold during fabrication then removed to allow for integration of the sensors. Motion sensors may be advantageously positioned at relatively distal sites within the prosthetic socket, inasmuch as movement tends to increase with increasing distance from the patient's hip.
Data that are collected from the sensors may be relayed to a remote or onboard microprocessor unit for immediate or future use and/or stored or saved remotely or onboard the prosthetic socket. Remote devices, such as a computer at the patient's home or at a clinic, may be configured to receive data. Mobile phones or dedicated mobile receivers or transceivers may be carried by the patient, and such devices may further transmit the data to a remote server or a local computer.
In some embodiments, sensed prosthetic socket data may be exploited to make immediate or nearly immediate mechanical adjustments or other forms of adjustment, such as changes to a pneumatic bladder, temperature change of the prosthetic socket, change to a phase changing material such as a phase changing alloy or phase changing fluid, and/or changes with thermoformable materials within the socket. Some prosthetic socket embodiments have sites where mechanical adjustments can be made that affect the sizing or shape-based fit of a socket to the residual limb of a patient. These adjustments may be made manually, either by the patient or by a prosthetist or other caregiver, for example. In other embodiments, the adjustments may alternatively or additionally be made automatically, when sensor data is processed by an associated microprocessor that is, in turn, enabled to appropriately activate an actuator to affect a responsive mechanical adjustment.
Sensed data or the total input data may also immediately or nearly immediately be used for notifications, predictive analysis, and prescriptive information. For example, sensor data can measure the amount of pressure inside regional aspects of the socket, a processor may analyze the sensor data to determine that the pressure in the proximal-medial aspect of the user's transfemoral smart socket is beyond the maximum threshold based on qualitative data, such as previous comfort scores and targets the user has set for their daily use. A notification can immediately be sent to the user's smart phone, with prescribed or recommended manual adjustment.
In some embodiments, sensed data may be used on a longer-range time line but may still result in one or more mechanical adjustments, such as those just described. Longer-range time refers to a data collection period of more than a single day, typically a duration of a week or more. The longer-range time line may allow for sensor data that is more highly resolved and dependable, and it may also allow the processor to analyze patterns in the data. Data and patterns may analyzed with software, mathematical equations, and algorithms built from clinical experts and users who are knowledgeable on extracting clinically relevant metrics from the total input data. Mechanical adjustments made in response to these types of accumulated data are typically made be a prosthetist, but a patient may participate as well.
As mentioned above, in some embodiments, a prosthetic socket may automatically adjust in response to sensed socket data, via one or more mechanical adjustment actuators on the prosthetic socket. Microprocessor-enabled automatic control of adjustable points in the prosthetic structure may occur in sites, by way of example, where angles or tensioning elements occur. U.S. Pat. No. 8,978,224, incorporated herein, describes several examples, including adjusting the angle of struts with respect to the base, adjusting the length of a telescopically enabled strut, adjusting the circumferential tension applied either to struts or brim elements, and automatically adjusting air or hydraulic pressure within a pneumatic or hydraulic bladder or camber system.
One type of automatic adjustment actuator that may be used in a sensor-enabled prosthetic socket is a sensor-controlled servomotor or stepper motor integrated into one or more tensioning or closure elements of the socket (or multiples of same). For example, servomotors may be disposed at two sites on the socket, across a joining element that connects two portions of a tensioning element. The effect of appropriately configured servomotors is to loosen or tighten tensioning elements, depending on sensed data and predetermined responses according to where the data land within predetermined ranges.
Another example is sensor-controlled servomotors that adjust the take-off angle of struts with respect to the distal base. The distal ends of struts attach to the distal base, and a separate intervening strut connector may be present in some embodiments. The takeoff angle affects the volume encompassed by the socket, particularly in the distal region of the socket. Tension or force sensors disposed within the socket can sense outward pressure that the residual limb is directing against the struts. The effect of the appropriately configured servomotors, by way of adjusting the takeoff angle, is to expand or contract the volume enclosed by the struts, depending on sensed data and predetermined responses according to where the data land within predetermined ranges.
Another example of an automatic adjustment mechanism for a prosthetic socket is inflatable and deflatable hydraulic or pneumatic bladders, which may be positioned within the smart socket along with an expulsion valve and a small compressor or pump, such as a mini-diaphragm air pump. For example, fabric strut and polyurethane bladders may be positioned on or along the struts of a prosthetic socket or between struts or against a socket liner. Sensors can be positioned on a strut sleeve, such that they are subjected to pressure that exists between the socket structure and the residual limb. Sensors may also be integrated into pneumatic bladders that are between struts within the brim or within cross-connectors, other textile components, the distal base, the distal cup, or other aspects of the socket. These sensors within the pneumatic or hydraulic bladders, which may communicate with a microprocessor, which may be operatively coupled to a bladder pump and vent or valve, can monitor pressure on the limb and respond in real time by inflating or venting, depending on sensed data and predefined commands. For example, the system may operate to maintain pressures/forces within the prosthetic socket within predefined ranges. Any type of bladder(s) may be used in such embodiments—e.g., pneumatic, air, fluid, gel or the like. Air or fluid bladders may also include baffles or geometric chambers designed to provide a specific pattern or motion or direction of force and pressure distribution when inflated or deflated. This direction-specific capability within the bladder can be designed to provide increased biomechanical control or can selectively increase inflation or deflation in regions where volume change typically applies. Any type of sensor-enabled bladder(s) described herein may be referred to as a “smart bladder” or “smart bladder system”.
In some embodiments, bladders may have force sensors integrated into them, such that the bladder can measure forces directly against the residual limb. A desired pressure range may be set by the user or clinician, for example via a smart phone application and smart socket microprocessor, and may be used to inflate or deflate as needed to maintain a certain amount of force or pressure. In some embodiments, the amount of force or pressure may be deliberately changed over time. Automated adjustment, in these bladder-based embodiments, would also increase or decrease the volume within the socket and may thereby automatically adjust for volume change of the residual limb. Inflation and deflation may also be used in cases where volume is consistent but the desired amount of pressure within the socket differs per amount of control that is needed per activity. For example, when a user sits down, the pressure may decreased, and when the user runs, the pressure may be increased, to provide more force needed to control the prosthesis while running.
Any type of smart bladder(s) described may also be used in such embodiments to maintain desired amounts of pressure per specific region of the socket, as set by the user or clinician, and/or intentional variation or oscillation of pressure per time spent using the prosthesis.
One example of the above smart bladder system is that the user may use a user interface application on a smart phone to establish that they want the pressure within the proximal-medial aspect of the socket and distal-lateral aspect of the socket to be maintained at an average pressure of 20 pounds, while the proximal-lateral aspect of the socket and distal-medial aspect of the socket are to be maintained at an average pressure of 10 pounds. This pressure would be automatically maintained as the user experiences changes in volume or changes their activity, and software is used to analyze or adjust pressure thresholds, average pressures, and/or peak pressure variance. A user may also use the described smart bladder system to reduce or increase pressure per activity. For example, the user may set the system to relax pressure to zero when they are sitting for more than five seconds. The smart socket would know the user is sitting, because data processing software includes pattern recognition of sensors that correspond with sitting. Settings may also be set for pounds per square inch (psi) within the bladder as opposed to pressure exerted onto the residual limb by the bladder.
Sensors that would enable these options in setting would be pressure sensors built within the layers of the bladder walls and/or pressure transducers built into the bladder system at the valve or at another location to test the pressure within the bladder. For some users, 5 pounds of pressure or 3 psi may be their high pressure setting, whereas others may select pressures of 20 pounds or 15 psi or greater for their high setting. For some users, 2 pounds of pressure or 2 psi may be their low pressure setting where as others may select for pressures of 5 pounds or 5 psi or lower for their low setting. These are merely examples, however, and any suitable combination of pressure settings may be used/selected in a given embodiment.
Another example of the above smart bladder system is that the user may use their user interface application on their smart phone to establish that they want the pressure within their smart bladder system to oscillate between 0 and 20 pounds 4 times at regular intervals throughout the day or upon request or allowance after notification via smart phone application. This oscillation or variation may be used to promote blood flow, other bodily fluid circulation, and/or other comfort of biological benefit.
Similar to the way these smart bladder systems can be used within a smart socket described herein, under-actuated mechanical joint systems may also be used to biomechanically control the residual limb, add comfort, and/or accommodate volume change. An under-actuated mechanical joint system is a system that provides minimal actuation or automation to control motion or response across multiple joints. For example, one cable may pass on the outside of a series of joints, while another cable may pass on the inside of multiple joints, and a motor or motors can actuate the motion of those cables to control the motion of multiple joints at once.
Another type of an automatic adjustment mechanism for a prosthetic socket is artificial muscle technology that may be used for automated response to sensed data or the total input data. This technology may be used in conjunction with the smart bladder systems described above or in other ways, such as shortening or lengthening a closure or biomechanical control mechanism.
Another type of automatic adjustment is height or length adjustment of one or more elements within the prosthetic socket, such as an automated height adjustment of the patellar tendon area of a transtibial prosthetic socket or the ischial seat area of a transfemoral socket. Some prosthetic socket embodiments have an ischial seat that is positioned at a medial and posterior aspect of the socket. The ischial seat engages the ischium of the patient and can bear weight that would otherwise be born by the residual limb, particularly the distal end of the residual limb. To bear weight effectively, the height of the ischial seat from the distal base of the socket should be properly adjusted. Small differences in height can have a large effect on the patient's comfort. For these reasons, it is advantageous for the ischial seat height to be adjustable. Pressure absorbed through the ischial seat is thus a useful indicator. A pressure sensor positioned such that it senses pressure through the ischial seat, and informed by a microprocessor, can be configured to drive movement of a linear actuator that adjusts the height to the ischial seat. Bladders as described above may also be used for this adjustment. Pneumatic or hydraulic cylinders may also be used for this type of adjustment and may also include a given range of shock absorption within the system so that the height may vary within a range that is predetermined by the user or clinician. For example, a user may set their ischial seat to maintain an average pressure of 30 pounds and to allow movement of plus or minus 5 mm from the selected baseline height or length. Adjustments may be made to accommodate for volume change or for changes in activity. For example, when a user sits, the ischial seat may automatically lower when it is not needed. Alternatively, when a user is walking and the pressure on their distal end gets past their self-selected socket threshold and, due to previous total input data, it is determined that raising the seat is more effective and comfortable for the user than tightening the closure system or tensioner, the seat can be automatically raised to reduce pressure on the bottom.
As mentioned above, any of a wide variety of sensors and any suitable combination of sensors may be incorporated into and/or used in conjunction with a prosthetic socket in providing a smart socket system as described in this application. In addition to using sensor-derived data for automatic adjustments of the socket, examples of which were just described, the data may also be used to guide manual adjustments of the socket and/or for general information about the patient's interaction with the socket and may include any of the above describe mechanisms or adjustment methods in a manual fashion. The general information may be used for any suitable purpose, such as in guiding the care of that patient, in developing clinical studies involving multiple patients, in collecting and analyzing data for multiple patients to help improve prosthetic socket manufacturing and/or use, or the like. What follows is a description of non-limiting examples of the types of clinically valuable information that may be derived from sensors incorporated into or used in conjunction with prosthetic sockets.
Pedometer devices may be used to track step count, the duration and times during the day when socket is worn, and the time it is not worn. Metrics of walking or running, such as speed and distance, may be tracked. Similarly, sedentary or standing time may be tracked. GPS location and altimeters may be used to capture details of location, distance, and terrain covered during walking or riding in a vehicle. Pattern recognition algorithms may be used to acquire data, characterize individual routines, such as donning or doffing the socket, and allow these routine activities to be tracked specifically. Pattern recognition may also be combined with gait analysis, so that gait features can be identified from data alone. Moisture levels within the space between the socket and a liner and/or in the space between the liner and the residual limb can be tracked with moisture sensors. Pressure in critical areas within the socket, as well as overall tightness of the socket, can be tracked. In a closely related function, overall volume of the residual limb may be tracked. In particular embodiments, blood pressure and pulse rate may be tracked. Photosensors or cameras can monitor skin color changes in critical or sensitive areas of the residual limb. Notifications, alerts, alarms, or instructions can be associated with any sensor (e.g., pressure, tension, moisture, temperature), with a range of values defined as normal and a range (or upper and/or lower limits, for example) defined as concerning. Changes in inductance may be used to track changes in circumference of the residual limb, and circumferences throughout the residual limb may be used to mathematically estimate the overall volume of the residual limb. These are merely examples of some of the parameters that may be sensed and measured with various embodiments of the sensor-enabled prosthetic socket systems described herein.
In some embodiments, data of the type described above may be used for making quick adjustments to the prosthetic socket and/or for creating awareness of a clinical issue immediately. The data may also be used by a prosthetist or healthcare provider who is following the patient over a longer period of time, and the longer-term data may also be used by the prosthetist or other caregiver to make socket adjustments, replace socket components, provide advice to the patient and/or the like. For example, the prosthetist may use the data to advise the patient regarding daily routines, patterns of activity, or levels of tension or pressure in the socket that may be optimal all over the socket and/or in specific regions of the socket.
In some embodiments, the sensor-enabled prosthetic socket system may include a user interface and data processing module or software, such as an application that may be used on a cell phone or mobile data capture device. The processing module may receive and transmit sensed data from the socket and in some embodiments may have broader uses as well. For example, the application can enable direct messaging between patient and prosthetist, or between a prosthetist and a researcher, or between the patient and the company that manufactures, markets, or distributes the prosthetic socket, or between the patient and patient interest or advocacy organizations. An application may have a questionnaire that can be filled out daily, or periodically, that captures the patient's subjective and qualitative experience. Such data can be usefully associated with the sensed data for cause and effect analysis by a prosthetist or researcher, or used by the software itself to identify patterns or adjust parameters as the increased data allows for improved learning about user needs and requirements. A software application may also provide connection directly to social media or to a private user group, thus providing peer and social support to the patient, in addition to more strictly clinical communications.
One example of a way in which sensed data from a prosthetic socket may be used relates to a scoring system to evaluate clinical outcomes for patients fitted with prosthetic sockets and such outcomes mapped against data sets obtained from patients. Such a system may evaluate the following criteria, for example:
(1) Comfort: For example, the 10-point socket comfort scoring system of Hanspal may be used (Disabil Rehabil 25(22):1278-80, 2003).
(2) Daily usage: For example, an accelerometer or pedometer may be used, and the data points may be time used per day and distance traveled or step count per day.
(3) Functionality: For example, Stevens, et al., (The Academy Today 5(1), February 2009) provides a comprehensive list of clinical relevant outcome measurements.
(4) Efficiency: This outcome relates to the efficiency of satisfactorily situating a patient in the socket with respect to the patient's time and effort, and the efficiency with respect to healthcare costs. This metric may be determined by tracking the number of clinical visits required from initial evaluation for a new socket to final socket delivery and also follow-up visits requested by a patient over a three-month period after initial socket delivery for socket related issues.
(5) Mobilization: This metric refers to the duration of time between (a) a patient's first evaluation for a prosthetic socket and the clinical decision to proceed with a new socket, and (b) delivery of a satisfactory socket to the patient for the initiation of ambulatory socket use. As a basic metric, mobilization can be expressed as the number of days from evaluation to ambulation. It relates primarily to operational efficiency of the clinic and turn-around factors specific to the socket. The metric can also be enriched by inclusion of other clinical factors or with details associated specifically with the initial trauma or limb condition, the level of amputation, etc.
A number of different assessments of a prosthetic socket and its interaction with the residual limb may be made, according to various embodiments. Examples include the following. (1) Changing residual limb volume, during the day or over the course of several days, reflecting changes in fluid pressure in the body. Compensating or correction intervention may include changes in tension of a tensioning belt (or belts) on the socket. (2) Changing residual limb volume over the longer-term volume, such as weeks or months, reflecting overall muscle and bone volume. Compensating or correction intervention may include changes in placement of a tensioning belt (or belts) on the socket. (3) Change in overall fit of the prosthetic socket: change in relative weight on pressure sensors (seat-strut-distal loading goes from 70%-20%-10% to 55%-30%-15%). (4) Activity level, for example as monitored by a motion sensor. Changes in activity level, per cumulative motion data, can be analyzed in conjunction with various measures of prosthetic socket “fit.” (5) Gait change, for example as monitored by a motion sensor. Example: change in average lateral motion during a step.
These preceding variables and the associated tracking data may be analyzed by a prosthetist that is following the patient. In response to the data, a prosthetist has a range of options, including making minor mechanical adjustments, changing ranges and responses that are automated, replacing the prosthetic socket or particular components, etc. Further, in some instances, patients have impaired cognition, perception, or communication, in which case sensor-based data can be particularly valuable.
In one embodiment, tension sensors are arranged at the junction of two connecting portions of a tensioning or closure element, such as a strap, belt, or webbing. Data from such tensioning sensors may be processed by a processor of the system to provide a profile of residual limb volume as it fluctuates during the day or as it changes gradually over weeks and months. Alternatively, change in inductance measurements can also be used for circumferential and volume measurements with a high degree of accuracy. Such an embodiment may include one or more tension sensors, such sensor(s) enabled or configured with features such as any of the options listed below.
(1) Feedback and sensing module (the device) may either be integrated into or in-line with a belt or webbing material that is typically under tension for use in prosthetics.
(2) Device may have onboard processing and feedback capabilities, wireless communication capabilities, and is battery-powered.
(3) Device may include a combination of sensors integrated into the belt or webbing module.
(4) Device may use raw sensor values to gather metrics related to wear time of prosthesis.
(5) Device may use raw sensor values to gather metrics related to activity level of a patient wearing the prosthetic device. Motion sensing of activity to be determined used to quantify activity levels.
(6) Device may use raw sensor values to gather metrics related to tightness (tensile force) present in belt or webbing of prosthesis.
(7) Device may use raw sensor values to gather metrics related to spatial orientation of prosthesis.
(8) Raw sensor values collected by device may be transformed into metrics that are useful for patients and care providers by way a series of custom algorithms developed for each prosthetic application.
(9) Device can deliver real time feedback to users via wireless streaming live data to a smart phone or computer.
(10) Device can operate in two modes of operation: (1) continuous data collection with non-real-time wireless streaming to a smart phone or computer (used to collect patient data over an approximate week time frame), or (2) continuous data collection with live wireless streaming to smart phone or computer (used for real testing and data collection experimentation.
(11) Device has modular belt attachment to swap out belt or interchange module with various prosthetic devices.
(12) Software (web and mobile app) that can display long-term trends, averages, and provides feedback to patients and/or care providers based up the data collection from the device. Data and relevant information is based on custom algorithms above stated prosthetic algorithms.
(13) Device may have on board general-purpose input/output ports to power and collect data from sensors external to the device casing. These include but are not limited to force, temperature, pressure, and humidity.
(14) The device may collect data from the above stated external sensors, store the data on-board, and transmit the data via a wireless data connection to a smart phone or computer.
Within the prosthetic socket, one or more of the following measurements are enabled by one or more sensors integrated into the prosthetic device; moisture, forces including pressure and tension, strain, shear, position of the socket and residual limb, geographic location of the user, relative movement/motion including changes in vertical position, lateral, and, horizontal position of the residual limb, user, or socket. Sensors within or adjacent to the socket or on the residual limb may also enable measurements of; residual limb volume, changes in residual limb volume, skin integrity or skin injury of the residual limb or contralateral limb, muscle or connective tissue integrity of the residual limb or contralateral limb, muscle or connective tissue activity of the residual limb or contralateral limb, as well as biometric measurements such as temperature, blood flow, heart rate, and blood glucose level/blood sugar.
One embodiment may include a modular addition that covers the inside and/or outside a prosthetic socket with instrumented textiles capable of sensing force in discrete regions. The textile force sensors are composed from a thin flexible piezoresistive film and conductive textile leads. Additionally, a 9 axis IMU is fastened to the modular addition. The array of force sensors and IMU are connected to a microcontroller for onboard data processing. This microcontroller parses the raw data from the previously mentioned sensors to a simplified list of gait metrics. This data is stored on onboard electronic flash memory, before being uploaded to the phone via Bluetooth Low Energy communication.
The mobile application acts as both a data transfer and user interaction tool. The phone communicates with the socket-side microcontroller/data transmission unit/battery via Bluetooth Low Energy, receiving the calculated gait metric data and validating their successful transfer. The mobile app then uploads the data to the cloud via cellular network or Wi-Fi.
Independent of the socket, the mobile app allows the user to interact with their data and access other social features. The phone retrieves past data from the cloud on user request. The amputee can then view their past data through visual display elements. These elements may include highlights of excellent or poor gait performance or activity, charts tracking gait progress and socket fit changes across time, comparison to self-determined goals, or others. Amputees may also upload other data manually, such as photos of painful areas or injuries, socket comfort score questionnaires, or abnormalities in socket function. The mobile app also contains a variety of social features for amputees to interact with each other, which may include chat capabilities with other amputees, activity comparisons and challenges with friends, tutorials and educational videos, or other social features.
The back-end server hosts a secure database that stores patient data. This server receives incoming data from the mobile app, both manually uploaded and transmitted from the socket. This server runs additional algorithms to further identify useful metrics from the collected data. The total input data is used across different users to establish standards or normality of data and then the individual's data can be compared to that of established standards to identify deviations or potential issues that may lead to injury or sub-optimal performance. The user entered targets as well as qualitative and quantitative data collected into the mobile application is also compared to the current socket state in order to identify any deviations or trends that may be lead to a targeted optimal state or to pain. The mobile application then sends data back to users on the web or mobile app for viewing. The server stores patient, prosthetist, and account info in addition to the collected sensor and app data.
The web-app has similar functionality to the mobile app for amputees, while allowing different permissions and functions for different types of users. Amputees can still retrieve data from the cloud, view their data history, manually input data, and access the various social features found on the mobile app. Other parties using the mobile-app will have access to different data and features, depending on their roles: e.g. prosthetists may access their patients' data, communicate securely with patients, communicate with other healthcare team members, and contact paying parties.
Qualitative data collected is collected through a smart phone application including; self-reported overall socket comfort score/socket comfort, regionally specific self-reported socket comfort, self-reported socket control or biomechanical control of the their residual limb within the prosthetic socket, suspension rating or how intimately the prosthesis stays on their residual limb, how tight or snug the socket feels around their residual limb, negative events such as skin breakdown or skin injury, muscle or connective tissue injury, reduced range of motion, falls, and discomfort as well as positive events such as increased range of motion, target socket states, goals, and presets, clinician or healthcare professional adjustments, clinician or healthcare professional goals, presets and target socket states.
In addition to input from sensors within and around the prosthetic socket and qualitative data collected through a smart phone application; additional input may be gathered through independent or third party applications, fitness trackers, and user interfaces. This may include a third party smart phone application that collects data on calories consumed or a fitness tracker that records the amount of sleep. Data transfer from a third party smart phone application is facilitated or aided by an application program interface (API) which is a set of routines, protocols, and tools for building software applications. An API specifies how software components should interact. Additionally, APIs are used when programming graphical user interface (GUI) components as it relates to routines, protocols, and parsing information created and acquired with in that systems network. The API allows the smart socket system to real time scrape a user's data from a third party application to the smart phone application and user interface associated with the described device and method.
Metrics described herein are referring to conclusions or meaningful results that are extrapolated or condensed from raw data that comes from one or more measurements or sensors. Metrics extrapolated from the measured data described above include; gait, weight distribution, user activity, and qualitative results. More specifically, weight distribution within the socket, abnormal gait patterns, gait speed, cadence, changes in direction, gait obstacles successfully navigated (such as a curb or set of stairs), changes in vertical distance, number of steps, distance of gait, duration of prosthetic use, gait distribution or comparisons of phases of gait such as stance phase vs. swing phase, changes in cadence, changes in activities (standing vs. sitting vs. walking), range of motion of the prosthetic limb, suspension movement on the residual limb (vertical movement between the residual limb and the prosthesis/socket), tension in the closure system vs. comfort in the socket, congruency or relative motion between the prosthetic socket and the residual limb, changes in volume, changes in tissue integrity, changes in tissue density, changes temperature, changes in muscle activity, calories burned (with prosthetic use which can be calculated from any metrics such as changes in vertical distance, gait distance, changes in direction, number of steps, gait speed, and other data), gait symmetry (percentage of time spent on prosthetic side vs. contralateral side or right and left for a bilateral amputee) and biological changes such as changes in blood sugar and changes in temperature vs. changes in activity calculated from biometric measurements. Qualitative result metrics may include; percentage of time spent in target socket weight distribution, percentage of step count goal accomplished, differences in quantitative data and qualitative reporting such as the difference in measured movement and reported perception of control, and longitudinal changes in their reported socket comfort score. These metrics can be used to improve care or service to the user, can facilitate safety and monitoring capabilities, can help facilitate insight and better service from healthcare professionals, and can confirm or refute usage for insurance payment.
Such an embodiment of a base model may include one or more of the following specific metrics: steps; stance versus swing; gait speed (ground speed); distance of gait; cadence (steps per time); K-Level (a combination of gait obstacles successfully navigated (such as a curb or set of stairs), changes in cadence, changes in vertical distance, gait distance, and changes in activities); socket state (pressure distribution per region, overall pressure, and pressure relationships); time of use (used vs not used, standing vs sitting vs walking); and/or geographic location. Such an embodiment of an expanded model may additionally include one or more of the following specific metrics: abnormal gait patterns; changes in direction; gait obstacles successfully navigated (such as a curb or set of stairs); changes in vertical distance; range of motion of the prosthetic limb; suspension movement on the residual limb (vertical movement between the residual limb and the prosthesis/socket); congruency or relative motion between the prosthetic socket and the residual limb; tension in the closure system vs comfort in the socket; residual limb volume, changes in residual limb volume; skin integrity or skin injury of the residual limb or contralateral limb; muscle or connective tissue integrity of the residual limb or contralateral limb; muscle or connective tissue activity of the residual limb or contralateral limb; calories burned (with prosthetic use which can be calculate from an metrics such as changes in vertical distance, gait distance, changes in direction, number of steps, gait speed, and other data); and/or gait symmetry (percentage of time spent on prosthetic side vs contralateral side or right and left for a bilateral amputee). Biometric measurements may include temperature; blood flow; heart rate; and/or blood glucose level/blood sugar. Qualitative result metrics may include: percentage of time spent in target socket weight distribution; percentage of step count goal accomplished; quantitative data and qualitative reporting such as the difference in measured movement and reported perception of control; and/or longitudinal changes in their reported socket comfort score.
What is described as an algorithm or algorithms refers to the set of steps, programming, mathematical formula, or process used to convert input or measured data into predictive analysis, prescriptive analysis, conclusions, or informative derivatives. Put simply, algorithms are used to accomplish valuable or meaningful outputs with the inputs of collected data. One example of an algorithm described is the set of steps needed to identify and send a notification to the user that they may need to tighten their socket brim in order to avoid pain and injury on the end of the residual limb. If the prosthetic socket device described includes an embodiment with an integrated mechanism for automatically adjusting the fit of their prosthetic socket. The primary process of determining what may need to change with the fit can still be applied, then an additional algorithm would be applied to use measured data in and derive an appropriate automated adjustment of the fit, to avoid user pain and discomfort.
A foundation or pool of data of existing research and tests may be generated and used as a body of knowledge or data that can be compared against itself without any users. These standard test and existing research may include Amputee Mobility Predictor (AMPnoPRO) and research on human gait. This data may be used to inform processing and conclusions from data collected from the user and the socket.
After a user is fit with a sensor-enabled prosthetic socket as described herein, baseline measurements may be measured on regular basis and statistical analyses may be conducted on the measurements to establish baseline for each user. The user is also prompted to record their optimal baseline or fit goal or target. This optimal baseline, fit goal, or target becomes a benchmark, where the total input data is saved at that moment as well as leading up to that moment and can then be compared to other states or sets of total input data and used for predictive algorithms, machine learning, prescriptive information, and notifications. Data used for said baseline may include all or any of the types described and may be stamped or tagged by the user, healthcare professional, or authorized family member through a smart phone application that includes a preset or record or comfort rating button, where the user can add qualitative information about their experience while simultaneously recording all sensor inputs.
In some embodiments, healthcare professionals can also set parameters and record goals/target states. These settings can be made along with adjustments to the fit or alignment of the prosthesis and are stored as independent variables from the user settings and can be compared to user settings. Furthermore, the resulting condition or experience of the amputee from these settings can be compared to that of their own settings to determine if the adjustment or alternative setting has been helpful. If the adjustment results in an improvement, supervised learning models will be applied to learn the new parameters or settings. Regression models or support vector machines can be applied to these parameter values. In machine learning, support vector machines (SVMs, also support vector networks) are supervised learning models with associated learning algorithms that analyze data used for classification and regression analysis. Each user's settings and models are individually adjusted per variation in data input.
In one method, the user's health state and mobility state may be calibrated against standard tests, such as Amputee Mobility Predictor (AMPnoPRO), four-square step test (FSST), up-and-go test/L-test, two minute walk test, four minute walk test, and research on human gait. The user may be asked to perform tests, such as the AMPnoPRO, and data can be collected while the user is participating in these tests. This process can be repeated multiple times with different control of fit parameters, and at different time length after the socket has been worn, and take the measurements for each. Conducting and recording the results for these tests is easier and more accurate through use of a smart phone application which can show the user and tester what the user needs to do and can start and stop the test while recording for improved accuracy. Statistical analyses are conducted on the measurements to calibrate the measurements against the actions and against the existing data pool of research conducted using these same tests. Measurements of the contralateral leg may also be used to calibrate results and as input into the predictive model to explain whether limb health is less than ideal. The baseline measurements along with calibration data points may also be used as training data for various machine learning models or methods.
Calibration can also be made by adjusting the socket a known amount, then recording the resultant sensor input, qualitative data via a smart phone application interfacing with the user. For example, the socket closure system or tensioner may be tightened by 3 cm or 50 pounds. of tension force, and the resulting change in pressure distribution and pressure profile as well as changes in the accelerometer and/or other input can be recorded. This may help to calibrate the data and may also be used to understand and set parameters for automated adjustment capabilities. For example, testing and data collection in this way may yield a conclusion that when distal end pressure exceeds 70 pounds. for a given user, increasing closure tension by 50 pounds. results in an average increase of socket comfort score of 40%.
Negative events relating to use of a prosthesis, such as an injury or fall, do occur. The described sensor enabled prosthetic socket and app-based user interface will allow the user, healthcare team, and/or user's family to record negative events when they occur. The recorded data leading up to these negative events become a valuable tool in avoiding future injury. For example, if a user records in their user interface app that they discovered a blister on the proximal-medial aspect of their socket this evening when they took off their prosthesis; the recorded settings from the day showing that the user had 55 pounds. of pressure in the proximal-medial aspect of their residual limb and that usually they have under 50 pounds. This metric of difference can be used to give notifications in pressure, to the proximal-medial aspect of their socket, stating that it has exceeded 50 pounds., and give an adjustment recommendation that has been learned, to move the user towards a target socket state. This can thereby help to avoid future injury and guide the user towards target states. Because there is a potential high variance between different data samples, learning from individual user events in this way and building sets of data across users is required in order to accurately and appropriately identify precursors of negative events and direct a user towards goals.
Negative events such as pain can also be predicted by detecting certain gait patterns that have been associated with pain in independent studies, such as antalgic gait and circumduction. Identifying such patterns in the amputee's gait distribution may help to detect or avoid pain. The frequency and velocity of such patterns in the measurement data can also be used as an indicator for the degree of pain. In order to improve accuracy, the predictive model herein pulls from multiple types of data and uses multiple analytical methods in order to improve accuracy and user or clinical relevance.
Amputees' pain points may vary by the time of day, tasks, and the physiological or dietary variables such as how much sodium the user consumed in the last 24 hours. With this device and method, users can set indicator flags of discomfort before the onset of the pain. These indicator flags are correlated with the input data for that specific time interval and used to build predictive models for pain. These inputs are treated as a binary classification problem (pain, no pain), a multi-class classification problem (such as no pain, light pain, medium pain, pain, extreme pain), or probability estimation problem.
Pain detection capabilities built into these models are also useful for people who have difficulty recognizing or sensing the pain themselves such as young children and/or people with nerve damage. Said methods may be able to drastically improve the user's ability to avoid pain or more precisely identify pain through these machine learning methods.
Detecting and avoiding falls may be made possible by data collected during controlled falls within harness systems and user input after falls have occurred. Sensors can also be used to determine the amount of clearance at the foot and can warn the user that the clearance in swing or the stability in stance is not sufficient. Furthermore, precursors or variables that increase the risk of a fall can be used to notify users of their increased risk. For example, lack of sleep, dehydration, and low caloric intake data that can be collected by a partner API such as Fitbit, or My Fitness Pal, can predict that a user has an increased chance of falling and can notify the user of this risk. This same partner data can also help to determine a likelihood or cause of volume reduction due to dehydration or excessive sodium intake for example.
In addition to comparisons of data for explicit time variables, it is also useful and valuable to longitudinally compare data. In other words, data is analyzed and compared over large periods of time, in addition to comparing data from specific time stamps. This analysis allows for broad conclusions to be drawn such as whether or not a user's activity level and mobility has increased or decreased. Mobility measurements such as AMPnoPRO can also be used as an additional data points to the overall metric of mobility over time.
Given the system can see and determine certain predefined antalgic gait patterns, these can act as an early stage warning system for a clinical expert. By example a user is seen in the system to be repeating a pattern that indicates osteoarthritis. The system would notify the associated specialist of a patient/user and this evaluation would take place in a prosthetist's office. This ability to sense gait patterns and this early stage warning system is an added benefit to patients and health care providers for creating opportunities for clinical intervention to manage potential problems before they get too out of hand, resulting in example of this case a total knee replacement in the sound side or good leg of an amputee.
Mobility improvement or deterioration may happen in a very slow process over long span of time. Very slow changes in measurements as people age or other conditions occur are to be detected by a fine-tuned amputee mobility assessment. This is accomplished by the calibration data obtained earlier, very detailed measurements data over long span, and statistical analysis for detecting shift in data. Users who allow for their data to be released and compared to other users can leverage increased machine learning capabilities through the increased pool of data.
An overall wellness index that takes data from user's daily input on the comfort/pain scale, pressure profile data, accelerometer data, partner API data, clinical metrics, and/or negative events such as antalgic gait, ulcers, or falls if they occur; can be used to track general trends in residual limb and overall wellness. Trending this index can provide a global view of whether the use of the socket is optimal and whether the usage is improving over time.
Qualitative recordings of the experience while simultaneously recording all inputs historically leading up to that time are prompted at regular time intervals and available for the user to enter whenever they desire to self-report information to the system. They are encouraged by the system to report whenever they feel a state change or discomfort they would like for the system to learn or take note of as it relates what their goals may be for comfort, control, tension or suspension. For example, the smart phone application may automatically prompt the user to enter their comfort score and user experience once a week and the user can additionally go to the interface home button and record their experience at any time. The user will also be asked to log any negative events or especially positive events such as a particularly comfortable and functional day on in their prosthetic socket. Predictive algorithms and machine learning principles such as Bayes' Theorem are applied to the described inputs, longitudinal data, cross-sectional data and other types of data in order to yield prescriptive information or notifications that can be useful to the user, healthcare team, researchers, and family.
The total input data for a given user can be typed or profiled into user categories or groups. Patterns can be identified within these groups and algorithms for these groups of similar people with amputation can be more accurate for predictive and prescriptive information, can more accurately and appropriately respond to automated adjustment of the socket, can have an improved feedback significance, and can more appropriately form user networks of similar people with higher likelihood of providing useful psychological and logistical peer support.
According to one embodiment, a method for using sensed prosthetic socket data may involve the following steps: foundational data is created (including user information like age, weight, height, etc); user baseline or targets are established; adjustments or targets may also be set by healthcare professionals; tests or calibration may also be conducted and recorded by healthcare professionals; any negative events are automatically and qualitatively recorded; longitudinal, cross-sectional, and other types of data collected over time are compared and analyzed automatically by the application programming or cloud based computing; qualitative recordings and their subsequent inputs are prompted at regular time intervals and available to the user at any time; and predictive algorithms and machine learning principles such as Bayes' Theorem are applied to yield prescriptive information or notifications that can be useful to the user, healthcare team, researchers, or family.
The steps described above may be repeated multiple times and increased machine learning from previous process cycles can be improved. Some cycles may only be portions of the steps before cycling back through again. For example, the last 4 steps may be repeated for many different cycles then the user decides to recalibrate their targets so then a loop may occur with all steps described.
The data or inputs, analysis of described inputs, conclusions from described inputs or analysis, and/or notifications may be provided to the patient, authorized healthcare professionals, family members, payers, researchers, and/or the like. One way of providing the data is via a computerized user interface, which may also serve, in some embodiments, as a cross-collaborative communication network where data, experience, insights, and other valuable elements can be shared across the network and from one party to another. For example, if a user experiences an ulcer on their residual limb after a trip to the store, they can take a picture of their skin and push it to their doctor, prosthetist, and caretaker along with the total input data recorded from that day or that trip.
Embodiments of the invention are directed toward using sensor-derived information to profile the physical state of a prosthetic socket when worn by a patient. “State”, as used herein, refers to a description of a system in terms of classical mechanics. Sensor-derived information from a prosthetic socket thus refers to any aspect of a prosthetic socket, as worn by a user, which is informative of the mechanics or physical properties of a socket in time and space. The status of forces within and flowing through the socket is an example of what may be broadly understood as the socket's report of what its physical state is, in terms of any of force, tension, acceleration, position, or phase of a material composition.
In a basic view of a prosthetic socket as an intermediary device bridging a proximal residual limb to a distal prosthetic limb, embodiments of the invention profile the distribution of force within the prosthetic socket as such force travels from the prosthetic residual limb to the distal prosthetic limb. Examples of sensor-reported data and socket state, as depicted in
Embodiments of the technology include systems (
Force is generated by the body weight of the user even when the user is standing still, and as body weight is amplified by acceleration and impact associated with ambulation. A prosthetic socket performance profile or socket state refers to comprehensively processing sensor-derived data into a mapping of the distribution of forces as they flow from the body weight of the user through the socket from its proximal end to its distal end. A force profile, as used herein, may further include processed sensor data reporting on acceleration occurring within the socket, by region, and position of socket components in space and in relation to each other.
This profile information can be used in two basic ways, per various embodiments: (1) a strictly “informational” system (such as
Informational system embodiments are useful as the basis for making manual adjustments of the prosthetic socket. Such embodiments may also provide alerts when data show a deviation from a predetermined desired profile of distribution of force through the prosthetic socket. In actuatable embodiments, building on an underlying informational system, processed sensor data can be applied to actuate adjustment features of the prosthetic socket that are activated when data report a deviation from a desired profile of distribution through the socket. The actuatable features, acting on instructions from system embodiments, respond by changing the configuration of the prosthetic socket in such a way as to drive the profile of force distribution toward a desired optimal profile. Force distribution profiles can be rendered as visually intuitive dashboard style diagrams, appropriate for a graphic user interface. In the actuatable aspect of the embodiments, the overall force profile is used as feedback to drive appropriate movement of motorized mechanical features of a prosthetic socket, such as, merely by way of example, a tension-based closure system, a hinge, or an adjustable height ischial seat.
By way of one particular example of a force distribution issue, it is generally advantageous to distribute force away from the distal end of a residual limb, diverting force through the peripheral structures of the prosthetic socket. Profiling the force distribution paths within a prosthetic socket when worn by a patient will reveal when an inordinate amount of force is conveyed through the distal end. When the force profile of an empirically determined optimum is compared to a profile that shows too much force is conveyed through the distal limb, one or more circumferential tensioning mechanisms may be actuated by the system to distribute the force profile back toward a previously established optimal or desired profile. (
As shown in
Referring now to
All proximal force 1 enters the prosthetic socket from the body weight of a user, positioned proximal to the socket. Proceeding distally, force distributes into central path 2 and peripheral path 3. Central path 2 represents the force conveyed through the residual limb (that which is not bled away into the peripheral path), and which is ultimately transferred through the distal end of the residual limb. Peripheral path 3 represents body weight driven force absorbed by a proximal component of a socket, such as a brim in the case of a transfemoral prosthetic socket (from the full circumference of the residual limb, but particularly from the ischial seat). Force from the brim is transferred distally through the longitudinal socket structures, such as struts. Force from the residual limb is also transferrable from the residual limb to the struts throughout their full length. Convergent path 4 represents the convergence of the peripheral path 2 and the central path 3 at the distal base of the prosthetic socket. The summed forces of the convergence path 4 are transferred distally from there to distal prosthetic components.
Each of these three exemplary prosthetic sockets has been previously described in detail. Examples of a modular transfemoral prosthetic socket are described in U.S. Pat. No. 8,978,224, entitled “Modular Prosthetic Sockets and Methods for Making Same”, and in U.S. Patent Application Pub. No. 2014/0277584, entitled “Modular prosthetic sockets and methods for making and using same,” now abandoned. Examples of a modular transtibial prosthetic socket are described in U.S. Provisional Patent Application Nos. 62/237,204, filed Oct. 5, 2015; 62/287,702, filed Jan. 27, 2016; and 63/305,477, filed on Mar. 8, 2016. Examples of a modular osseointegrated abutment support socket are described in U.S. Provisional Patent Application Nos. 62/197,427, filed Jul. 27, 2015; 62/267,820, filed Dec. 15, 2015; and 62/334,791, filed May 11, 2016. All of the above-listed patent applications are hereby incorporated in their entireties herein.
In one aspect, a function of prosthetic socket 90 is to divert the proximal force 1 peripherally as peripheral forces 3, thereby minimizing the magnitude of the central force 2 through the distal end of the femur 104, which is not physically suited to bear significant force, and which is sensitive to such force. As described herein, one of the functions of smart socket systems is to quantify the level of force driven through the distal end of femur 104 and provide instructions or make automatic adjustments within prosthetic socket 90 to direct force peripherally, away from the distal end of femur 104. This is true of the following two embodiments illustrated in
Referring now to
With reference now to
The microcontroller/data transmission unit/battery 59 (
Referring now to
With reference now to
Referring again to
Referring now to
Referring now to
Referring to
Referring now to
In some cases, a particular patient may have past data regarding patient performance and/or socket performance, and this past data can be used, for example, in generating desired patient and socket profiles. In other cases, a patient might not have any previous use data or that data might not be available. In those cases, it may be possible to use data from a broad group of users to create patient performance standards and/or socket profiles. In some cases, through previous use and/or testing of a prosthetic socket on a patient, a full range of available mechanical adjustments may be recorded, and optimal prosthetic socket performance profiles may be created for respective activities. Such activities may include, for example, lying down, sitting, standing, walking slowly and walking quickly. Each of these activities can be captured and recognized by profile signatures derived from sensor data (including input from force, acceleration, and position sensors). An overall preferred profile of each of these types of activities becomes the “individual standard,” against which any set of instant observations is compared for the patient. This individual standard can be compared and contrasted to general performance standards derived across different user profiles in order to extrapolate relevant clinical learning that may inform refinement of either individual standards or general performance standards.
Referring now to
Referring now to
Referring now to
The total input data or any portion therein may be associated with these or other specific regions and the relationship between these regions, including a comparison of like data per region, which offers a clinical or user resource that can improve prosthetic care. One example of using data for specific regions and the relationships between the regions to improve prosthetic care for the user is a transtibial prosthesis that can loosen automatically when the user sits down. This may be accomplished by a transtibial prosthesis that includes sensors in the regions shown above in
Another example of using data for specific regions and the relationships between the regions to improve prosthetic care for the user is a transtibial prosthetic socket that includes sensors in the regions shown above in
The regions of a prosthetic socket and residual limb specified in
Referring now to
Referring now to
Referring to
Referring to
Referring now to
The sensor-enabled pneumatic bladder system 710 may be set by the user, healthcare provider, and/or the others, to maintain a desired amount of average pressure regardless of volume fluctuation, to vary the amount of pressure per activity, to vary the amount of pressure over time or while sitting in order to help promote better circulation to the residual limb and body, and/or specific amounts of pressure per region of the residual limb or corresponding prosthesis.
With reference now to
Referring to
Referring now to
This application claims the benefit of U.S. Provisional Patent Application No. 62/364,930, filed on Jul. 21, 2016, which is titled, “Prosthetic Sockets That Are Sensor Enabled to Provide Data for Clinical Use and Mechanical Adjustments,” and which is hereby incorporated by reference. The present application is also related to the following patents and pending applications: U.S. Pat. No. 8,978,224, entitled “Modular Prosthetic Sockets and Methods for Making Same,” filed on Nov. 13, 2012; U.S. patent application Ser. No. 14/213,788, entitled “Modular Prosthetic Sockets and Methods for Making and Using Same,” filed Mar. 14, 2014, published as U.S. Patent Publication No. 2014/0277584, and now abandoned; U.S. Pat. No. 9,468,542, entitled “Prosthetic Socket and Socket Liner with Moisture Management Capability,” filed on Jun. 20, 2014; U.S. patent application Ser. No. 14/659,433, entitled “Modular Prosthetic Sockets and Methods for Making Same,” filed on Mar. 16, 2015, published as U.S. Patent Application No. 2015/0190252, and now abandoned; U.S. Provisional Patent Application No. 62/275,546, entitled, “Prosthetic Socket That is Sensor Enabled to Provide Data for Clinical Use and Mechanical Adjustments,” filed on Jan. 6, 2016; and U.S. Provisional Patent Application No. 62/334,791, entitled “Prosthetic Support Device for an Osseointegrated Femoral Abutment,” filed on May 11, 2016. All of the above-referenced patents and applications are hereby incorporated by reference, in their entireties, into the present patent application.
| Number | Date | Country | |
|---|---|---|---|
| 62364930 | Jul 2016 | US |
