Hand function rehabilitation device with intention perception function
By introducing intention perception functions and multiple training modes into the hand function rehabilitation device, the problems of large size, complex structure and poor training effects of hand function rehabilitation robots in the prior art are solved, and more efficient rehabilitation training and evaluation are achieved.
Patent Information
- Application Number
- CN202180003357.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-11-11
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2041-11-11
AI Technical Summary
The existing hand-function rehabilitation robots have problems such as large size, poor portability, complex structure, low comfort, few repetitive active training functions, single training mode and unintentional perception, resulting in unsatisfactory rehabilitation results.
A hand-function rehabilitation device with intention perception function was designed, including a power module, intention perception module, rehabilitation training module and evaluation and analysis module. The driving force of the power module is transmitted to the glove body through the transmission mechanism, and the finger is analyzed using the intention perception module. Exercise intention, provides auxiliary driving force, and has a variety of rehabilitation training modes and real-time evaluation functions.
The device structure is simplified, the comfort of use is improved, the active training of patients is realized, the rehabilitation training effect is improved, and real-time rehabilitation training evaluation is provided.
Smart Images

Figure CN114206292B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of rehabilitation robots, and in particular to a hand function rehabilitation device with intention perception function. Background Art
[0002] According to the "2019 China Stroke Prevention and Treatment Report", the total number of stroke patients in my country exceeds 13 million, and about 75% of survivors suffer from hemiplegia due to motor nerve damage, leading to motor dysfunction. Nearly 80% of stroke patients will experience motor dysfunction of hand dysfunction, which is one of the most common symptoms of stroke. At present, the main methods used for improvement and rehabilitation of hand dysfunction at home and abroad are to repair or reshape damaged nerves, strengthen muscle tendon strength training, and complete motor learning to rebuild or improve the motor function of the affected hand.
[0003] The clinical symptoms of hand dysfunction are mainly manifested as decreased muscle strength, decreased or increased muscle tension, paresthesia, decreased joint mobility, muscle atrophy, hand swelling and pain, etc., which in turn manifest as the inability to achieve simple grasping functions in daily life. This is mainly due to the loss of control of the lower centers by the higher centers of the brain and the loss of inhibition of the original reflexes of the lower centers. If patients with hand dysfunction undergo effective rehabilitation training in the later stages of rehabilitation, it is possible to avoid more serious complications or secondary disabilities, reduce the degree of disability, and prevent secondary disability caused by stroke. With the rapid development of science and technology and the increase in the number of patients with hand dysfunction, robotics technology has begun to be applied to the field of rehabilitation medicine, and hand function rehabilitation robots have emerged. The emergence of hand function rehabilitation robots has alleviated the impact of problems such as the shortage of medical resources, the high cost of one-on-one physicians, and the influence of differences in physicians' treatment methods and experience on the treatment effect in the process of hand rehabilitation training.
[0004] At present, there are many hand function rehabilitation robot products developed and applied in the process of hand rehabilitation training at home and abroad. However, there are still many challenges to overcome in actual use. At present, there are several problems with the products for hand rehabilitation training for stroke on the market: 1) large size and poor portability; 2) complex structure and low comfort; 3) few products with repeated active training functions, which cannot effectively restore the affected hand; 4) single training mode and unsatisfactory rehabilitation effect; 5) no intention perception, unable to evaluate the patient's rehabilitation status in real time.
[0005] Therefore, a more reliable solution is needed now. Summary of the invention
[0006] The technical problem to be solved by the present invention is to provide a hand function rehabilitation device with intention perception function in view of the deficiencies in the above-mentioned prior art.
[0007] In order to solve the above technical problems, the technical solution adopted by the present invention is: a hand function rehabilitation device with intention perception function, comprising:
[0008] A host computer, which includes a power module, an intention perception module, a rehabilitation training module, an evaluation and analysis module, and a control module;
[0009] The glove body has five independent single-finger sleeves that cooperate with the fingers;
[0010] and a transmission mechanism connected between the host and the glove body, the transmission mechanism transmitting the driving force output by the power module to the glove body to drive the single finger sleeve of the glove body to perform bending and stretching movements;
[0011] Among them, the intention perception module determines the user's finger movement intention by analyzing the collected pressure signals, bending angle signals of the user's fingers and the forearm electromyographic signals of the corresponding hand, thereby providing auxiliary driving force for the finger movement through the power module.
[0012] Preferably, the power module includes 5 power output mechanisms, which include a power housing, a motor disposed in the power housing, a screw connected to the motor, a screw nut sleeved on the screw, a pulley rotatably disposed inside the power housing, and a traction rope wound around the pulley;
[0013] After the traction rope is wound around the pulley, it forms two sections: a first traction rope section and a second traction rope section. The movable ends of the first traction rope section and the second traction rope section form two source output ends: a first source output end and a second source output end.
[0014] The first traction rope segment is fixedly connected to the lead screw nut, and the second traction rope segment can slide through the lead screw nut, so that the first source output end and the second source output end respectively generate a pulling force output toward the pulley through the reciprocating linear motion of the lead screw nut on the lead screw.
[0015] Preferably, the transmission mechanism comprises a transmission housing, five groups of transmission units arranged in the transmission housing, five groups of steel wire lasso units matched one by one with the input ends of the five groups of transmission units, and five groups of cored wire units matched one by one with the output ends of the five groups of transmission units;
[0016] A set of steel wire lasso units is used to transmit the driving force output by a power module to a set of transmission units, and then finally transmit the output to a single-finger sleeve on the glove body through a set of core wire units;
[0017] A transmission unit group includes an upper transmission member and a lower transmission member, a wire lasso unit group includes two wire lassoes respectively drivingly connected to the input ends of the upper transmission member and the lower transmission member: a first wire lasso and a second wire lasso, and a core wire unit group includes two core wires respectively drivingly connected to the output ends of the upper transmission member and the lower transmission member: a first core wire and a second core wire.
[0018] Preferably, the steel wire lasso comprises a spring tube, a polytetrafluoroethylene sleeve inserted in the spring tube, and a steel wire slidably inserted in the polytetrafluoroethylene sleeve, and the core wire comprises a sheath tube and a traction wire slidably inserted in the sheath tube;
[0019] The input end of the steel wire of the first steel wire lasso is connected to the output end of the first source, the output end of the steel wire of the first steel wire lasso is connected to the input end of the upper transmission member, the output end of the upper transmission member is connected to the input end of the traction wire of the first cored wire, and the output end of the traction wire of the first cored wire is connected to the back of a finger of a single finger sleeve;
[0020] The input end of the steel wire of the second steel wire lasso is connected to the output end of the second source, the output end of the steel wire of the second steel wire lasso is connected to the input end of the lower transmission member, the output end of the lower transmission member is connected to the input end of the traction wire of the second cored wire, and the output end of the traction wire of the second cored wire is connected to the fingertip of a single finger sleeve.
[0021] Preferably, the transmission housing comprises an upper cover, a lower cover and a mounting block arranged between the upper cover and the lower cover, at least five first T-shaped slide grooves are respectively provided on the inner walls of the upper cover and the lower cover, at least five second T-shaped slide grooves are respectively provided on the upper and lower surfaces of the mounting block; the upper and lower adjacent first T-shaped slide grooves and second T-shaped slide grooves form a slideway;
[0022] The upper transmission member and the lower transmission member have the same structure and are symmetrically arranged in the transmission housing. The upper transmission member and the lower transmission member each include a first slider and a second slider.
[0023] The first slider includes a T-shaped slider slidably disposed in the first T-shaped slide groove, an insert block connected to the front end of the T-shaped slider, and a first connecting hole penetrating the T-shaped slider and the insert block for connecting to the input end of the traction wire;
[0024] The second sliding block includes a T-shaped base, two blocking bars connected to the base in a direction parallel to the insert block, and a second connecting hole provided on the base for connecting to the output end of the steel wire;
[0025] The base can be slidably arranged in the second T-shaped slide groove, and a slot is formed between the two baffles. After the insert block is inserted into the slot, the end of the baffle contacts the side of the T-shaped slider, so that the first slider and the second slider can slide back and forth together in the slideway to realize the transmission of traction force between the wire lasso and the core wire.
[0026] Preferably, the output end of the steel wire of the first steel wire lasso is connected to the second connecting hole of the second slider of the upper transmission member, and the input end of the traction wire of the first cored wire is connected to the first connecting hole of the first slider of the upper transmission member;
[0027] The output end of the steel wire of the second steel wire lasso is connected to the second connecting hole of the second slider of the lower transmission member, and the input end of the traction wire of the second cored wire is connected to the first connecting hole of the first slider of the lower transmission member.
[0028] Preferably, the host further comprises a host housing, a touch display screen arranged on the host housing, and an electrical stimulation module connected to the rehabilitation training module, and the power module, the intention perception module, the rehabilitation training module, the evaluation and analysis module, and the control module are all arranged in the host housing;
[0029] The intention perception module includes a signal acquisition unit and an intention judgment model, wherein the signal acquisition unit includes a pressure sensor arranged on a single finger sleeve for collecting fingertip pressure signals, a bending sensor arranged on the single finger sleeve for collecting finger bending angle signals, and a surface electromyography armband worn on the user's arm for collecting forearm electromyography signals;
[0030] The intention judgment model is a classifier based on a random forest model, which receives signals collected by the pressure sensor, bending sensor and surface electromyography armband, and judges the finger movement intention through analysis.
[0031] Preferably, the intention judgment model includes a preprocessing module, a feature extraction module, a first classifier and a second classifier, and the working method of the intention judgment model includes the following steps:
[0032] 1) Train the intention judgment model;
[0033] 1-1) Four groups of training gestures were designed in advance: the first group included fist, open hand and finger bending, which were used to show different movements of all fingers together; the second group included bending of the index finger, middle finger, ring finger and pinky together, bending of the middle finger, ring finger and pinky together, bending of the thumb, ring finger and pinky together, bending of the thumb, middle finger, ring finger and pinky with the index finger pointing to something, and bending of the thumb, index finger and middle finger to simulate grasping, which were used to show the hand flexibility in performing different actions; the third group included bending of the thumb, and bending of the thumb and index finger, which were used to show the movement of the thumb and the coordination of the thumb and index finger; the fourth group included bending of each finger separately, which was used to show the movement of a single finger;
[0034] 1-2) Collecting training data: The subject wears the glove body and performs gesture actions according to the group training gestures in step 1-1), and collects signals through the signal acquisition unit at the same time, specifically: collecting fingertip pressure signals through the pressure sensor, collecting finger bending angle signals through the bending sensor, and collecting forearm electromyography signals through the surface electromyography arm ring; each subject collects multiple groups of training data, and the training data of multiple subjects are collected, so as to obtain N groups of training data in total;
[0035] The collected training data was first preprocessed using a Butterworth bandpass filter, and then feature extraction was performed using the segmentation overlap method. The features extracted from the pressure signal included the absolute average value, recorded as MAV1; the features extracted from the bending angle signal included the absolute average value, recorded as MAV2; the features extracted from the forearm electromyographic signal included: absolute average value MAV3, root mean square RMS, electromyographic variance VAV, integrated electromyographic value iEMG, simple square integral SSI, and wavelength WL;
[0036] 1-4) Taking the feature combination MAV1, MAV2, MAV3 and RMS extracted from the i-th group of training data as the input of the first classifier, taking the training gesture group corresponding to the i-th group of training data as the output of the first classifier, and training the first classifier; then taking the feature combination MAV1, MAV2, MAV3, RMS, VAV, iEMG, SSI, WL extracted from the i-th group of training data and the output result of the first classifier as the input of the second classifier, taking the specific training gesture corresponding to the i-th group of training data as the output of the second classifier, and training the second classifier;
[0037] 1-5) According to the method of step 1-4), the first classifier and the second classifier are trained by using N sets of training data, and finally a trained intention judgment model is obtained;
[0038] 2) The user wears the glove body for rehabilitation training, and the preprocessing module receives the signals collected by the pressure sensor, the bending sensor and the surface electromyography armband, and first preprocesses them using a Butterworth bandpass filter before transmitting them to the feature extraction module;
[0039] The feature extraction module uses the segmentation overlap method to extract the following features: MAV1, MAV2, MAV3, RMS, VAV, iEMG, SSI, WL;
[0040] Then, the feature combination MAV1, MAV2, MAV3 and RMS are input into the first classifier to classify and identify the training gesture group to which the current signal belongs;
[0041] The classification results and feature combinations MAV1, MAV2, MAV3, RMS, VAV, iEMG, SSI, and WL of the first classifier are then input into the second classifier. Finally, the second classifier is used to identify the training gesture to which the current signal belongs, thereby determining the user's intention to move the finger.
[0042] Preferably, the bending sensor is a flexible film sensor, which includes a plurality of rectangular sheet portions and an arc-shaped bending line portion connected between the plurality of rectangular sheet portions.
[0043] Preferably, the rehabilitation training module controls the power module to apply force to the single finger sleeve of the glove body so that the single finger sleeve drives the user's fingers to perform corresponding actions, thereby achieving rehabilitation training of the fingers;
[0044] The training methods of the rehabilitation training module include finger training mode, power training mode, grip training mode and stimulation training mode;
[0045] When the finger-to-finger training mode is adopted, the glove body applies force to make any finger of the user except the thumb perform repeated passive finger-to-finger exercises with the thumb;
[0046] When the power-assisted training mode is adopted, the intention sensing module is used to judge the user's finger movement intention, and then the glove body is used to apply auxiliary driving force to the finger movement to assist the patient in performing the finger movement action;
[0047] When the grip training mode is adopted, the glove body applies force to make the user's fingers perform continuous passive gripping movements;
[0048] When the stimulation training mode is adopted, the single finger sleeve of the glove body applies force to the user's finger for passive movement, while the electrical stimulation module electrically stimulates the flexor and extensor muscles on the user's forearm corresponding to the finger.
[0049] Preferably, the evaluation and analysis module collects the strength, curvature, and electromyographic signals of the user during training in real time, compares them with the training effect information of normal people's hands, scores the current rehabilitation training effect, and provides feedback on the user's training status.
[0050] The beneficial effects of the present invention are as follows: the hand function rehabilitation device with intention perception function provided by the present invention can carry out effective rehabilitation training for patients with hand dysfunction;
[0051] The present invention adopts a flexible lasso for power transmission, which can simplify the device structure, reduce the size of the equipment, improve the comfort of use, and overcome the problems of complex structure and poor comfort of existing products.
[0052] The present invention has an intention sensing function, which automatically provides auxiliary driving force for patients through intention sensing judgment, enabling patients to actively train and effectively improve the rehabilitation training effect;
[0053] The present invention has multiple rehabilitation training modes, which can solve the problems of single training mode and unsatisfactory rehabilitation effect in existing products;
[0054] The present invention also has a rehabilitation training result evaluation and analysis function, and can provide real-time feedback to evaluate the rehabilitation training effect of the patient. BRIEF DESCRIPTION OF THE DRAWINGS
[0055] Figure 1 It is a principle block diagram of the hand function rehabilitation device with intention sensing function of the present invention;
[0056] Figure 2 It is a schematic diagram of the external structure of the main machine and the transmission mechanism of the present invention;
[0057] Figure 3 It is a schematic diagram of the internal structure of the main engine and the transmission mechanism of the present invention;
[0058] Figure 4 An exploded view of the transmission mechanism of the present invention;
[0059] Figure 5 It is a structural schematic diagram of the transmission unit of the present invention;
[0060] Figure 6 It is a structural schematic diagram of the upper transmission member of the present invention;
[0061] Figure 7 It is a structural schematic diagram of the lower transmission member of the present invention;
[0062] Figure 8 It is a schematic diagram of the structure of the transmission unit and the lower cover of the present invention;
[0063] Fig. 9It is a schematic cross-sectional structural diagram of the transmission housing of the present invention;
[0064] Fig.10 It is a schematic cross-sectional structural diagram of the steel wire lasso of the present invention;
[0065] Fig.11 It is a schematic diagram of the external structure of the power output mechanism of the present invention;
[0066] Fig.12 It is a schematic diagram of the internal structure of the power output mechanism of the present invention;
[0067] Fig.13 An exploded view of the power output mechanism of the present invention;
[0068] Fig.14 is a schematic structural diagram of a bending sensor in Embodiment 2 of the present invention;
[0069] Fig.15 The tensile test results in Example 2 of the present invention;
[0070] Fig.16 This is a schematic diagram of a voltage value test gesture of a bending sensor at different bending degrees in Example 2 of the present invention;
[0071] Fig.17 The voltage value test result of the bending sensor under bending in Example 2 of the present invention;
[0072] Fig.18 is a repeated bending test result of the bending sensor in Example 2 of the present invention;
[0073] Fig.19 It is a schematic diagram of some training gestures in embodiment 2 of the present invention.
[0074] Description of reference numerals:
[0075] 1—host; 10—host 1 housing; 11—touch screen;
[0076] 2—Transmission mechanism;
[0077] 20—transmission housing; 200—upper cover; 201—lower cover; 202—mounting block; 203—first T-shaped slideway; 204—second T-shaped slideway; 205—slideway;
[0078] 21—wire lasso unit; 210—first wire lasso; 211—second wire lasso; 212—spring tube; 213—polytetrafluoroethylene sleeve; 214—steel wire;
[0079] 22—transmission unit; 220—upper transmission member; 221—lower transmission member; 222—first slider; 223—second slider; 224—slot; 2220—T-shaped slider; 2221—insert block; 2222—first connecting hole; 2230—base; 2231—stop bar;
[0080] 23—core wire unit; 230—first core wire; 231—second core wire;
[0081] 3—power module; 30—power output mechanism; 31—power housing; 32—motor; 33—screw; 34—screw nut; 35—pulley; 36—traction rope; 361—first traction rope segment; 362—second traction rope segment; 363—first source output end; 364—second source output end; 37—limiting groove; 38—limiting block;
[0082] 4—bending sensor; 40—rectangular sheet portion; 41—arc-shaped bending line portion. DETAILED DESCRIPTION
[0083] The present invention is further described in detail below in conjunction with embodiments so that those skilled in the art can implement the invention with reference to the description.
[0084] It should be understood that the terms such as “having”, “including” and “comprising” used herein do not exclude the existence or addition of one or more other elements or combinations thereof.
[0085] Example 1
[0086] Reference Figure 1 , a hand function rehabilitation device with intention perception function of this embodiment includes:
[0087] The host 1 includes a power module 3, an intention perception module, a rehabilitation training module, an evaluation and analysis module, and a control module;
[0088] The glove body has five independent single-finger sleeves that cooperate with the fingers;
[0089] and a transmission mechanism 2 connected between the host 1 and the glove body, the transmission mechanism 2 transmits the driving force output by the power module 3 to the glove body to drive the single finger sleeve of the glove body to perform bending and stretching movements;
[0090] Among them, the intention perception module determines the user's finger movement intention by analyzing the collected pressure signals, bending angle signals and forearm electromyographic signals of the corresponding hand of the user's fingers, thereby providing auxiliary driving force for the finger movement through the power module 3.
[0091] The control module is used to realize centralized control. The control module can receive instructions from the intention perception module and the rehabilitation training module, and perform corresponding control on the power module 3. The control module is connected to the touch screen 11, and instructions are input through the touch screen 11, and then the whole machine can be controlled through the control module.
[0092] Reference Figure 2-13 In this embodiment, the power module 3 includes five power output mechanisms 30, which include a power housing 31, a motor 32 disposed in the power housing 31, a screw 33 drivingly connected to the motor 32, a screw nut 34 cooperating with the screw 33, a pulley 35 rotatably disposed inside the power housing 31, and a traction rope 36 wound around the pulley 35;
[0093] The traction rope 36 is wound around the pulley 35 to form two sections: a first traction rope section 361 and a second traction rope section 362. The movable ends of the first traction rope section 361 and the second traction rope section 362 form two source output ends: a first source output end 363 and a second source output end 364.
[0094] The first traction rope segment 361 is fixedly connected to the screw nut 34, and the second traction rope segment 362 can slide through the screw nut 34, so that the reciprocating linear motion of the screw nut 34 on the screw 33 enables the first source output end 363 and the second source output end 364 to respectively generate a pulling force output toward the pulley 35.
[0095] Among them, the upper and lower surfaces of the power housing 31 are provided with limit slots 37, and the upper and lower ends of the screw nut 34 are provided with limit blocks 38. The limit blocks 38 are arranged in the limit slots 37, can slide back and forth, and can limit the rotation of the screw nut 34. The end of the screw 33 is rotatably connected to the power housing 31, and the motor 32 drives the screw 33 to rotate, so that the screw nut 34 moves back and forth, so that the first traction rope segment 361 and the second traction rope segment 362 can be driven by the screw nut 34 to move in the direction toward the pulley 35, so that the first source output end 363 and the second source output end 364 can generate two-way traction force toward the pulley 35, and the two-way traction force is finally used to apply force to the upper and lower directions of the finger, so as to pull the finger to bend or straighten.
[0096] In a preferred embodiment, the power output mechanism 30 can provide a traction force of no less than 200N, which can meet the finger bending training action.
[0097] In this embodiment, the transmission mechanism 2 includes a transmission housing 20, five transmission units 22 arranged in the transmission housing 20, five steel wire 214 lasso units 21 matched one by one with the input ends of the five transmission units 22, and five core wire units 23 matched one by one with the output ends of the five transmission units 22.
[0098] The transmission mechanism 2 is used to transmit the traction force of the first traction rope segment 361 and the second traction rope segment 362 to the single finger sleeve, so as to pull the single finger sleeve to bend or straighten.
[0099] A set of steel wire 214 lasso unit 21 is used to transmit the driving force output by a power module 3 to a set of transmission unit 22, and then finally transmit the output to a single finger sleeve on the glove body through a set of core wire unit 23;
[0100] A transmission unit 22 includes an upper transmission member 220 and a lower transmission member 221, a steel wire 214 lasso unit 21 includes two steel wires 214 lassos respectively driven and connected to the input ends of an upper transmission member 220 and a lower transmission member 221: a first steel wire 214 lasso 210 and a second steel wire 214 lasso 211, and a core wire unit 23 includes two core wires respectively driven and connected to the output ends of an upper transmission member 220 and a lower transmission member 221: a first core wire 230 and a second core wire 231.
[0101] That is, one power output mechanism 30, one set of steel wire 214 lasso units 21, one set of transmission unit 22, one set of cored wire unit 23, and one single finger sleeve correspond in sequence, and each set of steel wire 214 lasso units 21, one set of transmission unit 22, one set of cored wire unit 23, and one single finger sleeve correspond in sequence all have two-way forces for applying traction to the upper and lower parts of the single finger sleeve, thereby pulling the single finger sleeve to bend or straighten.
[0102] In this embodiment, refer to Fig.10 The steel wire 214 lasso includes a spring tube 212, a polytetrafluoroethylene sleeve 213 inserted in the spring tube 212, and a steel wire 214 slidably inserted in the polytetrafluoroethylene sleeve 213. The outermost spring tube 212 plays the role of an external bendable support, and the middle layer of the polytetrafluoroethylene sleeve 213 can reduce resistance, so that the steel wire 214 can move more smoothly in the polytetrafluoroethylene sleeve 213 to provide traction.
[0103] The core wire includes a sheath and a traction wire slidably inserted in the sheath; the sheath plays a supporting role, the traction wire moves in the sheath to provide traction, and the output end of the traction wire is fixed to the upper or lower part of each single finger sleeve after passing through a wire channel fixed on the palm of the glove body.
[0104] The input end of the steel wire 214 of the first steel wire 214 lasso 210 is connected to the first source output end 363, the output end of the steel wire 214 of the first steel wire 214 lasso 210 is connected to the input end of the upper transmission member 220, the output end of the upper transmission member 220 is connected to the input end of the traction wire of the first core wire 230, and the output end of the traction wire of the first core wire 230 is connected to the back of a finger (on the same plane as the back of the hand) of a single finger sleeve;
[0105] The input end of the steel wire 214 of the second steel wire 214 lasso 211 is connected to the second source output end 364, the output end of the steel wire 214 of the second steel wire 214 lasso 211 is connected to the input end of the lower transmission member 221, the output end of the lower transmission member 221 is connected to the input end of the traction wire of the second cored wire 231, and the output end of the traction wire of the second cored wire 231 is connected to the fingertip of a single finger sleeve (in the same plane as the palm).
[0106] Reference Figure 3-5 , taking the drive of a single finger sleeve as an example, the power transmission process in the device is:
[0107] The motor 32 in the power output mechanism 30 works to move the screw nut 34 to the left, pulling the first traction rope segment 361 to move to the left, so that the first source output end 363 generates a first traction force, and the first traction force is transmitted to the upper transmission member 220 of the group transmission unit 22 through the first steel wire 214 lasso 210, and then transmitted to the first core wire 230, and the first core wire 230 generates a pulling force on the back of the finger of the single finger sleeve, so that the finger is straightened;
[0108] The motor 32 in the power output mechanism 30 is reversed to move the screw nut 34 to the right, and the second traction rope segment 362 is pulled to the left through the action of the pulley 35, so that the second source output end 364 generates a second traction force, and the second traction force is transmitted to the lower transmission member 221 of the group transmission unit 22 through the second steel wire 214 lasso 211, and then transmitted to the second core wire 231, and a pulling force is generated on the fingertips of the single finger sleeve through the second core wire 231, so that the fingers are bent.
[0109] In this embodiment, the transmission housing 20 includes an upper cover 200, a lower cover 201, and a mounting block 202 disposed between the upper cover 200 and the lower cover 201. The upper cover 200 and the lower cover 201 are buckled and connected to each other and can be fixed by screws. At least five first T-shaped slide grooves 203 are respectively provided on the inner walls of the upper cover 200 and the lower cover 201, and at least five second T-shaped slide grooves 204 are respectively provided on the upper and lower surfaces of the mounting block 202; the upper and lower adjacent first T-shaped slide grooves 203 and second T-shaped slide grooves 204 form a slideway 205;
[0110] The upper transmission member 220 and the lower transmission member 221 have the same structure and are symmetrically arranged in the transmission housing 20. The upper transmission member 220 and the lower transmission member 221 each include a first slider 222 and a second slider 223.
[0111] The first slider 222 includes a T-shaped slider 2220 slidably disposed in the first T-shaped slide groove 203, an insert block 2221 connected to the front end of the T-shaped slider 2220, and a first connection hole 2222 penetrating the T-shaped slider 2220 and the insert block 2221 for connecting with the input end of the traction wire;
[0112] The second slider 223 includes a T-shaped base 2230, two blocking bars 2231 connected to the base 2230 in a direction parallel to the insert block 2221, and a second connecting hole (not shown) provided on the base 2230 for connecting to the output end of the steel wire 214;
[0113] The base 2230 can be slidably arranged in the second T-shaped slide groove 204, and a slot 224 is formed between the two baffles 2231. After the insert block 2221 is inserted into the slot 224, the end of the baffle 2231 contacts the side of the T-shaped slider 2220, so that the first slider 222 and the second slider 223 can slide back and forth together in the slideway 205 to realize the transmission of traction force between the steel wire 214 lasso and the core wire.
[0114] The output end of the steel wire 214 of the first steel wire 214 lasso 210 is connected to the second connection hole of the second slider 223 of the upper transmission member 220, and the input end of the traction wire of the first cored wire 230 is connected to the first connection hole 2222 of the first slider 222 of the upper transmission member 220;
[0115] The output end of the steel wire 214 of the second steel wire 214 lasso 211 is connected to the second connecting hole of the second slider 223 of the lower transmission member 221 , and the input end of the traction wire of the second cored wire 231 is connected to the first connecting hole 2222 of the first slider 222 of the lower transmission member 221 .
[0116] Reference Figure 5 and 6For the upper transmission member 220, the first slider 222 is above the second slider 223, and the insert block 2221 is inserted into the slot 224 from left to right. The steel wire 214 of the first steel wire 214 lasso 210 is connected to the left end of the second slider 223, and the traction wire of the first cored wire 230 is connected to the right end of the insert block 2221. The steel wire 214 generates a leftward pulling force on the second slider 223, and the traction wire generates a rightward pulling force on the first slider 222, so that the insert block 2221 is inserted into the slot 224 to the bottom, and the right end face of the T-shaped slider 2220 contacts the left end face of the baffle 2231. The first slider 222 and the second slider 223 are clamped and can slide back and forth together in the slideway 205 to realize the transmission of force between the steel wire 214 and the traction wire.
[0117] Example 2
[0118] In this embodiment, the host 1 also includes a host shell 10, a touch display screen 11 arranged on the host shell 10, and an electrical stimulation module connected to the rehabilitation training module. The power module 3, the intention perception module, the rehabilitation training module, the evaluation and analysis module and the control module are all arranged in the host shell 10.
[0119] The touch display screen 11 is used for human-computer interaction and can display training results, parameters and other data, and can also be used as an input port of the device.
[0120] The intention perception module includes a signal acquisition unit and an intention judgment model. The signal acquisition unit includes a pressure sensor arranged on a single finger sleeve for collecting fingertip pressure signals, a bending sensor 4 arranged on the single finger sleeve for collecting finger bending angle signals, and a surface electromyography armband worn on the user's arm for collecting forearm electromyography signals;
[0121] The intention judgment model is a classifier based on the random forest model, which receives signals collected by the pressure sensor, the bending sensor 4 and the surface electromyography armband, and judges the finger movement intention through analysis.
[0122] In one embodiment, the surface electromyography armband adopts the Myo electromyography armband.
[0123] In one embodiment, the bending sensor 4 is a FLEX2.2 bending sensor 4. When the metal surface of the sensor is bent outward, the resistance value of the sensor will change, so that the bending degree can be detected.
[0124] In one embodiment, the pressure sensor is a FSR403 thin film pressure sensor, which converts the pressure applied to the thin film area of the FSR403 pressure sensor into a change in resistance value, thereby obtaining pressure information.
[0125] Reference Fig.14In one embodiment, the bending sensor 4 is a flexible film sensor, which includes a plurality of rectangular sheet portions 40 and an arc-shaped bending line portion 41 connected between the plurality of rectangular sheet portions 40 .
[0126] The five bending sensors 4 (Sensor A-Sensor E) shown in the figure were subjected to a tensile test, and the results are as follows: Fig.15 As shown, it can be seen that the resistance value and the stretching rate of the five bending sensors 4 have good linearity when the stretching rate is 0-8%.
[0127] use Fig.15 The bending sensor 4 shown in A is used to test the voltage value of the index finger at different bending degrees, such as Fig.16 The test gestures include 0°, 45°, 90°, and 135°. The test results are as follows Fig.17 As shown in Figure 2, it can be seen that the voltage value has an obvious gradient at different bending angles. Then three tests were performed, and the results are as follows: Fig.18 As shown, it can be seen that it has good repeatability.
[0128] It can be seen from the above experimental results that the bending sensor 4 used in this embodiment has excellent performance and can be well used for collecting bending angle signals of fingers.
[0129] In one embodiment, the intention judgment model includes a preprocessing module, a feature extraction module, a first classifier and a second classifier, and the working method of the intention judgment model includes the following steps:
[0130] 1) Train the intention judgment model;
[0131] 1-1) Four groups of training gestures were designed in advance: the first group included fist (FH), open hand (HO) and finger curling (FF), which were used to show different movements of all fingers together; the second group included index finger, middle finger, ring finger and pinky finger curled together (IMRIF), middle finger, ring finger and pinky finger curled together (MRLF), thumb, ring finger and pinky finger curled together (TRL), thumb, middle finger, ring finger and pinky finger curled and index finger pointing to something (TMRI), and thumb, index finger and middle finger curled simulated grasping (TIM), which were used to show the hand flexibility of performing different actions; the third group included thumb curling (TF) and thumb and index finger curling (TIF), which were used to show the movement of the thumb and the coordination of the thumb and index finger; the fourth group included the individual curling of each finger, which was used to show the movement of a single finger; refer to Fig.19 , indicating various training gestures from the first to the third groups;
[0132] 1-2) Collecting training data: The subject wears the glove body and performs gesture actions according to the group training gestures in step 1-1), and collects signals through the signal acquisition unit at the same time, specifically: collecting fingertip pressure signals through the pressure sensor, collecting finger bending angle signals through the bending sensor 4, and collecting forearm electromyography signals through the surface electromyography arm ring; each subject collects multiple groups of training data, and the training data of multiple subjects are collected, so as to obtain N groups of training data in total;
[0133] 1-3) The collected training data is first preprocessed using a Butterworth bandpass filter, and then feature extraction is performed using the segmentation overlap method. The features extracted from the pressure signal include the absolute average value, recorded as MAV1; the features extracted from the bending angle signal include the absolute average value, recorded as MAV2; the features extracted from the forearm electromyographic signal include: absolute average value MAV3, root mean square RMS, electromyographic variance VAV, integrated electromyographic value iEMG, simple square integral SSI, and wavelength WL;
[0134] 1-4) Taking the feature combination MAV1, MAV2, MAV3 and RMS extracted from the i-th group of training data as the input of the first classifier, taking the training gesture group corresponding to the i-th group of training data as the output of the first classifier, and training the first classifier; then taking the feature combination MAV1, MAV2, MAV3, RMS, VAV, iEMG, SSI, WL extracted from the i-th group of training data and the output result of the first classifier as the input of the second classifier, taking the specific training gesture corresponding to the i-th group of training data as the output of the second classifier, and training the second classifier;
[0135] 1-5) According to the method of step 1-4), the first classifier and the second classifier are trained by using N sets of training data, and finally a trained intention judgment model is obtained;
[0136] 2) The user wears the glove body for rehabilitation training, and the preprocessing module receives the signals collected by the pressure sensor, the bending sensor 4 and the surface electromyography armband, and first preprocesses them using a Butterworth bandpass filter before transmitting them to the feature extraction module;
[0137] The feature extraction module uses the segmentation overlap method to extract the following features: MAV1, MAV2, MAV3, RMS, VAV, iEMG, SSI, WL;
[0138] Then, the feature combination MAV1, MAV2, MAV3 and RMS are input into the first classifier to classify and identify the training gesture group to which the current signal belongs;
[0139] Then the classification result of the first classifier and the feature combination MAV1, MAV2, MAV3, RMS, VAV, iEMG, SSI, WL are input into the second classifier, and finally the training gesture to which the current signal belongs is identified by the second classifier, so as to judge the user's intention of finger movement. Among them, the feature combination MAV1, MAV2, MAV3 and RMS can realize the rapid classification of the training gesture group, and then the more refined feature combination MAV1, MAV2, MAV3, RMS, VAV, iEMG, SSI, WL can realize the accurate classification of the specific training gesture; using different feature combinations to first classify the group and then classify the specific gesture can improve the classification recognition speed and accuracy.
[0140] In one embodiment, the rehabilitation training module controls the power module 3 to apply force to the single finger sleeve of the glove body so that the single finger sleeve drives the user's finger to perform corresponding actions, thereby achieving rehabilitation training for the finger;
[0141] The training methods of the rehabilitation training module include finger training mode, power training mode, grip training mode and stimulation training mode;
[0142] When the finger-to-finger training mode is adopted, the glove body applies force to make any finger of the user except the thumb perform repeated passive finger-to-finger exercises with the thumb;
[0143] When using the power-assisted training mode, the intention sensing module determines the user's finger movement intention, and then the glove body applies auxiliary driving force to the finger movement to assist the patient in performing the finger movement.
[0144] When the grip training mode is adopted, the glove body applies force to make the user's fingers perform continuous passive gripping movements;
[0145] When the stimulation training mode is adopted, the single finger sleeve of the glove body applies force to the user's finger for passive movement, while the electrical stimulation module electrically stimulates the flexor and extensor muscles on the user's forearm corresponding to the finger.
[0146] In one embodiment, the evaluation and analysis module collects the strength, curvature, and electromyographic signals of the user during training in real time, compares them with the training effect information of a normal person's hand, scores the current rehabilitation training effect, and provides feedback on the user's training status.
[0147] Although the embodiments of the present invention have been disclosed as above, they are not limited to the applications listed in the specification and the implementation modes. They can be fully applied to various fields suitable for the present invention. For those familiar with the art, additional modifications can be easily implemented. Therefore, without departing from the general concept defined by the claims and the scope of equivalents, the present invention is not limited to specific details.
Claims
1. A hand function rehabilitation device with intention perception function, characterized in that: include: The host includes a power module, an intention perception module, a rehabilitation training module, an evaluation and analysis module, and a control module; The glove body has five independent single-finger sleeves that fit the fingers; and a transmission mechanism connected between the host and the glove body, the transmission mechanism transmitting the driving force output by the power module to the glove body to drive the single finger sleeve of the glove body to perform bending and stretching movements; The intention sensing module analyzes the collected pressure signals, bending angle signals and forearm electromyographic signals of the user's fingers and the corresponding hand to determine the user's finger movement intention, thereby providing auxiliary driving force for the finger movement through the power module; The host further includes a host housing, a touch screen display provided on the host housing, and an electrical stimulation module connected to the rehabilitation training module. The power module, intention perception module, rehabilitation training module, evaluation and analysis module, and control module are all provided in the host housing. The intention perception module includes a signal acquisition unit and an intention judgment model. The signal acquisition unit includes a pressure sensor provided on a single finger sleeve for collecting fingertip pressure signals, a bending sensor provided on the single finger sleeve for collecting finger bending angle signals, and a surface electromyography armband worn on the user's arm for collecting forearm electromyography signals. The intention judgment model receives signals collected by the pressure sensor, the bending sensor and the surface electromyography armband, and judges the finger movement intention through analysis; The intention judgment model includes a preprocessing module, a feature extraction module, a first classifier and a second classifier. The working method of the intention judgment model includes the following steps: 1) Train the intention judgment model; 1-1) Pre-design four groups of training gestures: the first group includes fist, open hand, and finger curling, which are used to show different movements of all fingers together; the second group includes curling the index finger, middle finger, ring finger, and pinky together, curling the middle finger, ring finger, and pinky together, curling the thumb, ring finger, and pinky together, curling the thumb, middle finger, ring finger, and pinky with the index finger pointing at something, and curling the thumb, index finger, and middle finger to simulate grasping, which are used to show the flexibility of the hand to perform different actions; the third group includes curling the thumb, and curling the thumb and index finger, which are used to show the movement of the thumb and the coordination of the thumb and index finger; the fourth group includes curling each finger separately, which is used to show the movement of a single finger; 1-2) Collecting Training Data: The subjects wear the glove and perform the four training gestures in step 1-1). Simultaneously, the signal acquisition unit collects signals. Specifically, the pressure sensor collects fingertip pressure signals, the bend sensor collects finger bend angle signals, and the surface electromyography armband collects forearm electromyography signals. Multiple sets of training data are collected for each subject, and training data from multiple subjects is collected, resulting in a total of N sets of training data. 1-3) The collected training data was first preprocessed using a Butterworth bandpass filter, and then feature extraction was performed using the segmentation overlap method. The features extracted from the pressure signal included the absolute mean value, recorded as MAV1; the features extracted from the bending angle signal included the absolute mean value, recorded as MAV2; the features extracted from the forearm electromyography signal included the absolute mean value MAV3, root mean square RMS, electromyography variance VAV, integrated electromyography value iEMG, simple square integral SSI, and wavelength WL; 1-4) Using the feature combination MAV1, MAV2, MAV3, and RMS extracted from the i-th group of training data as the input of the first classifier, and using the training gesture group corresponding to the i-th group of training data as the output of the first classifier, the first classifier is trained; then using the feature combination MAV1, MAV2, MAV3, RMS, VAV, iEMG, SSI, WL extracted from the i-th group of training data and the output of the first classifier as the input of the second classifier, and using the specific training gesture corresponding to the i-th group of training data as the output of the second classifier, the second classifier is trained; 1-5) Following the method in steps 1-4), train the first classifier and the second classifier using N sets of training data to ultimately obtain a trained intent judgment model; 2) The user wears the glove body for rehabilitation training, and the preprocessing module receives the signals collected by the pressure sensor, the bending sensor, and the surface electromyography armband, preprocesses them using a Butterworth bandpass filter, and then transmits them to the feature extraction module; The feature extraction module uses the segmentation overlap method to extract the following features: MAV1, MAV2, MAV3, RMS, VAV, iEMG, SSI, WL; Then, the feature combination MAV1, MAV2, MAV3 and RMS is input into the first classifier to classify and identify the training gesture group to which the current signal belongs; The classification results of the first classifier and the feature combination MAV1, MAV2, MAV3, RMS, VAV, iEMG, SSI, and WL are then input into the second classifier. Finally, the second classifier recognizes the training gesture to which the current signal belongs, thereby determining the user's intention to perform finger movement. The bending sensor is a flexible film sensor, which includes a plurality of rectangular pieces and an arc-shaped bending line portion connecting the plurality of rectangular pieces.
2. The hand function rehabilitation device with intention perception function according to claim 1, characterized in that: The power module includes five power output mechanisms, each of which includes a power housing, a motor disposed in the power housing, a screw connected to the motor, a screw nut fitted on the screw, a pulley rotatably disposed inside the power housing, and a traction rope wound around the pulley; After being wound around the pulley, the traction rope forms two sections: a first traction rope section and a second traction rope section. The movable ends of the first traction rope section and the second traction rope section form two source output ends: a first source output end and a second source output end. The first traction rope segment is fixedly connected to the lead screw nut, and the second traction rope segment can slide through the lead screw nut, so that the reciprocating linear motion of the lead screw nut on the lead screw causes the first source output end and the second source output end to respectively generate a pulling force output toward the pulley.
3. The hand function rehabilitation device with intention perception function according to claim 2, characterized in that: The transmission mechanism includes a transmission housing, five transmission units disposed in the transmission housing, five steel wire lasso units matched one-to-one with the input ends of the five transmission units, and five cored wire units matched one-to-one with the output ends of the five transmission units. A set of steel wire lasso units is used to transmit the driving force output by a power module to a set of transmission units, and then finally transmits the output to a single finger sleeve on the glove body through a set of cored wire units; A transmission unit group includes an upper transmission member and a lower transmission member, a wire lasso unit includes two wire lassoes respectively driven and connected to the input ends of the upper transmission member and the lower transmission member: a first wire lasso and a second wire lasso, and a cored wire unit includes two cored wires respectively driven and connected to the output ends of the upper transmission member and the lower transmission member: a first cored wire and a second cored wire.
4. The hand function rehabilitation device with intention perception function according to claim 3, characterized in that: The steel wire lasso comprises a spring tube, a polytetrafluoroethylene sleeve inserted in the spring tube, and a steel wire slidably inserted in the polytetrafluoroethylene sleeve; the core wire comprises a sheath and a traction wire slidably inserted in the sheath; The input end of the steel wire of the first steel wire lasso is connected to the output end of the first source, the output end of the steel wire of the first steel wire lasso is connected to the input end of the upper transmission member, the output end of the upper transmission member is connected to the input end of the traction wire of the first cored wire, and the output end of the traction wire of the first cored wire is connected to the back of a finger of a single finger sleeve; The input end of the steel wire of the second steel wire lasso is connected to the output end of the second source, the output end of the steel wire of the second steel wire lasso is connected to the input end of the lower transmission member, the output end of the lower transmission member is connected to the input end of the traction wire of the second cored wire, and the output end of the traction wire of the second cored wire is connected to the fingertip of a single finger sleeve.
5. The hand function rehabilitation device with intention perception function according to claim 4, characterized in that: The transmission housing includes an upper cover, a lower cover, and a mounting block disposed between the upper cover and the lower cover. The inner walls of the upper cover and the lower cover are respectively provided with at least five first T-shaped slide grooves. The upper and lower surfaces of the mounting block are respectively provided with at least five second T-shaped slide grooves. The upper and lower adjacent first T-shaped slide grooves and second T-shaped slide grooves form a slideway. The upper transmission member and the lower transmission member have the same structure and are symmetrically arranged in the transmission housing. The upper transmission member and the lower transmission member each include a first slider and a second slider. The first slider includes a T-shaped slider slidably disposed in the first T-shaped slot, an insert block connected to the front end of the T-shaped slider, and a first connecting hole penetrating the T-shaped slider and the insert block for connecting to the input end of the traction wire; The second slider includes a T-shaped base, two blocking bars connected to the base in a direction parallel to the insert block, and a second connecting hole provided on the base for connecting to the output end of the steel wire; The base can be slidably arranged in the second T-shaped slide groove, and a slot is formed between the two baffles. After the insert block is inserted into the slot, the end of the baffle contacts the side of the T-shaped slider, so that the first slider and the second slider can slide back and forth together in the slide to realize the transmission of traction force between the wire lasso and the cored wire.
6. The hand function rehabilitation device with intention perception function according to claim 5, characterized in that: The output end of the steel wire of the first steel wire lasso is connected to the second connecting hole of the second slider of the upper transmission member, and the input end of the traction wire of the first cored wire is connected to the first connecting hole of the first slider of the upper transmission member; The output end of the steel wire of the second steel wire lasso is connected to the second connecting hole of the second slider of the lower transmission member, and the input end of the traction wire of the second cored wire is connected to the first connecting hole of the first slider of the lower transmission member.
7. The hand function rehabilitation device with intention perception function according to claim 6, characterized in that: The rehabilitation training module controls the power module to apply force to the single finger sleeve of the glove body so that the single finger sleeve drives the user's fingers to perform corresponding movements, thereby achieving rehabilitation training for the fingers; The training methods of the rehabilitation training module include finger training mode, power training mode, grip training mode and stimulation training mode; In the finger-to-finger training mode, the glove body applies force to make the user's thumb and any finger other than the thumb perform repeated passive finger-to-finger exercises. When using the power-assisted training mode, the intention sensing module determines the user's finger movement intention, and then the glove body applies auxiliary driving force to the finger movement to assist the patient in performing the finger movement. In the grip training mode, the glove body applies force to make the user's fingers perform continuous passive gripping movements; When the stimulation training mode is adopted, the single finger sleeve of the glove body applies force to the user's finger for passive movement, while the electrical stimulation module electrically stimulates the flexor and extensor muscles on the user's forearm corresponding to the finger.
8. The hand function rehabilitation device with intention perception function according to claim 7, characterized in that: The evaluation and analysis module collects the user's pressure signals, bending angle signals and forearm electromyographic signals in real time during the training process, compares them with the training effect information of normal people's hands, scores the current rehabilitation training effect, and provides feedback on the user's training status.
Citation Information
Patent Citations
Multi-mode fusion hand function rehabilitation training and intelligent evaluation system
CN105963926A
Multi-dimensional surface myoelectricity signal artificial hand control method based on principal component analysis
CN107378944A
Intelligent exoskeleton rehabilitation mechanical arm based on normal side biological electrical control and method
CN108743223A
Finger exercise rehabilitation training robot based on lasso drive and electromyographic control
CN109907939A
Electromyographic signal gesture recognition method combined with cross-modal incidence relation model
CN110443309A
