Intelligent ankle foot orthosis with sole touch reconstruction function

Abstract

The invention provides an intelligent ankle-foot orthosis with a sole touch reconstruction function. The intelligent ankle-foot orthosis comprises a main control unit, a six-axis inertial measurement unit, two vibration units and a dual-channel electrical stimulation unit, the two vibration units are fixed at the anterior tibia muscle of the body surface and the achilles tendon of the body surface and alternately or simultaneously vibrate when corresponding to each other so as to build a space tactile reconstruction passage; a six-axis inertial measurement unit is adopted to collect inertial data during walking, and a sliding window is used to extract feature vectors; real-time classification of gait phases is carried out on the feature vectors through a random forest model, and real-time correction is carried out on classification results through an objective sequential logic module; in a corresponding gait time phase, electric pulses of the double-channel electrical stimulation unit are used for arousing pressure sensation feelings of different subareas of the sole to complete a real-time sole tactile reconstruction process, and a space coding tactile reconstruction process is completed through mechanical vibration at the corresponding body surface tibia anterior muscle and the body surface achilles tendon. According to the invention, touch feelings of different areas of the sole can be activated in real time to realize ankle-foot touch reconstruction when the user walks.

Description

Technical Field

[0001] The invention relates to the technical field of foot and ankle orthoses, and in particular to an intelligent ankle and foot orthosis capable of reconstructing plantar tactile sensation. Background Art

[0002] The foot is an important sensory structure in the postural control mechanism, as the sole of the foot is the body's direct and often only contact surface with the ground, allowing us to sense and interact with the environment. However, the sensitivity of the plantar skin naturally decreases with age and the effects of neurological diseases including diabetes and chemotherapy. This reduction in sensitivity often leads to a weakened postural control ability, increasing the risk of falls, especially in the elderly and patients with certain medical conditions.

[0003] As a medical device used to support the ankle joint and foot, the main functions of the ankle-foot orthosis include fixing the normal functional position of the joint, pulling the tendons for a long time, and compensating for some muscle weakness. Dynamic ankle-foot orthoses can further help correct and improve the patient's walking function. However, traditional ankle-foot orthoses usually lack tactile reconstruction functions, which makes it impossible for users to accurately perceive the tactile sensation of the foot space, thereby increasing the risk of falling and injury. Summary of the invention

[0004] In view of the shortcomings of the prior art, the present invention provides an intelligent ankle-foot orthosis with plantar tactile reconstruction. The present invention uses electric pulse and vibration technology to activate the tactile sensation of different areas of the sole of the foot in real time when the user walks, thereby achieving plantar tactile reconstruction and significantly improving the user's walking experience.

[0005] The technical solution of the present invention is: an intelligent ankle-foot orthosis with plantar tactile reconstruction, comprising a main control unit, and a six-axis inertial measurement unit, two vibration units, and a dual-channel electrical stimulation unit connected to the main control unit;

[0006] Wherein, the electrical pulse channel 1 of the dual-channel electrical stimulation unit stimulates the tibial nerve and the sural nerve of the ankle and can cause tactile sensation in the heel; the electrical pulse channel 2 only stimulates the tibial nerve and can cause tactile sensation in the front of the sole;

[0007] The two vibration units are fixed on the tibialis anterior muscle and the Achilles tendon on the body surface, and the two vibration units vibrate alternately or simultaneously in corresponding phases to build a spatial tactile reconstruction pathway;

[0008] The six-axis inertial measurement unit collects inertial data during walking, and uses a sliding window to extract feature vectors from the collected inertial data;

[0009] The main control unit is integrated with a random forest model and an objective sequential logic module; the feature vector is classified in real time according to the gait phase by the pre-trained random forest model, and the classification result is corrected in real time by the objective sequential logic module;

[0010] In the corresponding gait phase, the electrical pulses of the dual-channel electrical stimulation unit are used to evoke the pressure sensation of different partitions of the sole of the foot to complete the real-time plantar tactile reconstruction process. At the same time, the mechanical vibration of the vibration unit corresponding to the tibialis anterior muscle and the Achilles tendon on the body surface is used to complete the spatial encoding tactile reconstruction process.

[0011] Preferably, the ankle-foot orthosis system also includes a Bluetooth module and a 3.7V rechargeable lithium battery. The 3.7V rechargeable lithium battery increases the voltage from 3.7V to 5V through a voltage conversion circuit to power the main control unit, and reduces the voltage from 3.7V to 3.3V to power the six-axis inertial measurement unit.

[0012] Preferably, the dual-channel electrical stimulation unit utilizes a dual-channel biphasic rectangular wave generating circuit to generate rectangular wave electrical pulses with different pulse widths and frequencies. The dual-channel biphasic rectangular wave generating circuit includes a boost circuit and two H-bridge circuits using a DC motor driver chip. The boost circuit boosts the 5V voltage to 12V to 35V; and generates a dual-channel biphasic pulse waveform through two H-bridge circuits using a DC motor driver chip and outputs it through a hydrogel electrode pair.

[0013] Preferably, the main control unit adopts a single-chip microcomputer with an ARM architecture; the six-axis inertial measurement unit adopts a six-axis inertial sensor; and the vibration unit adopts a flat vibration motor with a model number of XHXDZ-1234.

[0014] Preferably, the voltage of the electric pulse of the dual-channel electric stimulation unit is 10-35V, the frequency is 2-10 Hz, and the pulse width is 100-800 μs.

[0015] Preferably, the two hydrogel electrode patches of the electric pulse channel 1 are respectively arranged on the surface of the tibial nerve on the inner side of the ankle and the surface of the sural nerve on the outer side of the ankle, and are responsible for causing tactile sensation in the heel.

[0016] Preferably, the two hydrogel electrode patches of the electric pulse channel 2 are respectively arranged on the surface of the tibial nerve above the inner side of the ankle and on the surface of the tibial nerve below the inner side of the ankle, and are responsible for causing tactile sensation in the forefoot.

[0017] Preferably, during the swing phase of walking, neither the electric pulse channel 1 nor the electric pulse channel 2 emits a pulse;

[0018] When the heel touches the ground, the electric pulse channel 1 sends out pulses, and the electric pulse channel 2 does not send out pulses, which causes tactile sensation in the heel;

[0019] When the whole foot is touching the ground, the electric pulse channel 1 and the electric pulse channel 2 send pulses simultaneously, causing tactile sensations in the heel and forefoot, that is, sensations in the whole sole of the foot;

[0020] During the forefoot push-off phase, the electrical pulse channel 2 sends out pulses, while the electrical pulse channel 1 does not send out pulses, thereby causing tactile sensation in the forefoot.

[0021] Preferably, the first and second vibration units are respectively arranged at the Achilles tendon and the tibialis anterior muscle on the body surface, wherein the first vibration unit causes tactile sensation in the back of the calf when vibrating, and the spatial encoding corresponds to the heel, and the second vibration unit causes tactile sensation in the front of the calf when vibrating, and the spatial encoding corresponds to the front of the foot.

[0022] Preferably, the spatial tactile reconstruction pathway is constructed by controlling the first and second vibration units to vibrate alternately or simultaneously in corresponding phases, as follows:

[0023] During the swing phase of walking, neither the first nor the second vibration unit vibrates;

[0024] During the heel-strike phase, the first vibration unit vibrates, while the second vibration unit does not vibrate. At this time, the tactile sensation of the heel is encoded to the back of the calf through spatial tactile encoding;

[0025] During the full-foot strike phase, both the first and second vibrations vibrate, encoding the full-foot sensory space to the front and back of the lower leg;

[0026] During the forefoot push-off phase, the second vibration unit vibrates and the first vibration unit does not vibrate. At this time, the tactile sensation of the forefoot is spatially encoded to the front of the calf.

[0027] Preferably, the inertial data includes walking data collected synchronously using a plantar pressure sensor and a six-axis inertial sensor;

[0028] The sliding window adopts a moving mode of a window with a step length of 5 and a sliding step length of 2. The window slides forward 2 inertial data points each time, and a new feature vector is calculated after each sliding.

[0029] Preferably, the gait phase is classified in real time on the feature vector by a random forest model, and the classification result is corrected in real time by an objective sequential logic module; specifically, the following steps are included:

[0030] S51), using fuzzy rules to label the feature vector extracted from the sliding window, and using the labeled feature vector to train the random forest model;

[0031] S52), using the pre-trained random forest model to perform real-time classification of gait phases on the feature vector;

[0032] S53), the classification results are corrected in real time through the objective temporal logic module.

[0033] Preferably, in step S51), the feature vector extracted from the sliding window is labeled using fuzzy rules, as follows:

[0034] S511), using the plantar pressure sensor and the six-axis inertial sensor to synchronously collect walking data for a certain period of time;

[0035] S512), divide the plantar pressure data into three areas: heel area, midfoot area and forefoot area, and sum them up respectively, then set pressure thresholds for the three areas respectively, and judge it as "1" if it is greater than the pressure threshold, that is, "stepping down"; judge it as "0" if it is less than the pressure threshold, that is, "lifting up"; and use A, B, C, and D to represent "swing phase", "heel landing phase", "whole foot landing phase", and "forefoot push-off phase" respectively; to set fuzzy rules.

[0036] Preferably, in step S53), the classification result is corrected in real time by an objective temporal logic module, as follows:

[0037] After the "swing phase" state of the preset time, the state will be changed only when the gait classification result is the "heel strike phase", and the state will not be changed when it is the other classification results;

[0038] In the "heel-on-the-ground phase" state, the state will only change when the classification result of the random forest model is "full-foot-on-the-ground phase", and the state will not change when it is the other classification results;

[0039] In the "full foot landing phase" state, the state will only change when the gait classification result is "forefoot push-off phase", and the state will not change when it is the other classification results;

[0040] In the state of "forefoot push-off phase", the state is changed only when the gait classification result is "swing phase", and the state does not change when it is the other classification results;

[0041] When the "heel touching the ground phase" and the "whole foot touching the ground phase" appear successively in a short period of time, the "forefoot pushing off phase" appears immediately after the "whole foot touching the ground phase" ends.

[0042] Preferably, the present invention also provides a method for foot orthotics with plantar tactile reconstruction, comprising the following steps:

[0043] S1), setting a dual-channel electrical stimulation unit at a specific position of the ankle, wherein the electrical pulse channel 1 of the dual-channel electrical stimulation unit stimulates the tibial nerve and the sural nerve of the ankle and can induce tactile sensation in the heel; the electrical pulse channel 2 only stimulates the tibial nerve and can induce tactile sensation in the front of the sole;

[0044] S2), fixing two vibration units at the tibialis anterior muscle and the Achilles tendon on the body surface, and the two mechanical vibration units vibrate alternately or simultaneously at corresponding phases to build a spatial tactile reconstruction pathway;

[0045] S3), using a six-axis inertial measurement unit to collect inertial data during walking, and using a sliding window to extract feature vectors from the collected inertial data;

[0046] S4), real-time classification of gait phases of feature vectors through pre-trained random forest model,

[0047] And make real-time corrections through the classification results of the objective sequential logic module;

[0048] S5) In the corresponding gait phase, the electric pulses of the dual-channel electrical stimulation unit are used to evoke the pressure sensation of different areas of the sole to complete the real-time plantar tactile reconstruction process, and at the same time, the mechanical vibration of the corresponding tibialis anterior muscle and Achilles tendon on the body surface is used to complete the spatial encoding tactile reconstruction process.

[0049] The beneficial effects of the present invention are:

[0050] 1. The present invention can activate the tactile sensation of different areas of the sole of the foot in real time when the user walks, realize ankle and foot tactile reconstruction, and significantly improve the user's walking experience;

[0051] 2. The present invention uses a sliding window to extract the feature vector of the six-axis inertial measurement unit in real time; uses an efficient random forest algorithm to intelligently divide the gait phase to ensure accurate gait analysis and reduce the risk of overfitting; and uses an objective sequential logic module to correct the classification results in real time to improve the reliability of the results;

[0052] 3. The present invention solves the deficiencies of traditional ankle-foot orthoses in tactile perception and gait control, and has good social benefits and broad application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 It is a schematic diagram of the process of the present invention;

[0054] Figure 2 It is a structural framework diagram of the ankle-foot orthosis system of the present invention;

[0055] Figure 3 A schematic diagram of the positions of the dual-channel electrical stimulation unit and the vibration unit of the present invention;

[0056] Figure 4 It is a schematic diagram of stimulation by a dual-channel electrical stimulation unit of the present invention;

[0057] Figure 5 This is a schematic diagram of extracting feature vectors through a sliding window according to the present invention;

[0058] Figure 6 This is a structural diagram of the random forest model pre-trained in the present invention;

[0059] Figure 7 It is a schematic diagram of the objective sequential logic module of the present invention. DETAILED DESCRIPTION

[0060] The specific implementation of the present invention will be further described below in conjunction with the accompanying drawings:

[0061] Example 1

[0062] like Figure 1 As shown, this embodiment provides an ankle-foot orthosis method with plantar tactile reconstruction, comprising the following steps:

[0063] S1), wearing the ankle-foot orthosis on the user's ankle and foot; Figure 2 As shown, the ankle-foot orthosis system includes a main control unit, and a six-axis inertial measurement unit, two vibration units, and a dual-channel electrical stimulation unit connected to the main control unit; the ankle-foot orthosis system also includes a Bluetooth module and a 3.7V rechargeable lithium battery. The 3.7V rechargeable lithium battery increases the 3.7V voltage to 5V through a voltage conversion circuit to power the main control unit, and reduces the 3.7V voltage to 3.3V to power the six-axis inertial measurement unit. In this embodiment, the Bluetooth module is connected to a mobile terminal, such as a mobile phone, a tablet, etc., and the intensity of the vibration unit and the dual-channel electrical stimulation unit can be controlled through the mobile terminal.

[0064] The dual-channel electrical stimulation unit adopts two pairs of hydrogel electrode patches and uses a dual-channel two-phase rectangular wave generating circuit to generate rectangular wave electrical pulses with different pulse widths and frequencies. The dual-channel two-phase rectangular wave generating circuit includes a boost circuit and two H-bridge circuits using DC motor drive chips. The boost circuit boosts the 5V voltage to 12V~35V; and generates a dual-channel two-phase pulse waveform through two H-bridge circuits using DC motor drive chips and outputs it through the hydrogel electrode pairs.

[0065] In motor control, the voltage input of the H-bridge remains unchanged, and by controlling the switching state of the four diode switches, the positive and negative polarity of the output voltage can be reversed, thereby achieving the purpose of controlling the direction and speed of the motor. The four diodes are controlled in pairs, with the upper left and lower right switches forming a pair. To create the positive part of the output pulse, one pair of diodes is turned on and the other pair of diodes is turned off. Conversely, to form a negative pulse, the previously closed pair of diodes is now turned on and the previously opened pair of diodes is now turned off. For the rest of the cycle, all diodes are turned on and the output is 0V.

[0066] In this embodiment, the main control unit adopts an ARM architecture microcontroller of model STM32F407ZGT6; the six-axis inertial measurement unit is a six-axis inertial sensor; and the vibration unit adopts a flat vibration motor of model XHXDZ-1234.

[0067] S2), placing a dual-channel electrical stimulation unit at a specific location on the ankle, wherein Figure 3 As shown, the electrical pulse channel 1 of the dual-channel electrical stimulation unit stimulates the tibial nerve and the sural nerve of the ankle and can cause tactile sensation in the heel; the electrical pulse channel 2 only stimulates the tibial nerve and can cause tactile sensation in the front of the sole;

[0068] In this embodiment, the voltage of the electric pulse of the dual-channel electric stimulation unit is 10-35V, the frequency is 2-10 Hz, and the pulse width is 100-800 μs.

[0069] The two hydrogel electrode patches of the electric pulse channel 1 are respectively arranged on the surface of the tibial nerve on the inner side of the ankle and the surface of the sural nerve on the outer side of the ankle, which are responsible for causing tactile sensation in the heel. The two hydrogel electrode patches of the electric pulse channel 2 are respectively arranged on the surface of the tibial nerve on the upper inner side of the ankle and the surface of the tibial nerve on the lower inner side of the ankle, which are responsible for causing tactile sensation in the forefoot.

[0070] In this embodiment, Figure 4 As shown, during the swing phase of walking, the electric pulse channel 1 and the electric pulse channel 2 do not send out pulses;

[0071] When the heel touches the ground, the electric pulse channel 1 sends out pulses, and the electric pulse channel 2 does not send out pulses, which causes tactile sensation in the heel;

[0072] When the whole foot is touching the ground, the electric pulse channel 1 and the electric pulse channel 2 send pulses simultaneously, causing tactile sensations in the heel and forefoot, that is, sensations in the whole sole of the foot;

[0073] During the forefoot push-off phase, the electrical pulse channel 2 sends out pulses, while the electrical pulse channel 1 does not send out pulses, thereby causing tactile sensation in the forefoot.

[0074] S3), fixing two vibration units at the tibialis anterior muscle and the Achilles tendon on the body surface, and the two mechanical vibration units vibrate alternately or simultaneously at corresponding phases to build a spatial tactile reconstruction pathway; setting the position as follows Figure 3 shown.

[0075] In this embodiment, the first and second vibration units are respectively arranged at the Achilles tendon and the tibialis anterior muscle on the body surface, wherein the first vibration unit causes tactile sensation in the back of the calf when vibrating, and the spatial encoding corresponds to the heel, and the second vibration unit causes tactile sensation in the front of the calf when vibrating, and the spatial encoding corresponds to the front of the foot.

[0076] The spatial tactile reconstruction pathway is constructed by controlling the first and second vibration units to vibrate alternately or simultaneously in corresponding phases, as follows:

[0077] During the swing phase of walking, neither the first nor the second vibration unit vibrates;

[0078] During the heel-strike phase, the first vibration unit vibrates, while the second vibration unit does not vibrate. At this time, the tactile sensation of the heel is encoded to the back of the calf through spatial tactile encoding;

[0079] During the full-foot strike phase, both the first and second vibrations vibrate, encoding the full-foot sensory space to the front and back of the lower leg;

[0080] During the forefoot push-off phase, the second vibration unit vibrates and the first vibration unit does not vibrate. At this time, the tactile sensation of the forefoot is spatially encoded to the front of the calf.

[0081] S4), using a six-axis inertial measurement unit to collect inertial data during walking, and using a sliding window to extract feature vectors from the collected inertial data;

[0082] In this embodiment, the inertial data includes walking data synchronously collected using a plantar pressure sensor and a six-axis inertial sensor; in this embodiment, the collection frequency of the six-axis inertial measurement unit is 50 Hz, that is, one piece of inertial data is collected every 20 ms, which is recorded as a 1x6 data vector.

[0083] like Figure 5As shown, in this embodiment, the window width of the sliding window is set to 5 steps, that is, the length of a window is 100ms, forming a 5x6 data matrix. For each axis in the inertial data, that is, each column in the 5x6 data matrix, six eigenvalues ​​are calculated, namely the maximum value, minimum value, mean value, root mean square value, current value and waveform height; after calculation, a 1x36 eigenvector is extracted from the 5x6 data matrix to represent the gait characteristics within the 100ms window; in addition, this embodiment adopts a window movement mode with a sliding step length of 2, and the window slides forward 2 inertial data points each time, and a new eigenvector is calculated after each sliding.

[0084] S5), using the pre-trained random forest model to classify the gait phase of the feature vector in real time, and using the objective timing logic module to perform real-time correction on the classification result; specifically including the following steps:

[0085] S51), using fuzzy rules to label the feature vector extracted from the sliding window, and using the labeled feature vector to train the random forest model; the details are as follows:

[0086] S511), using the plantar pressure sensor and the six-axis inertial sensor to synchronously collect walking data for a certain period of time;

[0087] S512), the plantar pressure data is divided into three areas: the heel area, the midfoot area and the forefoot area, and the sum is calculated respectively. Then, pressure thresholds are set for the three areas respectively. If the pressure is greater than the pressure threshold, it is judged as "1", that is, "stepping down"; if the pressure is less than the pressure threshold, it is judged as "0", that is, "lifting up"; and A, B, C, and D are used to represent "swing phase", "heel landing phase", "whole foot landing phase", and "forefoot push-off phase" respectively; to set fuzzy rules. The fuzzy rule settings are shown in Table 1:

[0088] Table 1 Fuzzy rules

[0089]

[0090] For the synchronously collected six-axis inertial data, sliding window feature extraction is performed with a window width of 5 and a step length of 2 to obtain a feature vector, and the true label of the feature vector is annotated by the above fuzzy rules. The annotated feature vector is then used to train and test the random forest model on the server, and then deployed to the main control unit.

[0091] S52), using the pre-trained random forest model to perform real-time classification of gait phases on the feature vector;

[0092] In this embodiment, Figure 6As shown in the figure, the pre-trained random forest model has about 30 decision trees, and the average depth of each decision tree is about 15. By integrating the results of multiple decision trees, it can effectively handle complex nonlinear relationships and resist the overfitting problem that may occur in a single decision tree. The real-time performance of the random forest model also ensures that the intelligent ankle-foot orthosis system can quickly respond to the user's gait changes, thereby providing the necessary support and adjustments to optimize the user's walking experience.

[0093] S53), the classification results are corrected in real time through the objective temporal logic module. The details are as follows:

[0094] like Figure 7 As shown, after the "swing phase" state of the preset time, the state is changed only when the gait classification result is the "heel strike phase", and the state does not change when it is the other classification results;

[0095] In the "heel-on-the-ground phase" state, the state will only change when the classification result of the random forest model is "full-foot-on-the-ground phase", and the state will not change when it is the other classification results;

[0096] In the "full foot landing phase" state, the state will only change when the gait classification result is "forefoot push-off phase", and the state will not change when it is the other classification results;

[0097] In the state of "forefoot push-off phase", the state is changed only when the gait classification result is "swing phase", and the state does not change when it is the other classification results;

[0098] When the "heel touching the ground phase" and the "whole foot touching the ground phase" appear successively in a short period of time, the "forefoot pushing off phase" appears immediately after the "whole foot touching the ground phase" ends.

[0099] S6) In the corresponding gait phase, the electric pulses of the dual-channel electrical stimulation unit are used to evoke the pressure sensation of different areas of the sole to complete the real-time sole tactile reconstruction process. At the same time, the mechanical vibration of the vibration unit corresponding to the tibialis anterior muscle and the Achilles tendon on the body surface is used to complete the spatial encoding tactile reconstruction process.

[0100] Example 2

[0101] like Figure 2 As shown, this embodiment provides an intelligent ankle-foot orthosis with plantar tactile reconstruction, including a main control unit, and a six-axis inertial measurement unit, two vibration units, and a dual-channel electrical stimulation unit connected to the main control unit; the ankle-foot orthosis system also includes a Bluetooth module and a 3.7V rechargeable lithium battery. The 3.7V rechargeable lithium battery increases the 3.7V voltage to 5V through a voltage conversion circuit to power the main control unit, and reduces the 3.7V voltage to 3.3V to power the six-axis inertial measurement unit. The main control unit controls the two vibration units and the dual-channel electrical stimulation unit through a PWM signal.

[0102] In this embodiment, the main control unit adopts an ARM architecture microcontroller of model STM32F407ZGT6; the six-axis inertial measurement unit includes a six-axis inertial sensor; and the vibration unit adopts a flat vibration motor of model XHXDZ-1234.

[0103] In this embodiment, the dual-channel electrical stimulation unit generates rectangular wave electrical pulses of different pulse widths and frequencies using a dual-channel biphasic rectangular wave generating circuit, and the dual-channel biphasic rectangular wave generating circuit includes a boost circuit and two H-bridge circuits using a DC motor driver chip, and the boost circuit boosts the 5V voltage to 12V to 35V; and generates a dual-channel biphasic pulse waveform through two H-bridge circuits using a DC motor driver chip and outputs it through a hydrogel electrode pair. In motor control, the voltage input of the H-bridge remains unchanged, and the positive and negative polarity of the output voltage can be reversed by controlling the switching state of the four diode switches, thereby achieving the purpose of controlling the direction and speed of the motor. The four diodes are controlled in pairs, and the upper left and lower right switches form a pair. In order to create the positive part of the output pulse, one pair of diodes is turned on, and the other pair of diodes is turned off. Conversely, in order to form a negative pulse, the previously closed pair of diodes is now turned on, and the previously opened pair of diodes is now turned off. For the remaining cycle, all diodes are turned on and the output is 0V.

[0104] The electrical pulse channel 1 of the dual-channel electrical stimulation unit stimulates the tibial nerve and the sural nerve of the ankle and can cause tactile sensation in the heel; the electrical pulse channel 2 only stimulates the tibial nerve and can cause tactile sensation in the front of the sole;

[0105] In this embodiment, the voltage of the electric pulse of the dual-channel electric stimulation unit is 10-35V, the frequency is 2-10 Hz, and the pulse width is 100-800 μs.

[0106] The two hydrogel electrode patches of the electric pulse channel 1 are respectively arranged on the surface of the tibial nerve on the inner side of the ankle and the surface of the sural nerve on the outer side of the ankle, which are responsible for causing tactile sensation in the heel. The two hydrogel electrode patches of the electric pulse channel 2 are respectively arranged on the surface of the tibial nerve on the upper inner side of the ankle and the surface of the tibial nerve on the lower inner side of the ankle, which are responsible for causing tactile sensation in the forefoot.

[0107] In this embodiment, during the swing phase of walking, the electric pulse channel 1 and the electric pulse channel 2 do not emit pulses;

[0108] When the heel touches the ground, the electric pulse channel 1 sends out pulses, and the electric pulse channel 2 does not send out pulses, which causes tactile sensation in the heel;

[0109] When the whole foot is touching the ground, the electric pulse channel 1 and the electric pulse channel 2 send pulses simultaneously, causing tactile sensations in the heel and forefoot, that is, sensations in the whole sole of the foot;

[0110] During the forefoot push-off phase, the electrical pulse channel 2 sends out pulses, while the electrical pulse channel 1 does not send out pulses, thereby causing tactile sensation in the forefoot.

[0111] In this embodiment, two vibration units are fixed on the tibialis anterior muscle and the Achilles tendon on the body surface, and the two mechanical vibration units vibrate alternately or simultaneously in corresponding phases to build a spatial tactile reconstruction pathway;

[0112] In this embodiment, the first and second vibration units are respectively arranged at the Achilles tendon and the tibialis anterior muscle on the body surface, wherein the first vibration unit causes tactile sensation in the back of the calf when vibrating, and the spatial encoding corresponds to the heel, and the second vibration unit causes tactile sensation in the front of the calf when vibrating, and the spatial encoding corresponds to the front of the foot.

[0113] The spatial tactile reconstruction pathway is constructed by controlling the first and second vibration units to vibrate alternately or simultaneously in corresponding phases, as follows:

[0114] During the swing phase of walking, neither the first nor the second vibration unit vibrates;

[0115] During the heel-strike phase, the first vibration unit vibrates, while the second vibration unit does not vibrate. At this time, the tactile sensation of the heel is encoded to the back of the calf through spatial tactile encoding;

[0116] During the full-foot strike phase, both the first and second vibrations vibrate, encoding the full-foot sensory space to the front and back of the lower leg;

[0117] During the forefoot push-off phase, the second vibration unit vibrates and the first vibration unit does not vibrate. At this time, the tactile sensation of the forefoot is spatially encoded to the front of the calf.

[0118] In this embodiment, a six-axis inertial measurement unit is used to collect inertial data during walking, and a sliding window is used to extract feature vectors from the collected inertial data;

[0119] The main control unit is integrated with a random forest model and an objective sequential logic module; this embodiment uses the random forest model to perform real-time classification of gait phases on feature vectors, and uses the objective sequential logic module to perform real-time correction on the classification results.

[0120] The above embodiments and descriptions are only for illustrating the principles and best embodiments of the present invention. Without departing from the spirit and scope of the present invention, the present invention may be subject to various changes and improvements, all of which fall within the scope of the present invention to be protected.

Claims

1. An intelligent ankle-foot orthosis with plantar tactile reconstruction, characterized in that: It includes a main control unit, and a six-axis inertial measurement unit, two vibration units, and a dual-channel electrical stimulation unit connected to the main control unit; Wherein, the electrical pulse channel 1 of the dual-channel electrical stimulation unit stimulates the tibial nerve and the sural nerve of the ankle and can cause tactile sensation in the heel; the electrical pulse channel 2 only stimulates the tibial nerve and can cause tactile sensation in the front of the sole; The two vibration units are fixed on the tibialis anterior muscle and the Achilles tendon on the body surface, and the two vibration units vibrate alternately or simultaneously in corresponding phases to build a spatial tactile reconstruction pathway; The six-axis inertial measurement unit collects inertial data during walking, and uses a sliding window to extract feature vectors from the collected inertial data; The main control unit is integrated with a random forest model and an objective sequential logic module; the feature vector is classified in real time by the pre-trained random forest model, and the classification result is corrected in real time by the objective sequential logic module; In the corresponding gait phase, the electrical pulses of the dual-channel electrical stimulation unit are used to evoke the pressure sensation of different partitions of the sole of the foot to complete the real-time plantar tactile reconstruction process. At the same time, the mechanical vibration of the vibration unit corresponding to the tibialis anterior muscle and the Achilles tendon on the body surface is used to complete the spatial encoding tactile reconstruction process.

2. The intelligent ankle-foot orthosis with plantar tactile reconstruction according to claim 1, characterized in that: The ankle-foot orthosis also includes a Bluetooth module and a 3.7V rechargeable lithium battery. The 3.7V rechargeable lithium battery increases the voltage from 3.7V to 5V through a voltage conversion circuit to power the main control unit, and reduces the voltage from 3.7V to 3.3V to power the six-axis inertial measurement unit.

3. The intelligent ankle-foot orthosis with plantar tactile reconstruction according to claim 1, characterized in that: The two hydrogel electrode patches of the electric pulse channel 1 are respectively arranged on the surface of the tibial nerve on the inner side of the ankle and the surface of the sural nerve on the outer side of the ankle, and are responsible for causing tactile sensation in the heel; The two hydrogel electrode patches of the electric pulse channel 2 are respectively arranged on the surface of the tibial nerve above the inner side of the ankle and on the surface of the tibial nerve below the inner side of the ankle, and are responsible for causing tactile sensation in the forefoot.

4. The intelligent ankle-foot orthosis with plantar tactile reconstruction according to claim 3, characterized in that: During the swing phase of walking, the electric pulse channel 1 and the electric pulse channel 2 do not emit pulses; When the heel touches the ground, the electric pulse channel 1 sends out pulses, and the electric pulse channel 2 does not send out pulses, which causes tactile sensation in the heel; When the whole foot is touching the ground, the electric pulse channel 1 and the electric pulse channel 2 send pulses simultaneously, causing tactile sensations in the heel and forefoot, that is, sensations in the whole sole of the foot; During the forefoot push-off phase, the electrical pulse channel 2 sends out pulses, while the electrical pulse channel 1 does not send out pulses, thereby causing tactile sensation in the forefoot.

5. The intelligent ankle-foot orthosis with plantar tactile reconstruction according to claim 1, characterized in that: The first and second vibration units are respectively arranged at the Achilles tendon and the tibialis anterior muscle on the body surface, wherein the first vibration unit causes tactile sensation in the back of the calf when vibrating, and the spatial encoding corresponds to the heel, and the second vibration unit causes tactile sensation in the front of the calf when vibrating, and the spatial encoding corresponds to the front of the foot.

6. The intelligent ankle-foot orthosis with plantar tactile reconstruction according to claim 5, characterized in that: The spatial tactile reconstruction pathway is constructed by controlling the first and second vibration units to vibrate alternately or simultaneously in corresponding phases, as follows: During the swing phase of walking, neither the first nor the second vibration unit vibrates; During the heel-strike phase, the first vibration unit vibrates, while the second vibration unit does not vibrate. At this time, the tactile sensation of the heel is encoded to the back of the calf through spatial tactile encoding; During the full-foot strike phase, both the first and second vibrations vibrate, encoding the full-foot sensory space to the front and back of the lower leg; During the forefoot push-off phase, the second vibration unit vibrates and the first vibration unit does not vibrate. At this time, the tactile sensation of the forefoot is spatially encoded to the front of the calf.

7. The intelligent ankle-foot orthosis with plantar tactile reconstruction according to claim 1, characterized in that: The gait phase is classified in real time by the feature vector through the random forest model, and the classification result is corrected in real time by the objective timing logic module; the specific steps include the following: S51), using fuzzy rules to label the feature vector extracted from the sliding window, and using the labeled feature vector to train the random forest model; S52), using the pre-trained random forest model to perform real-time classification of gait phases on the feature vector; S53), the classification results are corrected in real time through the objective temporal logic module.

8. The intelligent ankle-foot orthosis with plantar tactile reconstruction according to claim 7, characterized in that: In step S51), the feature vector extracted by the sliding window is labeled using fuzzy rules, as follows: S511), using the plantar pressure sensor and the six-axis inertial sensor to synchronously collect walking data for a certain period of time; S512), the plantar pressure data is divided into three areas: the heel area, the midfoot area and the forefoot area, and the sums are calculated respectively. Then, pressure thresholds are set for the three areas respectively. If the pressure is greater than the pressure threshold, it is judged as "1", that is, "stepping down"; if the pressure is less than the pressure threshold, it is judged as "0", that is, "lifting up"; and A, B, C, and D are used to represent the "swing phase", "heel landing phase", "whole foot landing phase", and "forefoot push-off phase" respectively; to set fuzzy rules.

9. The intelligent ankle-foot orthosis with plantar tactile reconstruction according to claim 1, characterized in that: In step S53), the classification result is corrected in real time by the objective temporal logic module, as follows: After the "swing phase" state of the preset time, the state will be changed only when the gait classification result is the "heel strike phase", and the state will not be changed when it is the other classification results; In the "heel-on-the-ground phase" state, the state will only change when the classification result of the random forest model is "full-foot-on-the-ground phase", and the state will not change when it is the other classification results; In the "full foot landing phase" state, the gait classification result will change state only when it reaches the "forefoot push-off phase", and will not change state for the other classification results; In the state of "forefoot push-off phase", the state will be changed only when the gait classification result is "swing phase", and the state will not be changed when it is the other classification results; When the "heel touching the ground phase" and the "whole foot touching the ground phase" appear successively in a short period of time, the "forefoot pushing off phase" appears immediately after the "whole foot touching the ground phase" ends.

10. A method for performing ankle-foot orthosis using the ankle-foot orthosis according to any one of claims 1 to 9, characterized in that: The following steps are involved: S1), setting a dual-channel electrical stimulation unit at a specific position of the ankle, wherein the electrical pulse channel 1 of the dual-channel electrical stimulation unit stimulates the tibial nerve and the sural nerve of the ankle and can induce tactile sensation in the heel; the electrical pulse channel 2 only stimulates the tibial nerve and can induce tactile sensation in the front of the sole; S2), fixing two vibration units at the tibialis anterior muscle and the Achilles tendon on the body surface, and the two mechanical vibration units vibrate alternately or simultaneously at corresponding phases to build a spatial tactile reconstruction pathway; S3), using a six-axis inertial measurement unit to collect inertial data during walking, and using a sliding window to extract feature vectors from the collected inertial data; S4), real-time classification of gait phases of feature vectors through pre-trained random forest models, and real-time correction of classification results through objective timing logic modules; S5) In the corresponding gait phase, the electric pulses of the dual-channel electrical stimulation unit are used to evoke the pressure sensation of different areas of the sole to complete the real-time sole tactile reconstruction process. At the same time, the mechanical vibration of the vibration unit corresponding to the tibialis anterior muscle and the Achilles tendon on the body surface is used to complete the spatial encoding tactile reconstruction process.

Images