TECHNICAL FIELD

The invention of the application relates to an smart hand prosthesis that can be easily learned with short-term training without the need for long rehabilitation processes, and thus encourages individuals who have undergone hand amputation to use the smart hand prosthesis continuously, supports the majority of daily activities, can easily and naturally adapt to the environment and surrounding objects by changing their stiffness like human hand, appeals to a large audience at a lower cost compared to the existing prostheses in the market, can be designed according to the physical dimensions of the user, and can be controlled by the natural control interface. The position change of the prosthesis is achieved through the motor control, while the stiffness change is achieved by controlling the mechanical properties of the shape memory materials.

STATE OF THE ART

Half of the upper extremity amputee individuals in the world receive prosthesis service, while only half of this group of service recipients can consistently continue to use prosthesis according to the researches conducted by the World Health Organization. Among the leading reasons for the decrease in the usage rate are the fact that these prostheses, which lack a natural control interface, are not easy and convenient to use during the realization of daily activities and yet they are offered for sale at high prices in the market. This invention includes a low-cost hand prosthesis design with a natural control interface, which can be produced in sizes specific to the needs of the person, which can change its stiffness naturally according to the nature of the object in order to solve the specified problem.

The way of controlling commercial hand prostheses in the market is fairly different from the control architecture that human beings have learned from birth today. Amputees stop using these prostheses due to their interfaces that require long learning process. The subject of multifaceted comprehension and manipulation in unpredictable environments is among the active and challenging research subjects of prosthesis and robotics. Studies are carried out on multi-finger robot hands in both academic [1-3] and commercial [4,5] fields in order to achieve various tasks. Anthropomorphism (the ability to imitate the human-like hand shape) and skill (the ability to successfully manipulate even under uncertain conditions) are defined as the basic characteristics required to perform tasks successfully. Anthropomorphism stands out as an imperative design criterion in the design of robotic end-effectors, particularly and prosthesis [6,7] that replaces with a missing hand since tools in our environment such as consoles, handles, keys are designed according to the anthropomorphic structure of human hands. Meanwhile, since the anthropomorphic designs have a natural appearance, they are preferred by amputees for aesthetic and psychological reasons. However, anthropomorphism alone is not enough for an optimal hand prosthesis design; robust design, ease of use and sufficient dexterity should be evaluated as important design criteria.

Dexterity is an optimal performance criterion for robotic and prosthetic hands to have human abilities such as capturing and holding objects and performing sensitive manipulations. It should be able to perform most of the human-hand taxonomy required during daily life activities in order for a prosthetic hand to be considered as a dexterous design [8,9], Studies in the literature [10,11] emphasize that the human-hand taxonomy required for more than 50% of daily activities (especially 63% for households and 56% for machine workers) is power grasp. Pinch grasp ranks second in the order of preference in daily activities [9], Therefore, being able to hold a wide variety of objects by grasping them and carrying out the task by holding the objects in the other majority are sufficient to achieve the majority of activities of daily living with hand prostheses.

Successful manipulation (holding, grasping) also requires a very important feature of the human hand, impedance modulation (the ability to change its stiffness). Gaining impedance modulation property to the design of hand prothesis in hand prosthesis design allows the prosthesis to be adaptable according to the objects/tasks it interacts with, in other words, to adapt. The fact that people can successfully interact physically with their environment is due to the fact that they can adjust their impedance levels in the most appropriate way according to the changing needs of the task. For example, in activities that require precise position control such as writing and painting, while the stiffness level of our fingers reaches a very high value, we reduce the stiffness of our fingers to low levels in the manipulation of soft or fragile objects such as holding an egg without breaking, holding a sponge without deforming or holding an apricot without harming. The ability of human to adjust limb impedance (stiffness) has been particularly inspiring for the design and control strategies of robotic and prosthetic hands, which will undertake tasks that require interaction with the environment in unpredictable circumstances.

In particular, hand prostheses are safer and more functional if they have an appropriate level of stiffness stifness based on the physical conditions of the environment they are in/interact with [12-14], Recent studies [15-17] provide strong evidence that the performance of the amputee is improved if the stiffness level of the hand prosthesis is regulated based on the impedance level of the activity.

Impedance modulation approaches in robotics can be divided into three main categories. The compatibility of the device can be modulated through the software using strategies such as impedance/admittance control. In this approach, impedance modulation is limited to the controllable bandwidth of the actuators. Therefore, a prosthesis whose impedance is modulated by such a control strategy acts as a rigid above the frequencies exceeded the controllable bandwidth, such as impacts [13-14], Moreover, this approach considerably reduces the energy efficiency of the system as it requires the continuous use of the actuators. Alternatively, impedance modulation may also be accomplished by embedding the special mechanical elements into the robotic systems.

In this approach, the impedance of the robotic manipulator is adjusted through special mechanisms consisting of passive elastic elements, such as springs. The impedance variation takes place physically (mechanically) and across the entire frequency spectrum, including frequencies above the controllable bandwidth of the actuators such as a variable stiffness actuator (VSA). In addition, stiffness modulation by means of VSA is also known as an energy-efficient approach since the energy is only consumed during the impedance modulation. Materials with controllable stiffness have made many new applications possible in this field, as well [18,19], Two categories that are frequently used and stand out from their derivatives in terms of features are shape memory polymers (SMP) and shape memory alloys (SMA). SMP and SMAs can change their stiffness with an external stimulus, and this can act as a trigger for the desired action to be changed [20,21], This stimulus for SMP can be a chemical or light, the most prevalently preferred stimulator for both SMPs and SMAs is stimulating by means of heat. When SMPs are heated above glassy transition temperatures (Tg), they soften and a sudden variation occurs in their stiffness. Especially low melting point alloys (LMPA) are noteworthy as they provide ease of application among SMAs. LMPAs soften when heated, but this variation occurs on a smaller scale than SMPs [21], Heat is applied directly to SMAs to capture and utilize this variation as they are conductive, while SMPs are usually applied by a secondary heating system embedded between the polymer layers. This disadvantage in SMPs has been tried to be overcome by mixing these polymers with conductive materials, but in this case, the additive materials causes a variation in the mechanical properties of the polymer. On the other hand despite LMPA stands out, SMPs were considered as relatively cost-effective ones in shape mamory materials considering the economic perspective.

Fully actuated hand prostheses with high degree of freedom require the use of complex control algorithms [22], There are myoelectrically controlled hand prostheses in the literature that can perform various manipulation tasks [23-25], However, in these systems the control of each finger joint can be performed through complex learning algorithms that expose amputees to long training periods in these systems.

The control of such prosthetic devices that needs complex control agortihms and accordingly require long rehabilitation periods to be adapted by the amputees lead to abandonment of its use by up to 40% [26], A large percentage of amputees refuse to use active prostheses, since, considering the expectations of the amputees and the complexity of the prostheses, the functionality of the devices is not satisfactory and pleasing [27], Several research groups have focused on simplifying mechanical design and ensuring the control of prostheses in an easy way without losing their basic functionality in response to the low level of adaptation to active prostheses and the high rate of abandonment problems.

In the literature, many robotic hands used for tasks requiring human machine interaction have been designed using hand variable stiffness actuator (VS A) [28-30]; however, no such application has been reported in the field of anthropomorphic hand prostheses yet. In particular, each degree of active freedom of the DLR hand-arm system is controlled by two motors connected to antagonistically regulated non-linear spring elements [28,29], Similarly, the impedance modulation of the anthropomorphic Shadow Hand is provided by pneumatically operated and antagonistically regulated artificial muscles [30], The DLR handarm system and the Shadow Hand have sophisticated mechanical designs with many active degrees of freedom; accordingly, they are not suitable and convenient to use as a hand prosthesis owing to their bulky size, high weight and high cost.

Besides, grasp (pinch) planning and impedance modulation require the processing of control algorithm in the control interface, which makes the use of prosthetics quite difficult.

The use of less actuated mechanisms in hand designs is considered as a promising approach to provide solutions to problems in this field. The prominent reason for this situation is that underactuated hand mechanisms with missing actuators can adapt to various geometries quite well without the need for complex control algorithms and sensors. Underactuated mechanisms are usually included in linkage or tendon based finger designs that can perform the closing sequence of the human finger. Fingers have the ability of shape adaptation and withstand greater forces than tendon based underactuated mechanisms in linkage-based underactuated mechanisms [31-34]; on the other hand, they are not suitable for anthropomorphic hand prostheses due to their bulky design. Applications that are the most successful in terms of being both anthropomorphic and underactuated are observed in tendon-applicable mechanisms [35-38], It is possible to apply both thin and light fingers with this approach. For example, a robotic hand [39-41] with underactuated design, elastic joint that can be applied with sensor integrated tendon was produced by support decomposition manufacturing (SDM) technique. These low-cost hand systems are capable of self-adaptation to different shaped objects under simple control methods thanks to the missing actuator finger mechanisms with elastic joints. Similarly, a robot hand with underactuated design that can be driven with the tendon that investigates the synergies of human hand movements has been developed [42] . However, none of the mentioned missing actuator hands have impedance modulation feature.

More than 40 million amputee individuals live in developing countries [44], and this number is expected to increase in the future [45] according to the World Health Organization. Approximately 50,000 people lose limbs every year in the United States, with upper limb amputations accounting for a quarter of this number in the report published by the national health statistics center [46], Prosthetic devices have been proposed for years in order to eliminate the losses such as the inability of amputated people to do their daily work alone due to limb losses or to leave the job and thus to increase their living standards [47],

The vast majority of people with upper limb amputation do not prefer to wear prostheses even though it provides many benefits. The average rate of rejection of electric and body -moving prostheses is reported to be 35% and 45% in the pediatric population, 23% and 26% in the adult population, respectively in the literature [48], Among the reasons underlying the low acceptance of body-moving prostheses are basically their slow movement, being too heavy, inadequate grip strength, limited functionality, unnatural use, and therefore not being comfortable [48], Myoelectric controlled prostheses are superior to body -moving prostheses in that they can produce a stronger grip force; however, they require a lower amount of muscle strength, are more functional and do not require a fastening belt to attach the prosthesis to the body. Myoelectric controlled hand prostheses have negative effects on consistent use due to reasons such as high cost, being heavy and thus causing pain, charging their batteries frequently on a daily basis, and requiring periodic maintenance despite all these benefits [48,49,82],

Many research groups have tried to close the gap in this area by focusing their studies on enriching the skills and functionality of hand prostheses. The path followed to design a dextrous hand prosthesis in both academia [50-52] and commercial areas [53-56] is to collect surface EMG signal from different muscle groups, assigning a holding pattern to each class by classifying each collected signal. Some recent studies have suggested using different sources of information (MMG [57], NIRS [58], IMU [59]) to increase the success rates of the classification methods used to give their prostheses a large number of functions. Natural and non-intuitive control interfaces cause amputations to be exposed to long training periods [60] even though such studies aim to make the lives of individuals who have undergone amputation easier by giving more function to hand prostheses, and this situation ends with the gradual increase in the rate of release of such prosthetic devices [61],

It is thought that one of the leading features of the human neuromuscular system, which is competent to perform a wide range of physical activities, can solve the mentioned problem. In particular, most of the daily activities that require the interaction of the human hand with the environment are successfully performed using the unrivalled ability in human adaptation. Such an ability is formed by predicting the result of the interaction and regulating the impedance (stiffness level) depending on the activity [62-67], Impedance regulation of the limbs (stiffness modulation takes advantage of reflexive responses that directly contribute to neuromotor control, which plays a role in the interaction of the antagonistic muscle pair with the environment, as well as adjusting the contraction levels. All these abilities enable people to actively engage in their daily activities in a natural way. For example, people guarantee precise position control against perturbations by considerably reducing and hardening the level of stiffness of their muscles during tasks that require high precision (such as writing or drilling holes); on the other hand, people prevent damage to the object to be held by raising the level of stiffness of their limbs (such as fingers) as much as possible when interacting with a soft or fragile object [68],

The human capability to regulate his/her impedance, which is an important feature, has also been inspiring for applications in the field of robotics. Many robotic studies on prosthetic devices have been conducted to mimic the ability to adjust the level of stiffness when interacting physically with one's environment [69-71], For example, a systematic study has provided evidence that task-dependent impedance modulation improves human performance when using virtual arm prosthesis [72-74], Recently, researchers have proposed transradial hand prosthesis with variable stiffness [75], The variable elastic actuator of this prosthesis consists of antagonistically regulated non-linear springs; however, it takes up a large part of the prosthesis.

Surface electromyography (sEMG) signals are widely used for motion (position and speed) control of prosthetic devices [76-78], The surface electromyography (sEMG) signal has a wide range of uses for the diagnosis of disorders in muscle and nerve activities, from clinical studies to physical therapy and rehabilitation training, from biomechanical studies to multifunctional prostheses. sEMG signals are tools that enable communication between man and machine. The possibility of prostheses to have the chance to look like a real limb emerged as a result of the discovery of EMG signals and their association with this field. In particular, commercial hand prostheses such as i-Limb Ultra [53], Bebionic Hand [54], Ottobock Hand [55] and Motion Control Hand [56] are controlled by motion controllers managed in proportion to the amplitude of the sEMG signals measured over the remaining muscle group of the amputee. Some models of commercial prostheses are also integrated with tactile and force feedback sensors, giving the prostheses automatic grip to prevent possible slippage of physically interacting objects. However, performing motion control of these devices requires the amputees to “intentionally” modulate the levels of sEMG signals collected from several different muscle groups. In other words, the amputee is expected to show a high concentration, which in turn brings along long learning processes during the control of the said prostheses. The fact that the control interface, which plays a role in the realization of daily activities of hand prostheses, is far from natural functioning and therefore its use is difficult and difficult to learn by amputees, and their marketing at high prices in the market are among the main reasons indicating a low usage rate [81],

The majority of prostheses in the market are capable of controlling each joint with separate motors, but the functionality of the prosthesis is limited by the ability of the amputee to control the prosthesis. Since the ability of amputees to control such prostheses using EMG signals requires long-term rehabilitation and learning processes, they are shelved without prostheses. The records of users and the comments of the evaluators should be taken into consideration, except for the commercial sites related to the existing prostheses on the market and their own data sheets. Ottobock Michelangelo Hand, which is widely advertised and said to be functional but cannot even perform position control successfully, is only one of them [82],

In the state of the art, there are commercially available multifunctional hand prostheses [81- 83] that require high concentration during education and use.

BRIEF DESCRIPTION OF THE INVENTION

The present invention relates to an smart hand prosthesis that can be easily learned with shortterm training without the need for long rehabilitation processes that eliminate the disadvantages mentioned above, and thus encourages individuals who have undergone hand amputation to use the smart hand prosthesis continuously, supports the majority of daily activities, can easily and naturally adapt to the environment and surrounding objects by changing their stiffness like human hand, appeals to a large audience at a lower cost compared to the existing prostheses in the market, can be designed according to the physical dimensions of the user, can be controlled by the natural control interface.

The object of the invention

■ is to gain the ability to change the stiffness, which is one of the important features of the human hand, by using shape memory materials on the fingers, and

■ with the harmony of the natural human-machine interface and the design similar to the human hand enables the amputations to adapt and maintain in a short time without being exposed to long rehabilitation processes.

It can be marketed in budgets that will appeal to buyers from all walks of life.

The surface electromyography (sEMG) signal used in the human-machine interface in the hand prosthesis of the invention is located to perform its function in an innovative control interface, fairly different from the complex algorithms in commercial prostheses or the applications far from the simple but natural control architecture shown in user training. The amputee will be able to automatically control the stiffness level of the prosthesis according to the need of daily activity with the remote impedance (stiffness) control architecture over the sEMG signal, which we call the natural human-machine interface. The regulate in the impedance of the hand prosthesis is controlled naturally and automatically by the perceived sEMG signals from the live muscle group of the amputee, while the position regulation is controlled depending on the request of the amputee via the perceived sEMG signals from the muscle groups under the forearm (for hand amputations) or under the chest and shoulder (for transradial amputations). Thus, the intact parts of the body are not intervened for impedance control and ease of use for amputation is provided thanks to the interface proposed within the scope of the invention.

The invention has innovative aspects in the mechanical, electronic and control design parts of the hand prosthesis. In other words, the smart hand prosthesis, which is controlled by the natural control interface and has variable stiffness through the shape memory material, is a promising design for amputees to perform their daily activities with high ability and improve their living standards.

The method to be followed to introduce the variable stiffness feature to the prosthesis design will be carried out by including the shape memory materials mentioned in the literature summary in the finger design. All desired features in hand prosthesis such as anthropomorphic structure with underactuated design that can be driven with a tendon, simple control interface, lightness, energy saving and reproducibility according to the size of the person will be among the elements that will be prioritized in the system design in addition to variable stiffness. The harmonious finger design of the prosthesis will make significant contributions to the robustness of the system even under unexpected conditions that may come from the environment such as sudden impacts. It will be possible for amputee individuals to adapt quickly to the use of hand prosthesis and to reduce the long rehabilitation periods in other prostheses to very short periods thanks to the mechanical design in the invention and the natural control interface mentioned in the second part compatible with it.

The design, implementation and evaluation of a control architecture that has a natural humanmachine interface that works in harmony with the electromechanical design of the hand prosthesis with variable stiffness proposed in the first part, and is responsible for controlling both the stiffness and position of the hand prosthesis through the sEMG signals are included in the scope of the study in this part of the study. The elasticity control of the hand prosthesis will be carried out with a control architecture called tele-impedance control [75], An amputated individual becomes a direct part of the natural control interface thanks to the muscle activities perceived by the biopotential electrodes, i.e., the sEMG signals in this section.

In literature [76], a study is considered to estimate the impedance variation of the remnant segment of the limb of an amputee based on the relation between stiffness and EMG, such architecture does not exist in commercial hand prostheses.

The harmony between the design of the natural control interface to be designed and the design of the electromechanical system mentioned in the first part will gain importance in this part. Because the lower and upper limits of the stiffnessvalues of the variable stiffness hand prosthesis will be matched with the stiffness estimation made through sEMG. The angular rotation of the fingers will again be provided through the sEMG signals measured from the relevant muscle groups in the position control of the hand prosthesis. The sEMG signal, which is conditioned to the desired conditions, acts as a reference in the position control architecture of the hand prosthesis after passing through various signal processing methods.

DEFINITIONS OF THE FIGURES DESCRIBING THE INVENTION

The figures used to better explain the piston-driven lock mechanism developed by this invention are as follows. Figure 1 - Solid model of the finger front design (a) Top view of the finger (b) Side view of the finger (c) Placement of the heater in shape memory material

Figure 2 - Design of underactuated finger

Figure 3 - Design of underactuated hand prosthesis which can change its stiffness

Figure 4 - Schematic representation of modeling Shape Memory Materials as parallel coupled springs (a) Finger model with high stiffness level (b) Finger model with intermediate value (medium) stiffness level (c) Finger model with low stiffness level (stiff

DEFINITIONS OF THE ELEMENTS AND PARTS CONSTITUTING THE INVENTION

The parts and elements in the hand prosthesis developed by this invention are individually numbered and are given below.

1. Proximal Phalanx

2. Middle Phalanx

3. Distal Phalanx

4. Proximal Phalanx Tendon

5. Middle Phalanx Tendon

6. Distal Phalanx Tendon

7. Elastic Joint

8. 8

9. 9

10. 10

11. Elastic Joint

12. Finger Bottom Surface

13. Heater

14. 14

15. Metacarpal Bone DETAILED DESCRIPTION OF THE INVENTION

The invention relates to an smart hand prosthesis with variable stiffness, which is customizable, easy to use, in other words, does not require long learning and rehabilitation times before use, can adapt its stiffness to environmental conditions and the characteristics of the object to be held with a natural control interface just like healthy people, controlled by surface electromyography (sEMG) signals. Studies in the literature provide evidence that their ability to change the stiffness of hand prostheses positively affects the user's performance during their daily activities. The ability to adapt the stiffness of the human hand according to the conditions, which is one of the most important characteristics of the human hand, is integrated into finger design with biomedical engineering and chemistry science cooperation in light of biomimetic science for this purpose. Thus, upper extremity amputations (individuals subjected to hand or transradial amputation) can adjust the fingers of the hand prosthesis to high hardness value in situations requiring precise positioning such as picking and placing while making the prosthesis as flexible as necessary during the retention of soft or fragile objects and holding the objects without damaging them.

The smart hand prosthesis mentioned in the invention also provides a solution to one of the biggest problems in this field, "abandoning the use of prosthesis in the short term". Today, 40% of the upper extremity amputations stop using these types of prostheses due to the complex control algorithms within the functional hand prostheses in the market or the long rehabilitation processes and learning difficulties they require due to their incompatibility with the environment and objects with various characteristics (such as flexible, fragile objects), and the majority of the amputations have to prefer passive prostheses used only for cosmetic purposes. The hand prosthesis mentioned in the invention aims to greatly reduce the long rehabilitation process needed in other prostheses thanks to its control interface and electromechanical design compatible with the interface. Amputees are able to use the recommended prosthesis continuously in their daily lives and thus increase the quality of life thanks to its easy to learn use.

While the smart hand prosthesis, which can be controlled through the natural interface with electromyographic signals, is underactuated and can change its stiffness, aims to be economical in terms of energy consumption in terms of being underactuated, it provides the opportunity to address the important activities that a person often needs in terms of presenting the power grasp with all fingers constituting 60% of our daily activities and the pinch grasp functions constituting 20% of our daily activities to an amputee.

One of the original aspects of the hand prosthesis mentioned in the invention, which allows easy adaptation and fast learning process, is provided by the fact that the smart hand prosthesis contains the three basic features detailed below.

The first and most important of the features is that the stiffness of the smart hand prosthesis can be realized automatically according to the fragility, softness or hardness of the object to be held, that is, without requiring the individual to make special concentration and effort. This feature of the smart hand prosthesis is achieved by making use of special materials that can optionally change the stiffness of the fingers and by the cooperation of the natural control interface, in which the EMG signal compatible with this design plays a role. It has the ability to change the stiffness of its fingers not only in the position but also in the human hand as in other hand prostheses of the invention.

The second of the features is that the smart hand prosthesis has power grasp functions, which constitute 60% of the daily activities, and pinch grasp functions, which constitute 20% of the daily activities, and thus, it can hold objects with various geometric shapes by consuming much less energy without requiring complex control algorithms. This feature of the smart hand prosthesis is achieved by the fact that the whole system can be applied with the help of less than the degree of freedom, that is, it is designed as an underactuated system.

The third of the features that enable the realization of the original aspect of the invention is that the position control of the smart hand prosthesis offers the opportunity to be learned after a short training. This feature of the smart prosthesis is achieved by the simultaneous movement of the two motors and five fingers in its design, and thus only one muscle pair takes part in the position control. The only part in the control of the whole system that the amputee will make an effort to learn is to learn to use these two muscle groups, which are responsible for position control.

This direct and similar modularity relationship will be able to complete the rehabilitation process in a short time, as learning to modify muscle activation in a proportional way will enable fingers to be opened and closed in a proportional way. The ability to change its stiffness does not require the amputee to make a special effort for this function, sEMG signals measured from the remaining part of the individual are used, but the stiffness change of the fingers is carried out automatically.

The details of the design of the smart hand prosthesis, which will be easily learned by the user, has a natural control interface and has an electromechanical structure compatible with it, are listed below.

1. Design of Underactuated Smart Variable Stiffness Hand Prothesis Based on Shape Memory Material

Following the terminology proposed in [79] in hand prosthesis design, the performance criteria of a smart hand prosthesis proposed in this study will be evaluated in four separate groups as mandatory, optimal, primary and secondary requirements.

1.1. Design of Anthropomorphic Hand Prosthesis

Anthropomorphism (the association of human form or features with non-human objects/animals) is a mandatory design requirement for hand prostheses. The aesthetically pleasing natural appearance is suitable not only for the adaptation of prosthetic devices by the amputee, but also for interacting with common tools and environments designed for human use. The hand prosthesis in the present invention is an anthropomorphic hand prosthesis whose dimensions can be customized thanks to the availability of the production techniques used.

The part that will play a key role in the anthropomorphic hand prosthesis design developed by this invention is the fingers. Figure 1 shows the upper and side views and intersection of the anthropomorphic finger front design. The structure existing in the human finger is taken as an example in finger design. For example, the phalanxes, which are the three movable bone structures of the finger shown in Figures 4, 5, 6, constitute the main roof of the finger. As much force as necessary can be transferred to the object to be interacted in this way. The bone-type structures shown in Figure 1 with numbers 1, 2, and 3 are responsible for preventing uncontrolled extension. Thus, fingers are allowed to be opened at an angle suitable for human anthropomorphism. Compatible fingers that can easily adapt to objects in different geometries are also possible by including elastic joints (7,11) (elastic shape memory polymer materials) in the design. The soft finger surface (12), which is also anthropomorphic, is provided by integrating the polyurethane material into the lower finger surface (12) as shown in Figure 1.

The improvement of the preliminary design for the finger internal structure is the design of the thumb, four fingers and palms close to the human anthropomorphism [84] in the literature and the production of prototypes. Each cavity opened inside has an important function even though each phalanx in the finger has a very small volume. More free changes can be made in the interior design, in other words, it is possible to make improvements in the intricate design of the phalanxes with the use of 3D printers that can produce precision and in particular, it is possible for the finger design to meet the needs of the aforementioned design requirement.

1.2. Design of Dexterous Hand Prosthesis

Dexterity is an optimal performance requirement for a hand prosthesis design. A hand prosthesis should be able to grasp objects in different shapes (angled, spherical, cylindrical) and various properties (soft, brittle, smooth or irregular surfaces) without damage. The present invention relates to an underactuated hand prosthesis, which can mimic the opening/closing sequence of human fingers to accommodate a wide variety of geometries, and can adapt the stiffness of the prosthesis to the task/activity, thereby guaranteeing dexterity. Distal interphalangeal and proximal phalanxes are shown with numbers 1, 2 and 3, respectively, in Figure 2. The joints can be elastic joints, as well as the rigid joint using the shape memory alloy by using the shape memory polymer material.

Unlike previous studies, increasing the tendon width of the fingers of the present invention, which make the opening and closing movement of the hand based on the tendon (using stripetype tendons), as shown in Figures 2 with numbers 4, 5 and 6, reveals a unique underactuated design by driving each phalanx with its own tendon, but ultimately joining on the same reel with different paths. In the invention, the fingers in the aforementioned design perform the opening and closing movement with two de motors that can be placed in the palm of the hand in order to increase the lifting power of the fingers and to realize the sequential closing sequence as in the human-type hand. The said optimal performance criteria are provided with a design of the nature mentioned above. 1.3. Shortening Rehabilitation Process / Easy-to-Use Hand Prosthesis Design

The basic requirement for hand prosthesis is ease of use. First of all, the control of the hand prosthesis should be intuitive; the operation should take place naturally. In other words, the amputee should not be exposed to long-term rehabilitation and training processes in order to use the prosthesis.

• In this invention, the fact that the prosthesis can change its stiffness through shape memory materials and that the control interface can perform this function naturally, that is, without the need for the individual to show a special construction power, prevents the need for an additional rehabilitation process for this function.

• The use of an underactuated mechanism in this invention makes it considerably easier to control the device by the amputees. Because each finger does not need a separate controller for position and impedance control. The hand prosthesis in the present invention is a hand prosthesis that is fully compatible with the control interface, has an underactuated mechanism and has a mechanical design that allows both position and impedance to be changed.

• Thus, it can be easily adapted to the mechanical design with underactuated mechanism with its natural interface designed using the tele-impedance control sEMG signal, which will be mentioned in more detail in the second group. The compatibility of the control interface and the electromechanical design allows both position and impedance to be independently controlled.

The position change of the fingers is achieved by using two DC motors according to the needs of the improved design. Thus, hand prosthesis falls into the category of highly underactuated designs for the fingers with 14 degrees of freedom (since 12 degrees of freedom are considered as the sum of the degrees of freedom of index, middle, ring and pinky fingers, 2 degrees of freedom are considered as the sum of the degrees of freedom of thumbs or 9 degrees of freedom, middle, ring and pinky fingers, 5 degrees of freedom are considered as the sum of the degrees of freedom of thumbs and pinky fingers). The control unit will only be responsible for the motion control of the two motors in this case. 1.4. Design of Robust Hand Prosthesis

The secondary requirement of hand prosthesis is that it has a high level of robustness. The method followed in the present invention is to carry out a finger design with compliant and variable stiffness, so that said secondary requirement object is also achieved.

• The compatible finger property is achieved by including the elastic material (such as the shape memory polymer) to be used in the joint part in the finger design. The finger can easily adapt to an object of any geometry using only its elasticity feature, and there is no need to use a special control algorithm during this adaptation in this way.

• Materials that can change the stiffness of the fingers are used in order to be able to change the stiffness of the fingers depending on the environmental conditions, sudden reactions, and the hard, fragile and soft state of the object being held. i. Material Selection for Use on Fingers That Can Change Their Stiffness

SMPs and LMPAs have their own limitations and advantages as previously discussed. Therefore, the materials to be used in the hand prosthesis design within the scope of the invention are SMP, LMPA and SMA according to their adjustable elasticity and reaction rates. Each material is used as different versions for the hand prosthesis design proposed within the scope of the invention. ii. The Role of Placement of Shape Memory Materials in the Design of Variable Stiffness Finger

The shape memory materials can generally vary their stiffness between two values. A flexible, intermediate value and a finger design with a hard (less flexible) model come to the forefront, considering a variable stiffness finger model. The model in Figure 4 is presented as an example of the placement of shape memory materials. Here, SMP is placed on phalanxes parallel to each other. Thus, while the elastic material allows the angular change required for the opening and closing of the fingers, it shows resistance depending on the hardness level adjusted to the force applied when it changes its stiffness. The shape memory materials placed parallel to each other are modeled as two spring mechanisms connected parallel to each other in Figure 4. In this case, when the high stiffness level of the shape memory material is modelled with kl and the low stiffness level (i.e. rigidity) is modelled with k2 spring constant, the equivalent spring constant will be k= kl+ k2.

For example, the shape memory material with different stiffness level according to the temperature level will have the spring constant kl at temperature T1 and the spring constant k2 at temperature T2. Three different values of elasticity can be guaranteed for the level of elasticity of the finger in this case. The equivalent spring constant will be flexible= kl+ kl when both layers are brought to the temperature T1 in Figure 4a. In Figure 4b, it can be ensured that the first layer is brought to the temperature T2 and the material is brought to the spring constant k2, and the second layer is brought to the temperature T1 and the material is brought to the spring constant kl. We can reach a level of stiffness between the upper and lower values, which we can call the middle level in this case. In other words, the equivalent spring constant is modeled as kmiddle= kl+ k2. In Figure 4c, if both layers are brought to the temperature T2, the spring constants of the materials are brought to the k2 value. In this case, the finger is brought to the desired rigid position, that is, the most rigid form. The equivalent spring constant is modeled as krigit = k2+ k2.

The layout of the materials in the finger design that provides stiffness change is not limited to the model presented in Figure 4.

2. Design of Natural Control Interface Based on Surface Electromyography Signal

The second of the complementary arms of the invention is the design of the natural control interface, which enables amputee individuals to considerably shorten the learning/adaptation process of the hand prosthesis compared to commercial hand prostheses. The control architecture consists of two main parts. The first three methods mentioned below mention each stage in the first part of the control architecture. The fourth and fifth methods deal with the control cycle mentioned in the second part.

2.1. Identification of Relevant Muscle Groups to be involved in Machine Interaction The hand prosthesis in the present invention utilizes surface electromyography signals that enable human-machine communication. People who have suffered from congenital or subsequent loss of hands, arms or legs have live muscles and nerves that will enable them to control their missing limbs. It is possible to control the mechanical prosthesis that can replace the missing limb by using the electrical potential produced by the existing muscle cells. The first step followed in the control architecture design of the device is to first select the muscle groups in which the signals will be collected in this invention. Studies conducted during the study can be carried out on healthy people, considering the conditions of an amputee. 4 muscle groups should be selected considering the conditions of a hand or transradial amputee. For example, considering the condition of the transradial amputees, muscle groups can be selected as biceps-triceps, trapezius-pectoralis major muscles, and considering the conditions of an individual with only hand amputation, biceps-triceps and flexor digitorum profundis- extensor digitorum communis muscle groups can be selected. sEMG signals measured from bicep-triceps muscles play a role in impedance control because they are an antagonistic muscle pair that remains healthy in both amputations. The other muscles in question take part in the position control of the fingers of the hand prosthesis.

2.2. Filtering and Conditioning of the sEMG Signals through Signal Processing Methods

The second step followed is to perform sEMG signal measurements through the EMG device and electrodes from the muscle groups, and to remove the measured signals from the noises in real time. The signals outside the frequency band in the range of 20 and 500Hz, that is, the noises, must be cleared in order to benefit from the measured sEMG signal. Band-pass filter operating in the (20-500) Hz band is used for this reason. The next stage continues with the conditioning process of the sEMG signal. Here, the amplitude value of the signal is made ready for the stage of obtaining the control references by using signal processing methods such as moving average and envelope detector.

2.3. Calculation of Position and Elasticity References from sEMG Signal

The third step followed is to estimate the impedance and position control references required to ensure the stiffness and position control of the hand prosthesis. The estimation of the stiffness reference is carried out by using the relationship established in the literature [80] between the index of muscle co-contraction and the stiffness change of the human joint. The person undergoes a series of experiments before using the prosthesis. The resistance of the antagonistic muscle groups to the different load values applied is calculated and the level of stiffness of the muscle is estimated.

The amplitude information of the normalized EMG signal is used after the signal processing methods in the calculation of the position reference. A linear relationship is established between the angular change of the fingers and the normalized EMG signal.

2.4. Matching the Natural Control Interface with the Electromechanical Hand Prosthesis System

The fourth step is to address the stage of matching the position and impedance references obtained in the first part of the control architecture with the technical features of the electromechanical hand prosthesis. In other words, the maximum and minimum impedance levels calculated from the sEMG signal are matched to the maximum and minimum stiffness level that the electromechanical hand prosthesis can provide, that is, an amount of current corresponding to the three-class stiffness levels is sent to the heaters. Similarly, the maximum and minimum position levels calculated from the sEMG signal are matched to the maximum and minimum levels of the finger angles of the electromechanical hand prosthesis.

2.5. Control of DC Motors and Smart Materials According to the Given Reference Values

The fifth method followed is provided by the control of the DC motors responsible for the position of the fingers and the Joule heaters (13) responsible for the shape reformation of the shape memory materials. The control of the motors will be carried out with a simple proportional-integral-differential controls and closed-loop control. The heaters are set to the desired temperature level to ensure that the material has the stiffness at that temperature.

Finally, the control loop is closed through visual feedback given by human.

REFERENCES [1] T. Takaki and T. Omata, “High-performance anthropomorphic robot hand with grasping- force-magnification mechanism,” IEEE/ASME Transactions on Mechatronics, vol. 16, no. 3, pp. 583-591, 2011.

[2] W. Chen, C. Xiong, and S. Yue, “Mechanical implementation of kinematic synergy for continual grasping generation of anthropomorphic hand,” IEEE/ASME Transactions on Mechatronics, vol. 20, no. 3, pp. 1249-1263, 2015.

[3] S. A. Dailey, T. E. Wiste, T. J. Withrow, and M. Goldfarb, “Design of a multifunctional anthropomorphic prosthetic hand with extrinsic actuation,” IEEE/ASME Transactions on Mechatronics, vol. 14, no. 6, pp. 699-706, 2009.

[4] Ossur Inc., https://www.ossur.com/en-us/prosthetics/arms/i-limb-ultra Last accessed: 03.12.2019

[5] Ottobock, “Electric greifer”, http://www.ottobock.com.tr/en/prosthetics/products-from-a- to-z/system-electric-greifer-dmc-plus/ Last accessed: 03.12.2019

[6] A. Bicchi, “Hands for dexterous manipulation and robust grasping: a difficult road toward simplicity,” IEEE Transactions on Robotics and Automation, vol. 16, no. 6, pp. 652-662, 2000.

[7] B. Siciliano and O. Khatib, Eds., Springer Handbook of Robotics. Springer, 2008.

[8] T. Feix, R. Pawlik, H. Schmiedmayer, J. Romero, and D. Kragic, “A comprehensive grasp taxonomy,” 2009. http://grasp.xief.net Last accessed: 03.12.2019

[9] T. Feix, J. Romero, H. B. Schmiedmayer, A. M. Dollar, and D. Kragic, “The grasp taxonomy of human grasp types,” IEEE Transactions on Human-Machine Systems, vol. 46, no. 1, pp. 66-77, 2016.

[10] A. Bicchi, M. Gabiccini, and M. Santello, “Modelling natural and artificial hands with synergies,” Philosophical Trans, of Royal Soc. of London B: Biological Sciences, vol. 366, no. 1581, pp. 3153-3161, 2011.

[11] M. Catalano, G. Grioli, A. Serio, E. Farnioli, C. Piazza, and A. Bicchi, “Adaptive synergies for a humanoid robot hand,” in IEEE International Conference on Humanoid Robots, 2012, pp. 7-14.

[12] C. Abul-Haj and N. Hogan, “An emulator system for developing improved elbowprosthesis designs,” IEEE Transactions on Biomedical Engineering, vol. BME-34, no. 9, pp. 724-737, 1987.

[13] - , “Functional assessment of control systems for cybernetic elbow prostheses. I.

Description of the technique.” IEEE Trans. Biomed. Eng., vol. 37, no. 11, pp. 1025-1036, 1990. [14] - , “Functional assessment of control systems for cybernetic elbow prostheses. II. 5

Application of the technique.” IEEE Trans. Biomed. Eng., vol. 37, no. 11, pp. 1037-1047, 1990.

[15] A. Blank, A. Okamura, and L. Whitcomb, “Task-dependent impedance improves user performance with a virtual prosthetic arm,” in IEEE Int. Conference on Robotics and Automation, 2011, pp. 2235-2242. 10

[16] - , “User comprehension of task performance with varying impedance in a virtual prosthetic arm: A pilot study,” IEEE International Conference on Biomedical Robotics and Biomechatronics, pp. 500-507, 2012.

[17] - , “Task-dependent impedance and implications for upper-limb prosthesis control,”

International Journal of Robotics Research, vol. 3, no. 6, pp. 827-846, 2013.

[18] A.Firouzeh, S. Salehian, A. Billard, J. Paik, “An Under Actuated Robotic Arm with Adjustable Stiffness Shape Memory Polymer Joints”, IEEE International Conference on Robotics and Automation, 2015.

[19] J. Shintake, B. Schubert, S, Rosset, H. Shea, D. Floreano, “Variable stiffness actuator for soft robotics using dielectric elastomer and low-melting-point alloy”, IEEE/RSJ International Conference on Intelligent Robots and Systems, 2015.

[20] W. Shan, T. Lu and C. Majidi, “Soft-matter composites with electrically tunable elastic rigidity”, Smart Materials and Structures, Vol 22, No 8, 2013

[21] B. E. Schubert and D. Floreano, “Variable stiffness material based on rigid low meltingpoint-alloy microstructures embedded in soft poly(dimethylsiloxane) (PDMS)”, RCV Adv, 25 2013

[22] F. Touvet, N. Daoud, J. Gazeau, S. Zeghloul, M. A. Maier, and S. Eskiizmirliler, “A biomimetic reach and grasp approach for mechanical hands,” Robot. Auton. Syst., vol. 60, no. 3, pp. 473-486, 2012.

[23] G. Li and T. Kuiken, “EMG pattern recognition control of multifunctional prostheses by transradial amputees,” in Annual International Conference of the IEEE Engineering in Medicine and Biology Society, 2009, pp. 6914-6917.

[24] C. Castellini, P. van der Smagt, G. Sandini, and G. Hirzinger, “Surface EMG for force control of mechanical hands,” in IEEE International Conference on Robotics and Automation, 2008, pp. 725-730.

[25] J.-U. Chu, I. Moon, Y.-J. Lee, S.-K. Kim, and M. seong Mun, “A supervised featureprojection-based real-time EMG pattern recognition for multifunction myoelectric hand control,” IEEE/ASME Transactions on Mechatronics, vol. 12, no. 3, pp. 282-290, 2007. [26] L. McFarland, S. Hubbard Winkler, A. Heinemann, M. Jones, and A. Esquenazi, “Unilateral upper-limb loss: satisfaction and prosthetic device use in veterans and service members from Vietnam and OIF/OEF conflicts,” J Rehabil Res Dev, vol. 47, no. 4, pp. 299- 316, 2010.

[27] E. Biddiss and T. Chau, “Upper limb prosthesis use and abandonment: a survey of the last 25 years,” IEEE Robotics Automation Magazine, vol. 31, no. 3, pp. 236-257, 2007.

[28] M. Grebenstein, M. Chalon, G. Hirzinger, and R. Siegwart, “Antagonistically driven finger design for the anthropomorphic DLR hand arm system,” in IEEE Int. Conf, on Humanoid Robots, 2010, pp. 609-616.

[29] S. Wolf, O. Eiberger, and G. Hirzinger, “The DLR FSJ: Energy based design of a variable stiffness joint,” IEEE International Conference on Robotics and Automation, pp. 5082-5089, 2011.

[30] Shadow Robot Company Ltd., 2016. [Online], Available: https: //www. shadowrobot.com/products/dexterous-hand/

[31] D. Prattichizzo, M. Malvezzi, M. Gabiccini, and A. Bicchi, “On the manipulability ellipsoids of underactuated robotic hands with compliance,” Robotics and Autonomous Systems, vol. 60, no. 3, pp. 337-346, 2012.

[32] L. Birglen and C. Gosselin, “On the force capability of underactuated fingers,” in IEEE International Conference on Robotics and Automation, vol. 1, 2003, pp. 1139-31 145.

[33] T. Laliberte, L. Birglen, and C. Gosselin, “Underactuation in robotic grasping

[34] Y. Kamikawa and T. Maeno, “Underactuated five-finger prosthetic hand inspired by grasping force distribution of humans,” in IEEE International Conference on Intelligent Robots and Systems, 2008, pp. 717-722.

[35] C. Gosselin, F. Pelletier, and T. Laliberte, “An anthropomorphic underactuated robotic hand with 15 dofs and a single actuator,” in IEEE Int. Conference on Robotics and Automation, 2008, pp. 749-754.

[36] R. Ozawa, K. Hashirii, and H. Kobayashi, “Design and control of underactuated tendon- driven mechanisms,” in IEEE International Conference on Robotics and Automation, 2009, pp. 1522-1527.

[37] K. Xu, H. Liu, Z. Liu, Y. Du, and X. Zhu, “A single-actuator prosthetic hand using a continuum differential mechanism,” in IEEE International Conference on Robotics and Automation, 2015, pp. 6457-6462.

[38] B. Massa, S. Roccella, M. Carrozza, and P. Dario, “Design and development of an underactuated prosthetic hand,” in IEEE International Conference on Robotics and Automation, vol. 4, 2002, pp. 3374-3379.

[39] A. Dollar and R. Howe, “A robust compliant grasper via shape deposition manufacturing,” IEEE/ASME Transactions on Mechatronics, vol. 11, no. 2, pp. 154-161, 2006.

[40] A. Dollar, L. Jentoft, J. Gao, and R. Howe, “Contact sensing and grasping performance of compliant hands,” Autonomous Robots, vol. 28, no. 1, pp. 65-75, 2010.

[41] R. Ma, L. Odhner, and A. Dollar, “A modular, open-source 3D printed underactuated hand,” in IEEE Conf, on Robotics and Automation, 2013.

[42] R. Ma, L. Odhner, and A. Dollar, “A modular, open-source 3D printed underactuated hand,” in IEEE Conf, on Robotics and Automation, 2013.

[43] M. Catalano, G. Grioli, E. Farnioli, A. Serio, C. Piazza, and A. Bicchi,“ Adaptive synergies for the design and control of the Pisa/IIT SoftHand,” Int. Journal of Robotics Research, vol. 33, no. 5, pp. 768-782, 2014.

[44] M. Marino, S. Pattni, M. Greenberg, A. Miller, E. Hocker, S. Ritter, and K. Mehta, “Access to prosthetic devices in developing countries: Pathways and challenges,” in IEEE Global Humanitarian Technology Conference (GHTC), Oct 2015, pp. 45-51.

[45] K. Ziegler-Graham, E. MacKenzie, P. Ephraim, T. Travison, and R. Brookmeyer, “Estimating the prevalence of limb loss in the united states: 2005 to 2050,” Archives of Physical Medicine and Rehabilitation, vol. 89, no. 3, pp. 422-429, 2007.

[46] M. Lusardi, M. Jorge, and C. Nielsen, Orthotics and Prosthetics in Rehabilitation, 3rd ed. Saunders Elsevier, 2013.

[47] S. Millstein, D. Bain, and G. Hunter, “A review of employment patterns of industrial amputees-F actors influencing rehabilitation,” Prosthet Orthot Int, vol. 9, no. 2, pp. 69-78, 1985.

[48] E. Biddiss and T. Chau, “Upper limb prosthesis use and abandonment: a survey of the last 25 years,” Prosthet Orthot Int., vol. 31, no. 3, pp. 236-257, Sep 2007.

[49] L. McFarland, S. Hubbard Winkler, W. A., J. M., and A. Esquenazi, “Unilateral upperlimb loss: Satisfaction and prosthetic-device use in veterans and service members from Vietnam and oif/oef conflicts,” J Rehabil Res Dev., vol. 47, no. 4, pp. 299-316, 2010.

[50] N. Wang, K. Lao, and X. Zhang, “Design and myoelectric control of an anthropomorphic prosthetic hand,” Journal of Bionic

[51] Engineering, vol. 14, no. 1, pp. 47-59, 2017.

[52] F. Riillo, L. Quitadamo, F. Cavrini, E. Gruppioni, C. Pinto, N. C. Pasto, L. Sbernini, L. Albero, and G. Saggio, “Optimization of emg-based hand gesture recognition: Supervised vs. unsupervised data preprocessing on healthy subjects and transradial amputees,” Biomedical Signal Processing and Control, vol. 14, pp. 117-125, 2014.

[53] Ossur lnc., 2019. [Online], Available: https://www.ossur.com/.

[54] bebionic, 2019. [Online], Available: https://www.ottobockus.com/prosthetics/upper- limb-prosthetics/solution-overview/bebionic-hand/

[55] Ottobock, 2019. [Online], Available: https://www.ottobock.com/en/

[56] Motion Control Inc., 2019. [Online], Available: http://www.utaharm.com/motion- control -el ectri c-hand-terminal-devi ce . php

[57] M. Kurzynski, M. Krysmann, P. Trajdos, and A. Wolczowski, “Multi classifier system with hybrid learning applied to the control of bioprosthetic hand,” Computers in Biology and Medicine, vol. 69, pp. 286-297, 2016.

[58] W. Guo, X. Sheng, H. Liu, and X. Zhu, “Toward an enhanced human- machine interface for upper-limb prosthesis control with combined EMG and NIRS signals,” IEEE Transactions on Human-Machine Systems, vol. PP, no. 99, pp. 1-12, 2017.

[59] I. Kyranou, A. Krasoulis, M. S. Erden, K. Nazarpour, and S. Vijayakumar, “Real-time classification of multi-modal sensory data for prosthetic hand control,” in 6th IEEE International Conference on Biomedical Robotics and Biomechatronics (BioRob), June 2016, pp. 536-541.

[60] S. Herle, “Design of a reference signal generator for an upper limb prosthesis myoelectric controller,” in IEEE International Conference on Automation, Quality and Testing, Robotics (AQTR), May 2016, pp. 1-6.

[61] D. Atkins, D. Heard, and W. Donovan, “Epidemiologic overview of individuals with upper limb loss and their reported research priorities”, Journal of Prosthetics and Orthotics, vol. 8, no. 1, pp. 2-11, 1996.

[62] E. J. Perreault, R. Kirsch, and P. Crago, “Multijoint dynamics and postural stability of the human arm,” Experimental Brain Research, vol. 157, no. 4, pp. 507-517, 2004.

[63] F. Popescu, J. Hidler, and W. Rymer, “Elbow impedance during goal directed movements,” Experimental Brain Research, vol. 152, no. 1, pp. 17-28, 2003.

[64] D. Franklin, E. Burdet, R. Osu, M. Kawato, and T. Milner, “Functional significance of stiffness in adaptation of multijoint arm movements to stable and unstable dynamics,” Experimental Brain Research, vol. 151, no. 2, pp. 145-157, 2003.

[65] D. Franklin and T. Milner, “Adaptive control of stiffness to stabilize hand position with large loads,” Experimental Brain Research, vol. 152, no. 2, pp. 211-220, 2003.

[66] D. Franklin, U. So, M. Kawato, and T. Milner, “Impedance control balances stability with metabolically costly muscle activation,” Journal of Neurophysiology, vol. 92, no. 5, pp. 3097-3105, 2004.

[67] D. Franklin, M. Kawato, U. So, and T. Milner, “Interacting with our environment: Impedance control balances stability and metabolic cost,” in IEEE Engineering in Medicine and Biology Society, vol. 2, Sept 2003, pp. 1440-1443.

[68] N. Hogan, “Skeletal muscle impedance in the control of motor actions,” Journal of Mechanics in Medicine and Biology, vol. 02, no. 03 n04, pp. 359-373, 2002.

[69] C. Abul-Haj and N. Hogan, “Functional assessment of control systems for cybernetic elbow prostheses. i. description of the technique.” IEEE Trans. Biomed. Eng., vol. 37, no. 11, pp. 1025-1036, 1990.

[70] - , “Functional assessment of control systems for cybernetic elbow prostheses. ii. application of the technique.” IEEE Trans. Biomed. Eng., vol. 37, no. 11, pp. 1037-1047, 1990.

[71] S. Rao, R. Carloni, and S. Stramigioli, “Stiffness and position control of a prosthetic wrist by means of an EMG interface,” IEEE EMBS, pp. 495-498, 2010.

[72] A. Blank, A. Okamura, and L. Whitcomb, “Task-dependent impedance improves user performance with a virtual prosthetic arm,” in IEEE International Conference on Robotics and Automation (ICRA), May 2011, pp. 2235-2242.

[73] - , “User comprehension of task performance with varying impedance in a virtual prosthetic arm: A pilot study,” 4th IEEE RAS EMBS International Conference on Biomedical Robotics and Biomechatronics, pp. 500-507, 2012.

[74] - , “Task-dependent impedance and implications for upper-limb prosthesis control,”

International Journal of Robotics Research, vol. 3, no. 6, pp. 827-846, 2013.

[75] E. Hocaoglu and V. Patoglu, “Tele-impedance control of a variable stiffness prosthetic hand,” in IEEE International Conference on Robotics and Biomimetics (ROBIO), 2012, pp. 1576-1582.

[76 S. A. Dailey, T. E. Wiste, T. J. Withrow, and M. Goldfarb, “Design of a multifunctional anthropomorphic prosthetic hand with extrinsic actuation,” IEEE/ASME Transactions on Mechatronics, vol. 14, no. 6,pp. 699-706, Dec 2009.

[77] T. Lenzi, J. Lipsey, and J. W. Sensinger, “The RIC arm - a small anthropomorphic transhumeral prosthesis,” IEEE/ASME Transactions on Mechatronics, vol. 21, no. 6, pp. 2660-2671, Dec 2016.

[78] D. A. Bennett, J. E. Mitchell, D. Truex, and M. Goldfarb, “Design of a myoelectric transhumeral prosthesis,” IEEE/ASME Transactions on Mechatronics, vol. 21, no. 4, pp. 1868-1879, Aug 2016.

[79] J. Merlet, Parallel Robots, 2nd ed. Springer, 2006

[80] R. Osu, D. W. Franklin, H. Kato, H. Gomi, K. Domen, T. Yoshioka, and M. Kawato, “Short- and long-term changes in joint co-contraction associated with motor learning as revealed from surface EMG,” J Neurophysiol, vol. 88, pp. 991-1004, 2002.

[81] Wolf Schweitzer: Technical Below Elbow Amputee Issues - Otto Bock Michelangelo hand; published March 27, 2009, 00:32; URL: https://www.swisswuff.ch/tech/7pM45.

[82] Wolf Schweitzer: Technical Below Elbow Amputee Issues - What happened to the Otto Bock Michelangelo hand?; published May 12, 2010, 21 : 14; URL: https://www.swisswuff.ch/tech/7pM07.

[83] J. T. Belter, J. L. Segil, A. M. Dollar, R. F. Weir, "Mechanical design and performance specifications of anthropomorphic prosthetic hands: A review", vol. 50, no. 5, pp. 599 — 618, 2013.