This invention relates to devices for the rehabilitation of disabled persons with a neurological injury, such as stroke or spinal-cord injury, or otherwise impaired anatomical extremities, and novel methods and apparatus for facilitating the same.
A new and exciting branch of physical and occupational therapies is therapy assisted by a computer-directed robotic arm or device (sometimes also called a “manipulator” to distinguish it from the human arm that may engage it, in certain embodiments). These robotic systems leverage plasticity in the brain, which literally rewires the brain. Recent science has demonstrated that dosage (i.e., the amount of time engaged in therapy) is an essential element in order to benefit from this effect. The potential benefits of using a manipulator system for tasks such as post-stroke rehabilitative therapy, which typically involves moving a patient's limb(s) through a series of repeated motions, are significant. There exist some types of therapy, such as error-augmentation therapy, that simply cannot be implemented effectively by a human therapist. Furthermore, computer-directed therapy can engage the patient in games, thereby making the experience more enjoyable and encouraging longer and more intense therapy sessions, which are known to benefit patients. Finally, the therapist is able to work with more patients, e.g., the therapist is able to work with multiple patients simultaneously, the therapist is able to offer patients increased therapy duration (higher dosage) since the session is no longer constrained by the therapist's physical endurance or schedule, and the therapist is able to work more consecutive therapy sessions since the number of consecutive therapy sessions is no longer constrained by the therapist's physical endurance or schedule.
A useful way to categorize robotic rehabilitation systems is by the number of degrees of freedom, or DOFs, that they have. Generally speaking, for mechanical systems, the degrees of freedom (DOFs) can be thought of as the different motions permitted by the mechanical system. By way of example but not limitation, the motion of a ship at sea has six degrees of freedom (DOFs): (1) moving up and down, (2) moving left and right, (3) moving forward and backward, (4) swiveling left and right (yawing), (5) tilting forward and backward (pitching), and (6) pivoting side to side (rolling). The majority of commercial robotic rehabilitation systems fall into one of two broad categories: low-DOF systems (typically one to three DOFs) which are positioned in front of the patient, and high-DOF exoskeletal systems (typically six or more DOFs) which are wrapped around the patient's limb, typically an arm or leg. Note that these exoskeletons also need the ability to adjust the link lengths of the manipulator in order to accommodate the differing geometries of specific patients. Generally speaking, an exoskeletal system can be thought of as an external skeleton mounted to the body, where the external skeleton has struts and joints corresponding to the bones and joints of the natural body. The current approaches for both categories (i.e., low-DOF systems and high-DOF exoskeletal systems) exhibit significant shortcomings, which have contributed to limited realization of the potential of robotic rehabilitation therapies.
Low-DOF systems are usually less expensive than high-DOF systems, but they typically also have a smaller range of motion. Some low-DOF systems, such as the InMotion ARM™ Therapy System of Interactive Motion Technologies of Watertown, Massachusetts, USA, or the KINARM End-Point Robot™ system of BKIN Technologies of Kingston, Ontario, Canada, are limited to only planar movements, greatly reducing the number of rehabilitation tasks that the systems can be used for. Those low-DOF systems which are not limited to planar movements must typically contend with issues such as avoiding blocking a patient's line of sight, like the DeXtreme™ system of BioXtreme of Rehovot, Israel; providing an extremely limited range of motion, such as with the ReOGO® system of Motorika Medical Ltd of Mount Laurel, New Jersey, USA; and insufficiently supporting a patient's limb (which can be critically important where the patient lacks the ability to support their own limb). Most of these systems occupy space in front of the patient, impinging on the patient's workspace, increasing the overall footprint needed for a single rehabilitation “station” and consuming valuable space within rehabilitation clinics.
High-DOF exoskeletal systems, such as the Armeo®Power system of Hocoma AG of Volketswil, Switzerland, the Armeo®Spring system of Hocoma AG of Volketswil, Switzerland, and the 8+2 DOF exoskeletal rehabilitation system disclosed in U.S. Pat. No. 8,317,730, are typically significantly more complex, and consequently generally more expensive, than comparable low-DOF systems. While such high-DOF exoskeletal systems usually offer greater ranges of motion than low-DOF systems, their mechanical complexity also makes them bulky, and they typically wrap around the patient's limb, making the high-DOF exoskeletal systems feel threatening and uncomfortable to patients. Furthermore, human joints do not conform to axes separated by links the way robot joints do, and the anatomy of every human is different, with different bone lengths and different joint geometries. Even with the high number of axes present in high-DOF exoskeletal systems, fine-tuning an exoskeleton system's joint locations and link lengths to attempt to follow those of the patient takes considerable time, and even then, the high-DOF exoskeletal system frequently over-constrains the human's limb, potentially causing more harm than good.
Finally, there are a handful of currently-available devices which do not fit in either of the two categories listed above: for example, high-DOF non-exoskeletal devices, or low-DOF exoskeletal devices. To date, these devices have generally suffered the weaknesses of both categories, without leveraging the strengths of either. A particularly notable example is the KINARM Exoskeleton Robot™ of BKIN Technologies of Kingston, Ontario, Canada, which is an exoskeletal rehabilitation device designed for bi-manual and uni-manual upper-extremity rehabilitation and experimentation in humans and non-human primates. Like the KINARM End-Point Robot™ of BKIN Technologies of Kingston, Ontario, Canada (see above), the KINARM Exoskeletal Robot™ system provides only two degrees of freedom for each limb, limiting the range of rehabilitation exercises that it can conduct. Meanwhile, by implementing an exoskeletal design, the KINARM Exoskeletal Robot™ device can provide some additional support to the patient's limb, but at the cost of significant increases in device size, cost, complexity and set-up time.
While robot-assisted physical and occupational therapy offers tremendous promise to many groups of patients, the prior art has yet to match that promise. As the previous examples have shown, current therapy devices are either too simplistic and limited, allowing only the most rudimentary exercises and frequently interfering with the patient in the process; or too complex and cumbersome, making the devices expensive, intimidating to patients, and difficult for therapists to use. Thus there remains a need for a novel device and method that can provide patients and therapists with the ability to perform sophisticated 2-D and 3-D rehabilitation exercises, in a simple, unobtrusive and welcoming form factor, at a relatively low price.
In addition to the foregoing, with robot-assisted physical and occupational therapy in general, and upper-extremity robot-assisted physical and occupational therapy in particular, it has been found that patients often perform undesirable compensatory movements during such therapy that hinders achievement of desirable therapeutic outcomes. Clinical studies have shown that patients tend to use compensatory movements (i.e., any movements that deviate from a clinically desired movement), e.g., forward flexion of the trunk, extension of the trunk, lateral flexion of the trunk, shoulder hiking, etc., in order to overcome motor impairments and achieve particular exercise goals. Furthermore, it has been found that discouraging compensatory movement strategies developed by patients in response to motor impairments can be a dominant force in shaping post-stroke neural remodeling responses with mixed effects on functional outcome.
While a trained therapist can visually identify such compensatory movements as they occur if the trained therapist is watching the patient carefully, it is challenging for a therapist to simultaneously monitor a plurality of patients to identify such compensatory movements as they occur (and move to correct them). Furthermore, where a plurality of patients are to be monitored by a single therapist, there is a need for the system to be able to quickly identify each patient such that patient movements are attributed to the correct patient (and such that records pertaining to the same) are automatically stored in the proper electronic medial record.
Thus there is a need for a new and improved apparatus which facilitates the monitoring of a group of patients during robot-assisted physical and occupational therapy by a reduced number of therapists, so as to help identify and eliminate compensatory movements by the patient during therapy (while also identifying the patients).
It has also been found that it is clinically helpful to periodically assess a patient's progress during physical therapy (e.g., to assess whether continued therapy is likely to be fruitful, or whether it is more productive to switch to a different therapy). By way of example but not limitation, some conventional scales that may be used for such an assessment are the Wolf Motor Function Test (WMFT), the Function Ability Scale (FAS), and the Fugl-Meyer Assessment (FMA). As with compensatory movements, it would be desirable to be able to perform such assessments during therapy, particularly during group therapy in which a single therapist monitors a plurality of patients.
Thus there is also a need for a new and improved apparatus that can be used during a therapy session to assess patient progress according to conventionally-used assessment scales which does not require halting therapy or direct intervention by the therapist.
The present invention bridges the categories of low-DOF systems and high-DOF exoskeletal systems, offering the usability, mechanical simplicity and corresponding affordability of a low-DOF system, as well as the reduced footprint, range of motion, and improved support ability of a high-DOF exoskeletal system.
More particularly, the present invention comprises a relatively low number of active (powered) DOFs—in the preferred embodiment, three active DOFs, although the novel features of the invention can be implemented in systems with other numbers of DOFs—which reduces the device's cost and complexity to well below that of high-DOF exoskeletal systems. However, because of the innovative positional and orientational relationship of the system to the patient—unique among non-exoskeletal systems to date, as explained further below—the device of the present invention enjoys advantages that have previously been limited to high-DOF exoskeletal systems, such as more optimal torque-position relationships, better workspace overlap with the patient and a greater range of motion.
In addition, it has been discovered that a novel implementation of a cabled differential (with the differential input being used as a pitch axis and the differential output being used as a yaw axis relative to the distal links of the device) permits the mass and bulk of the power drives (e.g., motors) to be shifted to the base of the system, away from the patient's workspace and view. Through the combination of these two major innovations—the orientation and position of the device relative to the patient, and the implementation of a cabled differential with special kinematics—as well as other innovations, the present invention provides a unique rehabilitation device that fills a need in the rehabilitation market and is capable of a wide variety of rehabilitation tasks.
Significantly, the present invention enables a new method for bi-manual rehabilitation—a new class of rehabilitative therapy where multiple limbs, usually arms, are rehabilitated simultaneously—in which rehabilitative exercises can be conducted in three dimensions, by using two similar devices, simultaneously and in a coordinated fashion, on two different limbs of the patient.
The present invention also comprises a novel computer system comprising at least one camera for monitoring the patient during therapy, wherein the novel computer system is configured to utilize an AI-based platform to (i) utilize facial-recognition technology to identify a patient and link/record data concerning that patient to an electronic medical record particular to that patient, (ii) track movements of one or more patients during robot-assisted therapy in order to identify compensatory movements that can detract from therapy and notify the therapist of the same, (iii) track movements of one or more patients during robot-assisted therapy in order to perform real-time assessments of patient progress during therapy, and (iv) facilitate group robot-assisted therapy sessions in which a single therapist supervises a plurality of patients and the system acts to enhance patient safety while simultaneously providing diagnostic tools for enhancing therapy.
In one preferred form of the invention, there is provided a non-exoskeletal rehabilitation device, with as few as 2 active degrees of freedom, wherein the device is oriented and positioned such that its frame of reference (i.e., its “reference frame”) is oriented generally similarly to the reference frame of the patient, and motions of the patient's endpoint are mimicked by motions of the device's endpoint.
In another preferred form of the invention, there is provided a non-exoskeletal rehabilitation device, with as few as 2 active degrees of freedom, of which 2 degrees are linked through a cabled differential.
In another preferred form of the invention, there is provided a method for bi-manual rehabilitation, wherein the method utilizes a pair of rehabilitation devices, wherein each rehabilitation device is designed to be capable of inducing motion in three or more degrees of freedom, is easily reconfigurable to allow both right-handed and left-handed usage, and is located relative to the patient such that two devices may be used simultaneously without interfering with each other.
In another preferred form of the invention, there is provided a robotic device for operation in association with an appendage of a user, wherein the appendage of the user has an endpoint, the robotic device comprising:
In another preferred form of the invention, there is provided a method for operating a robotic device in association with an appendage of a user, wherein the appendage of the user has an endpoint, the method comprising:
In another preferred form of the invention, there is provided a robotic device comprising:
In another preferred form of the invention, there is provided a robotic device comprising:
In another preferred form of the invention, there is provided a robotic device comprising:
In another preferred form of the invention, there is provided a robotic device comprising:
In another preferred form of the invention, there is provided a method for providing rehabilitation therapy to a user, the method comprising:
In another preferred form of the invention, there is provided a method for providing rehabilitation therapy to a user, the method comprising:
In another preferred form of the invention, there is provided a method for providing rehabilitation therapy to a user, the method comprising:
In another preferred form of the invention, there is provided a method for providing rehabilitation therapy to a user, the method comprising:
In another preferred form of the invention, there is provided a robotic device for operation in association with a body of a user, wherein the body of the user comprises a torso and a limb, the robotic device comprising:
In another preferred form of the invention, there is provided a method for providing rehabilitation therapy to a user, the method comprising:
In another preferred form of the invention, there is provided a system for facilitating delivery of physical therapy to a patient, the system comprising:
In another preferred form of the invention, there is provided a method for delivering physical therapy to a patient, the method comprising:
These and other objects and features of the present invention will be more fully disclosed or rendered obvious by the following detailed description of the preferred embodiments of the invention, which is to be considered together with the accompanying drawings wherein like numbers refer to like parts, and further wherein:
Looking first at
The preferred embodiment shown in
To provide additional degrees of freedom, different endpoint attachments may be provided at the location of the coupling element 115, to permit different degrees of control over the patient's limb orientation, or to provide additional therapeutic modalities. By way of example but not limitation, different endpoint attachments may comprise a single-DOF endpoint attachment for performing linear rehabilitation exercises; or a three-DOF endpoint attachment to enable more complex motions, by enabling control over the orientation of the patient's limb; or an actively-controlled multi-DOF endpoint attachment. By reducing the number of degrees of freedom in the core of the robotic device to three in the preferred implementation (i.e., the robotic device 5 shown in
Looking next at
In this definition of kinematic frames, transmission components are excluded to simplify definition: a pulley within a transmission may be located away from a given joint, but rotate with that joint. Similarly, some pulleys in the system may be caused to rotate by the motion of more than one axis—for example, when they are part of a cabled differential, such as is employed in the preferred form of the present invention.
In the preferred embodiment, joints J1 and J2 are implemented through the use of a cabled differential transmission, designed similarly to that disclosed in U.S. Pat. No. 4,903,536, issued Feb. 27, 1990 to Massachusetts Institute of Technology and J. Kenneth Salisbury, Jr. et al. for COMPACT CABLE TRANSMISSION WITH CABLE DIFFERENTIAL, which patent is hereby incorporated herein by reference.
As described in U.S. Pat. No. 4,903,536, a cabled differential is a novel implementation of a differential transmission, in which two input pulleys (e.g., pulleys 505 in the robotic device 5 shown in
Stated another way, as described in U.S. Pat. No. 4,903,536, the cabled differential is a novel implementation of a differential transmission, in which two input pulleys (e.g., pulleys 505 in robotic device 5 shown in
In other words, as described in U.S. Pat. No. 4,903,536, the cabled transmission is a novel implementation of a differential transmission, wherein two input pulleys (e.g., pulleys 505 in robotic device 5 shown in
As seen in
By implementing this set of diametral relationships in the series of pulleys (i.e., input pulleys 505 and output pulley 540), progressively higher transmission ratios are achieved through the cabled transmission. In the preferred embodiment, a transmission ratio of 8.51:1 is implemented between motor pinions 510 and input pulleys 505, and a transmission ratio of 1.79:1 is implemented between input pulleys 505 and output pulley 540, generating a maximum transmission ratio between motor pinions 510 and output pulley 540 of 15.26:1. Throughout this cabled transmission, and all cabled transmissions of the present invention, care is taken to ensure that the ratio between the diameter of a given cable and the smallest diameter that it bends over is kept at 1:15 or smaller. Larger ratios, occurring when the cable is bent over smaller diameters, are known to significantly reduce cable fatigue life.
Still looking now at
As described in U.S. Pat. No. 4,903,536, this design has the benefit of moving the mass of motor 565 toward base 502 of robotic device 5, reducing the inertia of the system. In the preferred implementation, the motor's mass is positioned coaxial to axis 130 of joint J2, and as close as possible to axis 125 of joint J1, thereby reducing inertia about both axes. This design is particularly valuable in the preferred implementation shown, since the mass of motor 565 is moved close to both axis 130 of joint J2 and axis 125 of joint J1, thereby reducing inertia about both axes. A transmission ratio of 1.89:1 is preferably implemented between motor pinion 570 and intermediate pulleys 575, and a transmission ratio of 5.06:1 is preferably implemented between intermediate pulleys 575 and output pulley 580, yielding a maximum transmission ratio between motor pinion 575 and output pulley 580 of 9.55:1.
All transmission ratios listed here have been optimized based on a range of factors, including:
This optimization process is extensive and at least partially qualitative; it is not reproduced here, since both the optimization process and its outcome will change significantly as the above factors change. Based on data gathered from a number of sources and internal experimentation, these forces are estimated to be:
Beyond output pulley 580 of joint J3, there is generally an outer link 110 (
Robotic device 5 also comprises an onboard controller and/or an external controller for controlling operation of robotic device 5. The onboard controller and/or external controller are of the sort which will be apparent to those skilled in the art in view of the present disclosure. By way of example but not limitation,
There may also be other components that are included robotic device 5 which are well known in the art of robotic devices but are not shown or delineated here for the purposes of preserving clarity of the inventive subject matter, including but not limited to: electrical systems to actuate the motors (e.g., motors 500 and 565) of the robotic device; other computer or other control hardware for controlling operation of the robotic device; additional support structures for the robotic device (e.g., a mounting platform); covers and other safety or aesthetic components of the robotic device; and structures, interfaces and/or other devices for the patient (e.g., devices to position the patient relative to the robotic device, a video screen for the patient to view while interacting with the robotic device, a patient support such as, but not limited to, a wheelchair for the patient to sit on while using the robotic device, etc.).
Some specific innovative aspects of the present invention will hereinafter be discussed in further detail.
As discussed above, robotic device 5 is a non-exoskeletal rehabilitation device. Exoskeletal rehabilitation devices are generally understood as those having some or all of the following characteristics:
In
Because the aforementioned two “conditions” of an exoskeletal system are not met (i.e., the joint axes J1, J2 and J3 of the robotic device are not intended to be coaxial with the patient's joint axes 600, 605, 610 and 615, and because the segments of the patient's limb are not secured to corresponding segments of the arm of the robotic device), the robotic device of the present invention is not an exoskeletal rehabilitation device. While there are many non-exoskeletal rehabilitation devices currently in existence, the non-exoskeletal design of the present device is a critical characteristic distinguishing it from the prior art, since the device incorporates many of the beneficial characteristics of exoskeletal devices while avoiding the cost and complexity that are innate to exoskeletal designs.
Before further explaining this concept, it is helpful to provide some terminology. The “patient reference frame” (or PRF) 160 and the “device reference frame” (or DRF) 170, as used here, are located and oriented by constant physical characteristics of the patient and robotic device 5. As shown in
A similar reference frame is defined for the robotic device. The origin is placed at the centroid of the base of robotic device 5, which must also be fixed in space. The “forward” vector 172 is defined as the component of the vector pointing from the origin to the geometric centroid of the device's workspace. The “up” vector 171 and the “right” vector 173 may be defined in arbitrary directions, so long as they meet the following conditions:
In some cases, such as with the ReoGO® arm rehabilitation system of Motorika Medical Ltd. of Mount Laurel, New Jersey, USA, the aforementioned condition “4)” cannot be satisfied because the device's “forward” vector already points in the generally accepted “up” direction; consequently, the “up” vector may be defined arbitrarily subject to the three previous conditions. This case is further detailed below.
When existing rehabilitation devices are separated into exoskeletal and non-exoskeletal devices as per the description above, a further distinction between these two groups becomes apparent based on this definition of reference frames. In exoskeletal devices, the robotic device and the patient operate with their reference frames (as defined above) oriented generally similarly, i.e.,, “up”, “right” and “forward” correspond to generally the same directions for both the patient and the robotic device, with the misalignment between any pair of directions in the PRF (patient reference frame) and DRF (device reference frame), respectively, preferably no greater than 60 degrees (i.e., the “forward” direction in the DRF will deviate no more than 60 degrees from the “forward” direction in the PRF), and preferably no greater than 45 degrees. Meanwhile, to date, a non-exoskeletal device in which the device reference frame and the patient reference frame are generally oriented similarly in this way has not been created. Devices available today are oriented relative to the patient in a number of different ways, including the following:
The robotic device of the present invention is the first non-exoskeletal device which is designed to operate with its reference frame 170 oriented generally similarly to the reference frame 160 of the patient. This innovation allows the robotic device to leverage advantages that are otherwise limited to exoskeletal devices, including:
Like an exoskeletal device, robotic device 5 generally mimics the movements of the patient's limb, in that the endpoint of the device tracks the patient's limb, and a given motion in reference frame 160 of the patient produces motion in a generally similar direction in the device's reference frame 170. For example, if the patient moves their limb to the right in the patient's reference frame 160, the device's links will generally move to the right in the device's reference frame 170, as shown in
Because of the need for this distinction between the robotic device of the present invention and exoskeletal devices (i.e., that a relationship cannot easily be defined between the patient's limb and the links of robotic device 5), it is necessary to define the relationship between the robotic device and the patient as a function of the bases, endpoints and orientations of the robotic device and the patient. By defining device and patient reference frames in this manner, the previous statement that “robotic device 5 is designed such that its motions mimic those of the patient, in that a given motion of the patient's endpoint in reference frame 160 of the patient will be matched by a generally similar motion of the device's endpoint in reference frame 170 of robotic device 5” is satisfied only when robotic device 5 is oriented relative to the patient as described herein.
A series of simple logical tests have been developed to aid in determining whether a device meets the criteria outlined above. For these tests, the device is assumed to be in its typical operating position and configuration relative to the patient, and a PRF is defined for the patient's limb undergoing rehabilitation as described above.
Stated another way, generally similar orientation between the patient and the device can be examined by identifying a “forward” direction for both the user and the device. In the patient's case, the “forward” direction can be defined as the general direction from the base of the patient's arm undergoing rehabilitation, along the patient's limb, towards the patient's endpoint when it is at the position most commonly accessed during use of the device. In the device's case, the “forward” direction can be defined as the general direction from the base of the device, along the device's links and joints, towards the device's endpoint when it is at the position most commonly accessed during use of the device. If the “forward” direction of the device and the “forward” direction of the patient are generally parallel (e.g., preferably with less than 60 degrees of deviation, and more preferably with less than 45 degrees of deviation), then the device and the user can be said to be generally similarly oriented.
One preferred embodiment of the present invention is shown in
It should be noted that while this arrangement (i.e., with robotic device 5 positioned to the side of, and slightly behind, the patient) has been found to be preferable for certain rehabilitative therapies, there are other embodiments in which robotic device 5 is positioned differently relative to the patient which may be better suited to other applications, such as use as a haptic input/control device, or other rehabilitative activities. For example, in the case of advanced-stage arm rehabilitation, in situations where the patient is reaching up and away from the device, it may prove optimal to place the robotic device slightly in front of the patient.
Looking next at
Furthermore, in the preferred embodiment shown in
To date, however, the cabled differential has not been used in a configuration where neither of the differential axes is coaxial to one of the links. This configuration has been successfully implemented in the preferred embodiment of the present invention, as seen in both
Finally,
In an exemplary implementation shown in
The robotic device 5 described here is the first non-planar rehabilitation device to be purpose-designed for this type of dual-device, simultaneous use in a three-dimensional bi-manual system. As described earlier, the robotic device's innate symmetry allows its chirality to be easily reversed, allowing the same robotic device design to be used for rehabilitation of both right and left limbs. Furthermore, the device's small footprint facilitates simultaneous use of two systems, as shown in
There exists one known example of a system that is nominally capable of performing limited 3-dimensional bi-manual rehabilitation therapies with only uni-manual actuation, i.e., the 3rd-generation Mirror-Image Motion Enabler (MIME) rehabilitation robot, developed as a collaborative project between the Department of Veterans Affairs and Stanford University in 1999. See “Development of robots for rehabilitation therapy: The Palo Alto VA/Stanford experience.” Burgar et. al. Journal of Rehabilitation Research and Development. Vol. 37 No. 6, November/December 2000, pp. 663-673. The 3rd-generation MIME robot consists of a PUMA-560 industrial robot affixed to the patient's afflicted limb, and a passive six-axis MicroScribe™ digitizer affixed to a splint, which is in turn coupled to the patient's healthy limb. In the system's bi-manual mode, motions of the healthy limb are detected by the digitizer and passed to the robotic arm, which moves the afflicted limb such that its motions mirror those of the healthy limb. While this system can execute a limited set of bi-manual rehabilitation therapies, it is fundamentally limited by the uni-directional flow of information within the system: information can be passed from the healthy limb to the afflicted limb, but not from the afflicted limb back to the healthy limb to the healthy limb, since the digitizer is passive and does not have motors or other mechanisms with which to exert forces on the patient's healthy limb.
In the implementation described herein, the use of two similar, active robotic devices 5—in the preferred implementation, with similar kinematics, joint ranges, force output limits and static and dynamic performance characteristics—enables bi-directional information flow (i.e., bi-directional information flow wherein both devices send, receive and respond to information from the other device), creating a bi-manual rehabilitation system that is capable of monitoring the position of both the afflicted and healthy limbs, moving the patient's afflicted limb in three dimensions and potentially controlling its orientation simultaneously, and optionally providing simultaneous force feedback, support or other force inputs to the healthy limb. For example, the robotic device connected to the patient's healthy limb can be used to “drive” the robotic device connected to the patient's afflicted limb, while simultaneously supporting the healthy limb to prevent fatigue, and providing force feedback to the healthy limb as required by the therapy. In this respect it has been found that the cable drives used in the preferred implementation of the present invention are particularly well suited to this type of use, because of the high mechanical bandwidth of cable drive transmissions; however, alternative embodiments could be implemented using alternative mechanical drive systems. Regardless of the specific implementation, this bi-directional information flow—when executed between two similar devices with the facilitating characteristics described here—allows the device to be used for a far wider range of three-dimensional bi-manual rehabilitative therapies than prior art systems and enables the method disclosed herein.
In the foregoing sections, robotic device 5 was described as having a coupling element 115 for coupling outer link 110 to a patient, commonly to a limb of a patient, with outer link 110 being detachably connected to the remainder of the robotic device at the aforementioned mechanism 590 (
Note that in
Note also that in this form of the invention, U-shaped frame 140 may be supported above base 100 via a telescoping assembly 827 which allows the height of U-shaped frame 140 (and hence the height of the robotic arm) to be adjusted relative to base 100. This feature is highly advantageous, since it facilitates the use of robotic device 5 with patients who are both sitting (
Of course, the vertical height adjustment could be done by other means well known in the art, such as a manual foot-pumping hydraulic lift.
Novel attributes of these endpoint devices are listed below and described in further detail in the sections that follow:
In one preferred form of the invention, the endpoint device comprises a single yaw axis which is coincident with a point-of-interest (e.g., the user's hand). By way of example but not limitation, and looking now at
Another aspect of the present invention is the ability to provide a flexible connection between a forearm support (e.g., cradle 805) and the rest of the endpoint device. In this way the endpoint device is able to support the weight of the arm, but allows the user to outstretch their arm without uncomfortable pressure from the rear strap 810. By way of example but not limitation, and looking now at
Another aspect of the present invention is the provision of an adjustable pitch angle that: 1) enables left-hand to right-hand switching, and 2) enables small angular adjustments depending on user size, the workspace of interest, and the type of exercise. By way of example but not limitation, and looking now at
Still another aspect of the present invention is the provision of an off-axis-rotatable hand grip (e.g., ball grip) that enhances comfort while allowing for different hand sizes. By way of example but not limitation, and looking now at
Another feature of the present invention is the inclusion of an electronic hand-presence sensing system. More particularly, in one preferred form of the invention, a capacitive sensing system is provided which detects the presence of the user's limb on the endpoint device and signals the robotic device that a person's limb is (or is not) present on the endpoint device. This is a safety and functionality feature and is particularly important for some endpoint devices, e.g., ball endpoint 800B (
The status of the presence of the user is preferably made clear to the patient and therapist immediately by lighting up ball grip 820 (or another status light, not shown, provided on the endpoint device or elsewhere on robotic device 5) in one of several colors to report status, such as green when the patient engages the device and the device is active, or yellow to indicate that the system is ready to go and awaiting the patient or user. The system may also use audible sounds to help identify or confirm the status of the presence of the user.
By way of example but not limitation, cradle endpoint 800 may have its ball grip 820 configured with a capacitive sensing system which communicates with onboard controller 596 of robotic device 5. Such capacitive sensing systems are well known in the sensor art and are easily adaptable to ball grip 820. In accordance with the present invention, when the user grips ball grip 820, the capacitive sensing system associated with ball grip 820 detects user engagement and advises onboard controller 596 of robotic device 5 that the user is engaged with the endpoint device. Robotic device 5 may then proceed with the therapeutic regime programmed into onboard controller 596 of robotic device 5. However, if the user lets go of ball grip 820, the capacitive sensing system associated with ball grip 820 detects user disengagement and advises onboard controller 596 of robotic device 5 that the user is no longer engaged with the endpoint device. Robotic device 5 may then suspend the therapeutic regime programmed into onboard controller 596 of robotic device 5.
Another aspect of the present invention is the ability to easily “swap out” different endpoints on robotic device 5 and to have electrical connections occur automatically when the mechanical connection between the new endpoint and the robotic device is made. In one preferred form of the invention, this is accomplished with a mechanical latch (e.g., a mechanical latch such as one manufactured by SouthCo of Concordville, Pennsylvania), custom-designed nested tubes, and a floating electrical connector system (e.g., a “Molex Mini-Fit Blindmate” system such as one manufactured by Molex of Lisle, Illinois) which together provide mechanical and electrical connections which are able to account for mechanical misalignment without stressing the electrical connections.
In one preferred form of the invention, a mechanical switch is provided on robotic device 5 that detects the presence (or absence) of an endpoint device. Alternatively, an electrical switch may also be provided to detect the presence (or absence) of an endpoint device. Such mechanical and electrical switches are well known in the sensor art and are easily adaptable to the portion of robotic device 5, which receives outer link 110 of the endpoint devices. Endpoint-presence sensing is important for system safety—if the endpoint should become disconnected from robotic device 5 during operation of robotic device 5, the robotic device 5 can go into a safe (“motionless”) mode until the endpoint is re-attached (or another endpoint is attached in its place).
An important aspect of the modularity of the endpoints is that robotic device 5 is configured so that it can automatically sense and recognize the type of endpoint that is installed on the robotic device. This allows robotic device 5 to automatically adjust its operating parameters according to the particular endpoint which is mounted to the robotic device, e.g., it allows robotic device 5 to adjust various operating parameters such as the kinematics related to endpoint location, gravity-assist calculations (see below), etc. By way of example but not limitation, outer link 110 of each endpoint can comprise an encoded element representative of the type of endpoint and the portion of robotic device 5 which receives outer link 110 can comprise a reader element—when an endpoint is mounted to robotic device 5, the reader element on robotic device 5 reads the encoded element on the mounted endpoint and the reader element appropriately advises onboard controller 596 for robotic device 5.
In one preferred form of the invention, gravity compensation means are provided to make the user's limb feel weightless. This is done by applying an upward bias to the endpoint device which can offset the weight of the user's limb, thereby effectively rendering the user's limb “weightless”. Such gravity compensation may be achieved by having onboard controller 596 read the torque levels on motors 500 and 565 when a user's limb is engaging the endpoint device and then energizing motors 500 and 565 so as to apply an offsetting torque to the motors, whereby to offset the weight of the user's limb. Gravity compensation is important inasmuch as it allows a user to use the system for extended periods of time without tiring. However, this can be complex inasmuch as the weight of different people's limbs are different and because the weight of a single person's limb changes as he/she moves the limb to different locations and activates/adjusts different muscle groups. To this end, the gravity compensation means of the present invention includes various apparatus/algorithms/procedures which involve:
Note that onboard controller 596 may be configured to compensate for the effects of gravity when the endpoint device is engaged by a limb of a user in a single step, or onboard controller 596 may be configured to compensate for the effects of gravity in a series of incremental steps. This latter approach can be advantageous in some circumstances since the gradual application of gravity compensation avoids any surprise to the user. Note also that onboard controller 596 can apply the gravity compensation automatically or onboard controller 596 can apply the gravity compensation under the guidance of an operator (e.g., a therapist).
Robotic device 5 is configured so that it has the ability to easily flip from a right-hand to a left-hand configuration, e.g., using a cam-latch (similar to those found on front bicycle wheels) such as the aforementioned cam-latch 594 which allows outer link 110 of a given endpoint device to be quickly and easily attached to/detached from the remainder of robotic device 5. Furthermore, robotic device 5 has knowledge of the “handedness” of a given endpoint device due to the aforementioned automatic endpoint sensing switches. This allows robotic device 5 to automatically alter the software in its onboard controller 596 to account for the different kinematics of different endpoint devices. The various endpoint devices have been designed to accommodate this flipping and can be used in both right-hand and left-hand configurations.
To change from left-handed use to right-handed use, or vice versa, requires three 180-degree flips about three axes.
By way of example but not limitation, and looking now at
To change back from right-handed use to left-handed use, the flips are performed in the same order, but reversing the directions of the flips.
It is important to note that when the hand grip used with the endpoint device is not a symmetrical shape, or when mounting shaft 865 for ball grip 820 is disposed “off-axis” from the center of ball grip 820 (
In some situations it may be important to allow pronation/supination of the user's forearm/wrist while the user's forearm is strapped to cradle 805. Pronation/supination is the twist/rotation of the wrist about the longitudinal axis of the forearm.
To that end, in one form of the invention, and looking now at
Alternatively, other arcuate bearings of the sort well known in the bearing art may also be used.
However, the use of such Kaydon-style ring bearings and other arcuate bearings can increase the cost of the endpoint device.
Therefore, in another preferred embodiment of the present invention, and looking now at
In the foregoing sections, robotic device 5 was described as having a coupling element 115 for coupling outer link 110 to a patient, commonly to a limb of a patient, with outer link 110 being detachably connected to the remainder of the robotic device at the aforementioned mechanism 590 (
Various embodiments of cradle endpoints (e.g., cradle endpoint 800, cradle endpoint with actuated or spring-based hand-grip assist 800C, etc.) have been previously described in the foregoing sections. These previously-described cradle endpoints generally comprise a padded cradle 805 for receiving and supporting a limb (e.g., the forearm) of a patient, straps 810 for securing the limb to cradle 805, a connector 815 for connecting cradle 805 to outer link 110, a hand grip (e.g., ball grip 820) for gripping by the patient (e.g., by the hand of the patient), and multiple passive and manually-lockable degrees-of-freedom for making adjustments and for enabling a large range-of-motion. Each of the previously-described cradle endpoints are designed to be swapped in and out of robotic device 5 so as to allow patients with different size limbs, with different functional capabilities, and different therapeutic goals, to use robotic device 5. Furthermore, each of the hand grips provided with the various embodiments of endpoints are designed to be swapped in and out of the endpoint so as to allow patients with different size limbs, with different functional capabilities, and different therapeutic goals, to use robotic device 5.
In the following section, and looking now at
Endpoint 1000 generally comprises a cradle 1005 for receiving a limb (e.g., the forearm) of a patient, straps 1010 passing through slots 1012 for securing the limb to cradle 1005, a connector 1015 for connecting cradle 1005 to outer link 110, and the aforementioned outer link 110. Cradle endpoint 1000 preferably also comprises a stick grip 1020 for gripping by the patient (e.g., by the hand of a patient). If desired, a cushioned foam pad (not shown in
Cradle 1005 and stick grip 1020 are configured to move along a first yaw axis 1030 and a second yaw axis 1033, whereby to permit a limb of a user to swivel from left and right (i.e., along the flexion/extension axis of the wrist). Note that connector 1015 comprises a first portion 1035 for connection to outer link 110, and a second portion 1040 for connection to cradle 1005 and stick grip 1020. Preferably, a leaf spring 1050 is provided between cradle 1005 and second portion 1040, whereby to enable flexibility and allow a patient's arm to lift up during certain three-dimensional motions.
Another aspect of the present invention is the provision of a mechanism for permitting the pitch angle of cradle 1005 and connector 1015 to be adjusted along pitch axis 1045 relative to outer link 110, whereby to 1) enable left-hand to right-hand switching, and 2) enable small angular adjustments depending on user size, the workspace of interest, and the type of exercise. By way of example but not limitation, a cam-lever 1060 may be provided to allow the angular disposition of first portion 1035 to be adjusted relative to outer link 110. Cam-lever 1060 may be released to unlock first portion 1035 from outer link 110, whereby first portion 1035 can be adjusted (e.g., rotated about pitch axis 1045) relative to outer link 110, and then cam-lever 1060 may be re-locked once first portion 1035 is in the desired angular position.
As stated above, a unique feature of endpoint 1000 is the provision of an additional degree of freedom along a roll axis, whereby to enable passive and active pronation and supination of the wrist of a user. In order to provide this additional degree of freedom, stick grip 1020 is mounted to a rotatable plate 1065. Rotatable plate 1065 is free to rotate under the influence of a user's own power, however, rotatable plate 1065 is also configured to be rotated by an electric motor 1070 contained within a motor housing 1075 and connected to rotatable plate 1065. When actuated, motor 1070 rotates rotatable plate 1065 and stick grip 1020 along roll axis 1080, whereby to pronate and supinate the wrist of a user gripping stick grip 1020. Preferably, a geared transmission is provided within motor housing 1075 for reducing the speed of motor 1070 to rotatable plate 1065. If desired, motor housing 1075 can include a protective cover 1085 for protecting the user and/or healthcare professionals from the heat of the motor contained in motor housing 1075. Furthermore, if desired, a protective shield 1090 may be disposed around stick grip 1020 for covering potential finger pinch points as stick grip 1020 is rotated along roll axis 1080. Protective shield 1090 is preferably connected to rotatable plate 1065 so that protective shield 1090 rotates with rotatable sheath 1065 and stick grip 1020 as rotatable sheath 1065 and stick grip 1020 rotate.
In an additional embodiment of the present invention, a second motor (not shown) may be provided to enable powered movement of stick grip 1020 along second yaw axis 1033, whereby to provide powered movement (i.e., flexion and extension) of a wrist along second yaw axis 1033. Powered movement along second yaw axis 1033 can be beneficial to users who are unable to swivel their wrist from left to right under their own power due to a physical impairment.
As was discussed above in connection with the previously-discussed endpoints, endpoint 1000 can be easily “swapped out” for different endpoints on robotic device 5, with the electrical connections occurring automatically when the mechanical connection between the new endpoint and robotic device 5 is made. To this end, it is noted that mechanical and electrical connections between endpoint 1000 and robotic device 5 are made with a quick connect-disconnect mechanism 1100. Quick connect-disconnect mechanism 1100 comprises a mechanical fitting 1105 and an electrical port 1110 which together mechanically and electrically connect outer link 110 to coupling element 115 of robotic device 5. A threaded ring 1115 may be used to further secure mechanical fitting 1105 (and thus outer link 110) to coupling element 115.
Note that in
Note also that in this form of the invention, U-shaped frame 140 may be supported above base 100 via a telescoping assembly 827 which allows the height of U-shaped frame 140 (and hence the height of the robotic arm) to be adjusted relative to base 100. This feature is highly advantageous, since it facilitates the use of robotic device 5 with patients who are both sitting (
Of course, the vertical height adjustment could be done by other means well known in the art, such as a manual foot-pumping hydraulic lift.
As noted above, robotic device 5 is specifically configured so that it has the ability to easily flip from a right-hand to a left-hand configuration, e.g., using a cam-latch (similar to those found on bicycle wheels) such as the aforementioned cam-latch 594 which allows outer link 110 of a given endpoint device to be quickly and easily attached to/detached from the remainder of robotic device 5. Furthermore, robotic device 5 has knowledge of the “handedness” of a given endpoint device due to the aforementioned automatic endpoint sensing switches. This allows robotic device 5 to automatically alter the software in its onboard controller 596 to account for the different kinematics of different endpoint devices. The various endpoint devices have been designed to accommodate this flipping and can be used in both right-hand and left-hand configurations.
To change endpoint 1000 from left-handed use to right-handed use, or vice versa, requires three 180-degree flips about three axes.
By way of example but not limitation, the process of changing endpoint 1000 from left-handed use to right-handed use will now be described. First, the clamping mechanism connecting outer link 110 to inner link 105 (e.g., lever 593 shown in
To change back from right-handed use to left-handed use, the flips are performed in the same order, but reversing the directions of the flips.
In use, endpoint 1000 is mechanically and electrically connected to robotic device 5 by connecting mechanical fitting 1105 and electrical port 1110 of outer link 110 to tubular member 595 (
In one preferred form of the invention, stick grip 1020 may be provided with an electronic hand-presence sensing system. More particularly, a capacitive sensing system is provided which detects the presence of the user's limb on stick grip 1020 and signals the robotic device that a person's limb is (or is not) present on stick grip 1020. By way of example but not limitation, endpoint 1000 may have its stick grip 1020 configured with a capacitive sensing system which communicates with onboard controller 596 of robotic device 5. Such capacitive sensing systems are well known in the sensor art and are easily adaptable to stick grip 1020. In accordance with the present invention, when the user grips stick grip 1020, the capacitive sensing system associated with stick grip 1020 detects user engagement and advises onboard controller 596 of robotic device 5 that the user is engaged with the endpoint device. Robotic device 5 may then proceed with the therapeutic regime programmed into onboard controller 596 of robotic device 5. However, if the user lets go of stick grip 1020, the capacitive sensing system associated with stick grip 1020 detects user disengagement and advises onboard controller 596 of robotic device 5 that the user is no longer engaged with the endpoint device. Robotic device 5 may then suspend the therapeutic regime programmed into onboard controller 596 of robotic device 5.
In another form of the invention, stick grip 1020 may also, or alternatively, be provided with an electronic force sensing system. More particularly, a force sensing system may be provided to detect the force of the grip of the user's hand on stick grip 1020 and signal to the robotic device how much force the user's hand is providing to stick grip 1020.
The hand-presence sensing system and the force sensing system described above with respect to stick grip 1020 may also be implemented in any of the hand grips used with the previously-described endpoints (e.g., ball grip 820, ball grip 820B, actuated or spring-biased hand-grip 820C, etc.).
If desired, a contoured foam pad (not shown) could be positioned on cradle 1005 so as to provide a space under the wrist of the user which would allow the user to pronate and/or supinate their wrist without their wrist rubbing against the foam pad.
Furthermore, if desired, one or more of straps 1010 may be omitted so that the wrist has more freedom to rotate (i.e., pronate and supinate).
In another embodiment of the present invention, and looking now at
In another embodiment of the present invention, and looking now at
Angled handlebar grip 1150 is designed to be used in both right-hand and left-hand configurations. In a preferred form of the present invention, angled handlebar grip 1150 is mounted to rotatable base plate 1095 with a magnetic connection so as to enable angled handlebar grip 1150 to be rotated along yaw axis 1030 when angled handlebar grip 1150 is switched from right-handed use to left-handed use.
By way of example but not limitation, the process of changing angled handlebar grip 1150 from left-handed use to right-handed use will now be described. First, the clamping mechanism connecting outer link 110 to inner link 105 (e.g., lever 593 shown in
To change back from right-handed use to left-handed use, the flips are performed in the same order, but reversing the directions of the flips.
It is important to note that angled handlebar grip 1150 may be used as an alternative to any of the hand grips shown with the previously-described endpoints (e.g., ball grip 820, ball grip 820B, actuated or spring-biased hand-grip 820C, stick grip 1020, etc.). Preferably, the hand grips are mounted to the endpoint device (e.g., to base plate 1095) through a magnetic connection so as to enable one hand grip to be easily swapped in for another hand grip.
Furthermore, while angled handlebar grip 1150 of
In the foregoing disclosure, there is disclosed a novel multi-active-axis, non-exoskeletal robotic device for providing physical therapy and occupational therapy (sometimes collectively referred to herein as “physical therapy/occupational therapy” and/or simply “therapy”) to a patient.
In one form of the invention, the robotic device is configured to provide game-based rehabilitation. In this form of the invention, the patient views a two-dimensional (2D) or three-dimensional (3D) scene using a computer screen, a projector, glasses, goggles, or similar means. The 2D or 3D scene depicts a game which the patient “plays” by moving their limb (which is connected to the robotic device) so as to cause corresponding movement of a virtual object (or virtual character) within the 2D or 3D scene. As the patient endeavors to appropriately move their limb so as to cause appropriate movement of the virtual object (or virtual character) within the 2D or 3D scene of the game, the patient “effortlessly” participates in the therapy process. This form of the invention is a powerful tool, since it promotes increased engagement of the patient in the therapy process, and thereby yields higher “dosages” of the physical therapy or occupational therapy, which is known to be an essential element in successful recovery from stroke and many other injuries and diseases.
If desired, the 2D or 3D scene may take another non-game form, i.e., the 2D or 3D scene may be a non-game graphical or textual display, with the patient endeavoring to appropriately move their limb (which is connected to the robotic device) so as to cause appropriate movement of a virtual object within a graphical or textual display. This non-game approach, while less engaging for the patient than the game-based physical therapy or occupational therapy described above, is nonetheless capable of providing a valuable assessment measure.
In both of the foregoing forms of the invention, the patient is essentially endeavoring to appropriately move their limb (which is connected to the endpoint of the robotic device) so as to cause corresponding appropriate movement of a virtual object (or virtual character) on a computer screen, projector, glasses, goggles or similar means.
While the foregoing approaches provide excellent therapy for the patient, they do not lend themselves to Activity Based Training (ABT). With ABT, the patient learns to accomplish an important daily activity, e.g., feeding themselves with a spoon.
To this end, in another form of the present invention, the robotic device is configured so that the therapist guides (e.g., manually assists) the patient in moving their limb (which is connected to the robotic device) through a desired motion (e.g., feeding themselves with a spoon). As this occurs, the robotic device “memorizes” the desired motion (i.e., by recording the movements of the various segments of the robotic device), and then the robotic device thereafter assists the patient in repeating the desired motion, e.g., by helping carry the weight of the patient's limb and by restricting motion of the patient's limb to the desired path. Thus, with the robotic device operating in this activity-based mode, the patient is manipulating a real object in real space (and is not manipulating a virtual object on a computer screen, as with the game-based physical therapy).
However, it should be appreciated that the robotic device is also configured so that activity-based therapy may be provided without requiring physical intervention from the therapist, as it may be sufficient for the robotic device to simply suspend some fraction of the weight of the patient's limb, thereby allowing the patient to succeed at a given activity. The robotic device may also be provided with pre-conceived therapy modalities that go beyond just simply limb suspension, such as a generalized pre-defined path along which the patient movement is constrained, so that the robotic device acts in the sense of a guide.
In the preceding description, the present invention is generally discussed in the context of its application for a rehabilitation device. However, it will be appreciated that the present invention may also be utilized in other applications, such as applications requiring high-fidelity force feedback. By way of example but not limitation, these applications may include use as an input/haptic feedback device for electronic games, as a controller for other mechanical devices such as industrial robotic arms and/or construction machines, or as a device for sensing position, i.e., as a digitizer or coordinate-measuring device.
In another embodiment of the invention, the invention comprises a novel computer-based assisted therapy system comprising at least one camera for monitoring the patient during therapy and an AI-based platform for analyzing data provided by the at least one camera.
More particularly, in this form of the invention, the novel computer-based assisted therapy system is configured to (i) utilize facial-recognition technology to identify a patient and link/record data concerning that patient to an electronic medical record particular to that patient, (ii) track movements of one or more patients during robot-assisted therapy in order to identify compensatory movements that can detract from therapy and notify the therapist of the same, (iii) track movements of one or more patients during robot-assisted therapy in order to perform real-time assessments of patient progress during therapy, and (iv) facilitate group robot-assisted therapy sessions in which a single therapist supervises a plurality of patients and the system acts to enhance patient safety while simultaneously providing diagnostic tools for enhancing therapy.
To that end, and looking now at
Display 1215 may comprise a conventional LCD (or similar) flat screen monitor. Alternatively, display 1215 may comprise a virtual reality (VR)-enabled headpiece/goggles configured for mounting to patient P's head such that the at least one display is disposed directly in the field of view of patient P.
Camera 1220 may comprise a single camera (e.g., an Intel RealSense camera sold by Intel Corporation of Santa Clara, CA, USA), or camera 1220 may comprise a plurality of cameras (e.g., two monocular cameras).
Computer system 1225 comprises memory 1230 (e.g., non-transitory memory) comprising appropriate instructions for processing data and a microprocessor 1235 for use with the same, as will hereinafter be discussed in further detail.
In a preferred form of the invention, camera 1220 is used to identify a patient using facial recognition software and detect movement of the patient (e.g., movement of the patient's limbs, torso, head, etc.) in response to a prompt shown on display 1215. By using camera 1220 to detect movement of the patient, an AI-powered image processing tool can then be used to (i) automatically detect when patients perform undesirable compensatory movements during robot-assisted rehabilitation exercises, (ii) track progress of patient progress on conventional scales (e.g., the Wolf Motor Function Test, the Function Ability Scale, and the Fugl-Meyer Assessment) during robot-assisted rehabilitation exercises, and/or (iii) facilitate group robot-assisted therapy sessions in which a single therapist supervises a plurality of patients, as will be discussed in further detail below.
To this end, the present invention comprises a pretrained biomechanical model (sometimes hereinafter referred to as “OpenPose”) configured to detect 2D poses of humans in an image. See
In order to facilitate automated monitoring of the data received from camera 1220 and processed by the OpenPose biomechanical model, computer-based assisted therapy system 1200 further comprises an AI-based movement detection system 1240. AI-based movement detection system 1240 may be trained to recognize limb movements of patient P according to various approaches that will be apparent to one of skill in the art in view of the present disclosure.
In one form of the invention, and looking now at
In another form of the invention, and still looking at
Regardless of the approach employed to train AI-based movement detection system 1240, the result is that the system is configured to recognize not only movement of patient limbs as determined from camera 1220, but additionally the posture of the patient, i.e., the overall positioning of a plurality of limbs in the image obtained from camera 1220. Thus, it is possible to utilize camera 1220 and AI-based movement detection system 1240 to autonomously monitor patient P's limb movements and resulting posture during use of robot-assisted therapy device 1205, as will hereinafter be discussed in further detail.
Once AI-based movement detection system 1240 has been trained to identify a patient using facial recognition software, to recognize movement of patient limbs and to recognize resulting postures from video data, computer-based assisted therapy system 1200 is configured to utilize camera 1220, computer system 1225 and AI-based movement detection system 1240 to monitor a patient P during the performance of therapy delivered by a robot-assisted therapy device (e.g., the aforementioned multi-active axis, non-exoskeletal robotic device 5).
Specifically, AI-based movement detection system 1240 receives data from camera 1220 (i.e., image data of the patient P performing the therapeutic movements) and uses that data to determine the posture of patient P during the therapy, whereby to recognize when patient P engages in compensatory movement (defined as any movement that deviates from a desired therapeutic movement).
Looking now at
As discussed above, AI-based movement detection system 1240 may be configured to monitor for compensatory movements by patients P during therapy. In a preferred form of the invention, AI-based movement detection system 1240 interfaces with a centralized computer system 1225 so that data concerning patient movements is centralized in a single system. Appropriate software records patient movements as a function of therapy session time and generates a session report 1245 comprising metrics that relate to the compensatory-movement performance of the patient for storage in a session-report module 1250 (
If desired, clips of video showing the patient engaging in compensatory movements may be provided to therapist T to review and/or to show to the patient P and/or shown on session report 1245 and/or stored in session-report module 1250. Computer system 1225 is preferably configured to identify compensatory movements by a patient P in real-time and to provide a signal (e.g., a visual signal, an audible signal, a haptic signal, etc.) to therapist T so that therapist T can focus attention on a patient P engaging in compensatory movements and correct those movements in real-time. Alternatively and/or additionally, the therapist can review the session report stored in the session report module after a therapy session in order to plan future therapy sessions with a particular patient P. In a preferred form of the invention, session-report module 1250 is configured to generate information that includes statistics of identified compensatory strategies engaged in by the particular patient P during the therapy session, as well as video-clips with detailed 3D skeletons for further review by therapist T and/or patient P.
In a preferred embodiment of the invention, and looking now at
In a preferred embodiment of the invention, AI-based movement system 1240 preferably also comprises an offline-guidance module 1265 (
In order to determine whether therapy is working to benefit a particular patient (or whether continued therapy would benefit the particular patient), it is common to periodically halt therapy in order to perform assessments of the patient to ascertain progress according to certain assessment scales that are commonly used in the art. By way of example but not limitation, patients may be assessed for progress according to the Wolf Motor Function Test (WMFT), the Function Ability Scale (FAS), and the Fugl-Meyer Assessment (FMA).
Looking now at
Regardless of the assessment utilized, computer system 1225 is configured to provide a clinical score for one or more of the assessments during a therapy session particular to the patient performing the therapeutic movement, and to report the same to therapist T (e.g., in the form of a line graph with a plurality of points representing clinical scores graphed). See
As a result, therapist T is able to adjust the therapeutic movements performed by the patient in real-time as the patient progresses and their clinical scores improve (e.g., so that when a patient “plateaus” on a particular movement, therapist T can make the movement more challenging, etc.).
Stroke survivors frequently need upper-extremity rehabilitation that supports both unimanual and bimanual task training. To that end, computer-based assisted therapy system 1200 may be used to detect compensatory-movement strategies adopted by a patient during robot-assisted upper-extremity rehabilitation.
To that end, and looking now at
By tracking both the active limb and the contralateral, unengaged limb of patient P at the same time, interactive games (e.g., games displayed on display 1215 that include moving objects in a virtual space, etc.) that use unimanual and bimanual task training can be integrated into the therapy session. Importantly, when a modular endpoint such as modular endpoint 1000 is used with robot-assisted therapy device 1205, AI-based movement detection system 1240 may be configured to recognize and provide estimates of complex movements such as forearm pronation/supination. This can permit more complex, realistic games to be displayed on display 1215, whereby to enhance the therapy.
By way of example but not limitation, if desired, computer-based assisted therapy system 1200 may be used to provide therapy to stroke victims. With this form of the invention, computer-based assisted therapy system 1200 is used to track the position and orientation of a patient P's body segments by virtue of their engagement with arm 1210 of robot-assisted therapy device 1205 during the performance of bimanual training tasks that require tracking both a stroke-affected upper limb (which engages arm 1210) and the contralateral upper limb (which does not engage robot-assisted therapy device 1205). If desired, one or more “games” may be displayed on display 1215, with computer-based assisted therapy system 1200 tracking limb movement and moving objects displayed on display 1215 in real time to permit the patient to interact with the game.
By way of example but not limitation, an exemplary game may be a “Master Cook” game which simulates cooking activities in which the patient is asked to follow a recipe. First, instructions are provided through an interactive video tutorial displayed on display 1215, such as “cook the pasta until al dente; add cream cheese, pasta cooking water, parmesan, and stir well; drain and add pasta to the skillet; toss until well combined, adding some pasta water if needed; serve with parmesan cheese, black pepper and olive oil. Patient P is tasked with tasks such as setting a timer, i.e., a virtual timer displayed on display 1215 with which the patient engages by moving their limbs such that the movement is seen by camera 1220 and system 1200 acts to move the virtual object in concert with the real world movement of the patient's limbs.
Computer-based assisted therapy system 1200 is disclosed above as used in concert with a robot-assisted therapy device 1205, which robotic-assisted therapy device may be in the form of the aforementioned multi-active-axis, non-exoskeletal robotic device 5.
However, it should be appreciated that computer-based assisted therapy system 1200 (and/or AI-based movement detection system 1240) may be used to monitor patients performing therapeutic movement on substantially any therapy device that requires the patient to move their limbs. That is, although the novel system disclosed above is disclosed in the context of use with a robot-assisted therapy device for performing upper extremity therapy, the novel system of the present invention is not intended to be used only with the aforementioned robot-assisted therapy devices and/or only for upper extremity therapy. The present invention may be used with substantially any device the facilitates therapeutic movement of the patient's limbs (upper or lower extremities) and which would benefit from tracking movement of the patient's limbs (or torso, etc.) in real time.
It should be understood that many additional changes in the details, materials, steps and arrangements of parts, which have been herein described and illustrated in order to explain the nature of the present invention, may be made by those skilled in the art while still remaining within the principles and scope of the invention.
This patent application: (1) is a continuation-in-part of pending prior U.S. patent application Ser. No. 16/778,902, filed Jan. 31, 2020 by Barrett Technology, LLC and David D. Wilkinson et al. for MULTI-ACTIVE-AXIS, NON-EXOSKELETAL ROBOTIC REHABILITATION DEVICE (Attorney's Docket No. BARRETT-14), which patent application, in turn: (i) is a continuation-in-part of pending prior U.S. patent application Ser. No. 16/066,189, filed Sep. 30, 2016 by Barrett Technology, LLC and William T. Townsend et al. for MULTI-ACTIVE-AXIS, NON-EXOSKELETAL REHABILITATION DEVICE (Attorney's Docket No. BARRETT-0810 PCT US, which patent application: (a) is a 371 of International (PCT) Patent Application No. PCT/US2016/054999, filed Sep. 30, 2016 by Barrett Technology, LLC and William T. Townsend et al. for MULTI-ACTIVE-AXIS, NON-EXOSKELETAL REHABILITATION DEVICE (Attorney's Docket No. BARRETT-0810 PCT), which patent application claims benefit of (i) prior U.S. Provisional Patent Application Ser. No. 62/235,276, filed Sep. 30, 2015 by Barrett Technology, Inc. and Alexander Jenko et al. for MULTI-ACTIVE-AXIS, NON-EXOSKELETAL REHABILITATION DEVICE (Attorney's Docket No. BARRETT-8 PROV), and (ii) prior U.S. Provisional Patent Application Ser. No. 62/340,832, filed May 24, 2016 by Barrett Technology, LLC and William T. Townsend et al. for MULTI-ACTIVE-AXIS, NON-EXOSKELETAL REHABILITATION DEVICE (Attorney's Docket No. BARRETT-10 PROV);(b) is a continuation-in-part of prior U.S. patent application Ser. No. 14/500,810, filed Sep. 29, 2014 by Barrett Technology, LLC and William T. Townsend et al. for MULTI-ACTIVE-AXIS, NON-EXOSKELETAL REHABILITATION DEVICE (Attorney's Docket No. BARRETT-5), which patent application claims benefit of prior U.S. Provisional Patent Application Ser. No. 61/883,367, filed Sep. 27, 2013 by Barrett Technology, Inc. and William T. Townsend et al. for THREE-ACTIVE-AXIS REHABILITATION DEVICE (Attorney's Docket No. BARRETT-5 PROV);(ii) claims benefit of prior U.S. Provisional Patent Application Ser. No. 62/799,502, filed Jan. 31, 2019 by Barrett Technology, LLC and Michael Schiess et al. for A MOTORIZED END-EFFECTOR ENABLING WRIST PRONATION AND SUPINATION ON AN UPPER-EXTREMITY ROBOTIC THERAPY SYSTEM (Attorney's Docket No. BARRETT-14 PROV); and(2) claims benefit of pending prior U.S. Provisional Patent Application Ser. No. 63/403,107, filed Sep. 1, 2022 by Barrett Technology, LLC and William T. Townsend et al. for 3D CAMERA APPLIED TO REHABILITATION AND SPORTS TRAINING (Attorney's Docket No. BARRETT-20A PROV). The nine (9) above-identified patent applications are hereby incorporated herein by reference.
Number | Date | Country | |
---|---|---|---|
62235276 | Sep 2015 | US | |
62340832 | May 2016 | US | |
61883367 | Sep 2013 | US | |
62799502 | Jan 2019 | US | |
63403107 | Sep 2022 | US |
Number | Date | Country | |
---|---|---|---|
Parent | 16778902 | Jan 2020 | US |
Child | 18241486 | US | |
Parent | 16066189 | Jun 2018 | US |
Child | 16778902 | US | |
Parent | 14500810 | Sep 2014 | US |
Child | 16066189 | US |