Bionic Limbs


Prosthetics are artificial device extensions that replace missing body parts which help the individual regain some function and normalcy of their missing part. The development and study of prosthesis is considered as part of the field of biomechatronics, which is the science of using mechanical devices with human muscle, skeleton, and nervous systems to assist and enhance with motor control (Freudenrich, 2010). Orthotics are different than prosthetics in that they only provide support and corrective assistance for the limb or torso. The research of prosthesis has progress significantly over the years and has incorporated many new technologies and scientific developments.
Dr. Todd Kuiken, M.D., PhD (McCormick, 2007)

Today, the most natural and technically advanced prosthesis uses targeted reinnervation with a mechanical device to produce precise movements through thought (Kuiken, 2006). Individuals who undergo the targeted reinneravtion procedure are able to control the robotic prosthetic through thinking about moving their phantom limb. Surprisingly, this procedure has also given individuals a sensory component, being able to feel touch and pressure in their phantom limb. Developed and continually perfected by Dr. Todd Kuiken and his teams at DEKA Research and Development and the Rehabilitation Institute of Chicago, this technology has reworked how that world thinks about individuals with disabilities. Those who have suffered from limb amputation have something to look forward to with this new technology recently switching from the experimental stage to the consumer stage . The Pentagon is also now overseeing developments in this area and to assist those injured in the War (DEKA). It is unsure where this technology will take us and how far it will be able to be pushed for further improvements in functionality and sense in phantom limbs by using robots.


Our Central Nervous System (CNS), which includes our brain and spinal cord process and integrate both efferent and afferent information to allow for understanding of our surroundings and produce movement. Information is passed along ascending and descending pathways along the CNS and between structure for interpretation and recognition. In order to gain a full understand of the procedure, workings and inter-workings of this "Bionic Limb," it is necessary to have a general knowledge of muscle contraction, the tracts and integrations that are required for us to feel and produce movements. Voluntary muscle contraction and coordination to produce bodily movements are attributed to the Descending Motor Pathways, whereas our interpretation of cutaneous sensory information is attributed to several Ascending Pathways. This information will later help in the understanding of targeted reinnervation and electromyogram signaling.


All muscles contract in a similar way. An action potential passes electrical signals down the alpha-motor neuron which innervates a group of extrafusal muscle fibers. When the action potential triggers the end-plate potential of the muscle cell, an electric depolarization impulse is carried down and across the transverse tubules (T-tubules). This causes the terminal cistern to release calcium. The calcium ions that are released move throughout the cytosol of the cell and bind to troponin on the actin, causing a structural change. This structural change exposes tropomyosin binding sites on the actin. Heads of the myosin fibers cock and bind to these sites, pulling and causing a longitudinal force, which brings the Z-disks closer together. ATP is required to release and re-cock the myosin heads for another contraction. (Kendell, 2000) This process can be seen in this video for further understanding.


The Descending Motor Pathways take electrical impulses from the brain down efferent fibers to various muscle to perform movements. Dependent on the level of coordination and muscles involved, several pathways can be involved. Information is passes through and between the motor and premotor areas, basal ganglia and cerebellum to the brain stem and the spinal cord to the necessary level for the muscles that need to contract or relax. The motor and premotor areas allow for the conscious and subconscious planning and execution of movements. The basal ganglia facilitates movements, while the cerebellum is the major contributing factor in coordination and production. The cerebellum is also broken up into three functional regions. The cerebrocerebellum, consisting of the lateral hemispheres is responsible for motor planing of the extremities. The intermediate hemispheres and the vermis (middle) make up the spinocerebellem, which is responsible for distal and proximal limb and trunk coordination. The third, the vestibulocerebellum, consists of the floculonodular lobe and is responsible for balance and vestibular reflexes. The descending pathways allow for motor planning of extremities, limb and trunk coordination and movement, positioning and movement of head, eyes and neck, posture and gait control. Each descending tract has a particular origin, function, purpose and level of termination. These tracts include the Lateral Corticospinal tract, Rubrospinal tract, Anterior Corticospinal tract, Medial & Lateral Vestibulospinal tracts, Pontine & Medullary Reticulospinal tract and the Tectospinal tract. (Kendell, 2000) Further information on the Descending Motor Pathways and each particular tract can be seen on this website.


Our interpretation and understanding of the world is brought in through our sensory receptors and processed by the ascending pathways. For the purpose of this situation, we are interested in somatic sense receptors and their ascending pathways to the brain. Cutaneous receptors are found in the skin and include Meissner's corpuscles, Merkel discs, Pacinian corpuscles, Ruffini endings, and Free Nerve Endings. These receptors transduce information regarding stretching, vibration, pressure and texture to the brain. Thermal receptors are also found on the skin and convey information about the temperature; cold, cool, warm, and heat. Nociceptors are receptors on the skin that interpret extremes, such as sharp pain and pricking and burning and freezing sensations. Generally, two systems of pathways will carry the information from these receptors to the brain. The Dorsal Column Medial Lemniscal System bring information regarding touch and proprioception. The Anterolateral System, which includes the Spinothalamic and Spinoreticular tracts, carries the information regarding pain and temperature. Sensory information is continually processed as it ascends along the afferent fibers. The sensory information is projected to the thalamus where it is integrated with other information and interpreted for cognition and importance. From here the necessary information is then projected to the Somatosensory Cortex where it is further processed and passes along to other high cortical areas. (Kendell, 2000). Details and further, in depth explanation of the particular receptors, systems and tracts can be read in this online text.


Figure 1 (Stubblefield et al., 2009)

Targeted reinnervation is a method developed by Dr. Todd Kuiken which allows natural for control of a mechanical prosthetic through thought while also providing some sensory feedback of the amputated limb. This process requires a surgical procedure where a spare/target muscle, no longer biomechanically functional due to the missing part, is deactivated and reinnervated with the residual nerves from the amputated limb. These reinnervated nerves are salvaged, grow and develop strength, mass and motor units in their new muscle. Electromyogram (EMG) signals of the target muscle represent motor commands of the missing limb, causing new contraction patterns. These contractions cause a myoelectric signal that can be interpreted and converted by electrodes and the mechanical prosthetic to produce robotic movements. (Kuiken et al., 2004, 2006) This process can also be applied to the afferent sensory nerves, to allow for the sensation of touch and feeling of the phantom limb in the new target area (Kuiken et al. 2007). The technology is very new and is advancing, developing and progress each day as we learn more about what targeted reinnervation can provide for us and how to best engineer new robotic arms to accompany. Figure 1 shows a simple diagram of the reinnervated nerves passing information to the microprocessor of bionic limb.


Targeted Muscle Reinnervation (TMR) requires a surgical procedure to transfer the residual nerves to the target muscle, in order to provide a new surface for EMG signals that can control robotic prothestic devices. This procedure has successfully been performed on individuals with transhumeral and shoulder-disarticulation amputations (Kuiken et al. 2004, 2006, 2007). In these cases, the pectoral muscles were chosen as the target muscles because of their proximity to the limb and were non-functional due to the detachment of the arm (Kuiken, 2005). Multiple nerves are required for transfer for more
Figure 2 (Kuiken et al., 2009)
precise movements, however only the four main arm nerves have been transferred. After the pectoral muscles are denervated, their proximal ends are ligated to revent them from developing back into the muscle. Then the muscle is ready for the arm nerves to be transferred. For the should-disarticulation amputee, the musculocutaneous nerve is transferred to the clavicular head, the median nerve is moved to the upper sternal, and the radial nerve is transferred to the lower sternal head, all of the pectoralis major muscle (Kuiken, 2006). The fourth nerve, ulnar nerve, is connected to the pectoralis minor muscle. This muscle however, is translocated to the lateral chest wall, under the pectoralis major in order to not let EMG signals interfere (Kuiken, 2005). These nerves were integrated directly into the nerve fascicles of the pectoral muscles. Also all subcutaneous fat was removed so that the electrodes could gather optimal EMG signals (Kuiken, 2006) Overall however, the surgical procedures differ for each individual patient, dependent on what is available and viable after the amputation.

In Figure 2, we can see the (A) Normal muscular anatomy as compared to an individual who have undergone the surgery (B), for both the shoulder disartculation and transhumeral amputation. This figure shows the locations of the four main nerves in both cases. This is the typical and best case diagrams for each situation, shoulder and transhumeral amputation.

Once surgery is complete, there is a waiting time of five to six months for the nerves to regrow and develop in their new location (RIC, 2010). According to Kuiken (2006) however, evidence of a successful surgery can be seen at around three months. This is when the patient first had a twitch in their pectoral muscle just below the clavicle when he attempted to bend his phantom elbow. Within the next two months after that, he was able to contract four different regions of his pectoralis major by attempting different movements with his phantom limb; opening and closing the hand, elbow flexion and extension, and wrist supination and pronation.

After the healing and nerve regrowth has sufficiently taken place, training is necessary to strengthen and develop the muscles with their new nerve connections. Patients need to practice moving their phantom limb to produce contractions. The pectoral muscles are generally weak and degenerated due to the surgery and the non-use due to the loss of the limb. The training sessions in physical therapy and occupational therapy, will allow for greater strength of contraction and most precise contractions. This will be helpful to created better myoelectric signals for the EMG of the electrodes and prosthetic to interpret.


Targeted sensory reinnervation (TSR) is the surgical transfer of nerves for the purpose of reinnervating sensory feedback. Since the skin near and around the targeted muscle was denervates, the afferent nerve fibers are allowed to reinnervate the skin. It has the potential of providing sensory feedback to help the patient judge the amount of force needed to be exerted. (Kuiken, 2007)

Accidentally, this TSR was discovered in the targeted reinnervation procedures first patient, but then was purposefully incorporated in later surgeries to examine its
Figure 3 (Kuiken, 2004)
potential. The first patient had no nerve transfers that were aimed at sensory feedback. It was during a routine appointment that he said that he felt as though he was being touched on his phantom picky when touched in a certain area of his chest (Kuiken, 2004). Because the area had been removed of all subcutaneous fat, the chest skin had been denervated. This allowed for the afferent nerve fibers to regrow through the target muscles, pectoral muscles, to reinnervate the skin. Thus, the afferent sensory nerves were able to reinnervate the skin without any intended surgical procedures (Kuiken, 2004). Figure 3 displays a diagram mapping of this first patient's sensory feedback areas. Each circled area corresponds to an area of the phantom limb that can be felt. Thus in the next TMR surgery, researchers also integrated two major sensory nerves, in addition to the four nerves discussed above, into the procedure. The supraclavicular sensory nerve was cut and then connected to the ulnar nerve while the intercostobrachial cutaneous nerve was cut and then connected to the median nerve (Kuiken, 2007).

This addition transfer of sensory nerves provided significant input. The second patient experienced all sensation in all digits and areas of the missing limbs. She could also perceive all modalities of cutaneous sensation including vibration, light touch, temperature, stretching and pain. However, pressure felt more like a tingling feeling. She had a range of the sensations that where considered normal, however temperature was a little more narrowed. Still all sensations the patient experienced on her chest, were felt in her phantom arm and hand. (Kuiken, 2007) Thus, the procedure was successful and TSR is now incorporated into each TMR surgery to provide the individual with actual sensory feedback of their missing limb. As of right now, this procedure does not have any significance towards the use of the prosthetic.


Electrodes are uses to interpret the myoelectric signal to produce the movements of the prosthetic. Typically a 127-channel electrode array is used to record the signals of the
Figure 4 (RIC, 2010)
contractions during both training and testing. These electrodes are highly concentrated in the target area, pectoral muscle but are also
Figure 5 (Neurophilosophy, Word Press, 2006)
placed at the deltoid, latissimus dorsi, supraspinatus, and all three levers of the trapezius. Each electrode is placed 15mm from the next. This allows for the most natural interpretation of the desired movement. (Zhou, 2005) Other electrical noise is decreased through finding and setting the correct threshold for interpreting the muscles of interest.

In Figures 4 and 5, you can see the electrode array on two individuals who have undergone the TMR surgery and are now working with their prosthetic. Figure 4 shows the first TMR patient, Jesse Sullivan as he examines the array before his prosthetic is place on. Figure 5 shows the second TMR patient, Claudia Mitchell and her array of electrodes. Notice how the distribution of electrodes is different for the two individuals. This is due to the variabilities and uniqueness that exist with each individual procedure. Because Claudia is a female with some portion of her shoulder, the TMR surgery was more focused on the upper pectoral and deltoid muscle to save her breast and still allow for the same benefits (Kuiken, 2007)

Each specific prosthetic is custom-built to correspond with the unique features of each TMR surgery and the individual. Several fitting are required throughout the training process to make sure the fit is perfect (RIC, 2010). Figure 6 shows an example of the motorized prosthetic detached from the body and its electrodes. The strongest EMG signals are chosen for each nerve that was transferred to have the most influence of the desired
Figure 6 (Neurophilosophy, Word Press, 2006)
movement for the prosthetic to interpret. The musculocutaneous nerve is responsible for elbow flexion and extension. The median nerve is responsible for digit flexion and extension, closing and opening the hand. And the radial nerve was responsible for wrist rotation and flexion. (Kuiken, 2005) Each contraction of the target muscle allows for the electrode to pick up the myoelectric signal and pass it to the robotic arm to perform the correct action. Thus when the TMR patient thinks bend elbow, the bionic arm will bend at the elbow. All major motions are accounted for, including individual fingers for precise dexterity.

In 2009, Dr. Kuiken and his team investigated the real-time myoelectric control of the bionic arm. They tested the TMR patients against control participants who did not have an amputation. The TMR patients were able to perform ten different elbow, wrist and hand motions with the prosthetic arm. The completion times for the movements of those with the prosthetic were slower than that of the control group however, the mean elbow time was only 0.06 seconds longer while the wrist was only 0.21 seconds longer. All motions had similar findings. These results suggest that the reinnervated muscles can produce sufficient EMG information for real-time control of the prosthetic limb. (Kuiken, 2009) Further progression and modification of both the surgical procedure and the mechanical motorized arm, as well as training and practice will allow for increased precision, accuracy, dexterity and speed.


The general risks of the surgical procedure include permanent paralysis of the target muscle, recurrence of phantom limb pain and development of pain nerve neuromas (Kuiken, 2007). Other risks that are usually associated with surgery also apply. Also some nerve damage can not be determined until surgery has begun, so the nerves could be too damaged for successful transfer. It is also uncertain how each nerve will take to their new muscle. In the case of Jesse Sullivan, the first patient, his ulnar nerve did not successfully take to the pectoralis minor muscle (Kuiken, 2006). It is also uncertain if the transferred nerve will remain and survive permanently. The technology is still so new that there have been no longitudinal testing.

In order to be a candidate for the procedure, there are several standard which the individual must meet. This procedure is limited for only certain individuals since it has recently begun its commercial stage. The amputation of the arm, above the elbow or at the shoulder, must be with the last ten years, preferably within the last five years. Candidates must be at least 14 years of age and of a certain weight. Also, individuals must be amputees with good nerves still attached. Those born with a missing limb or portion of the limb or have severe nerve damage can not be candidates for the procedure. (RIC, 2010)


The future progression of the targeted reinnervation and motorized prosthetics is seemingly endless, including a potential reinnervation for lower limb amputees. Many
Jesse Sullivan and Dr. Kuiken (Rehab Management, 2006)
considerations have involved the splitting of nerves during surgery to allow for more independent signals that will provide for more precise prosthetic control with increased degrees of freedom. There has also been discussions regarding implantable electrodes for more localized, clear and specific readings of the myoelectric signals for the EMG. Also, there is work being done with TSR. The hope is to be able to allow the prosthetic to pass the sensory information to the targeted skin or the skin to the prosthetic. This will allow the individual to be able to feel with their prosthetic motorized arm. As of right now, most of the individual's ability to interpret the motion of their arm through seeing it move.
(Kuiken, 2006)

The DEKA Research and Development group is the government agency that is now funding and heading the development of the bionic limbs. They are working in conjunction with Dr. Kuiken and the Rehabilitation Institute of Chicago. The DEKA arm is now the a technology advanced arm with sensory feedback, but uses other body parts for control instead of TMR. Together hopefully, the two technologies will mere fully soon.


Overall the study of biomechatronics has allowed for incredible innovations in the field of prosthetics. Individuals who have suffered from amputations are now able to have hope for mo
Jesse Sullivan and Claudia Mitchell's First High Five (The Age, 2006)
tion and modality. The Phantom limb is now longer thought of as needless and annoying aspect of amputation, but now has meaning and a purpose. With Dr. Kuiken and his team, a surgical procedure known as Targeted Reinnervation transfers residual nerves to a no longer biomechanically functional muscle. These nerves regrow and develop in the new muscle, producing both contractions and sensations of the phantom limb. The contractions provide myoelectric signals that can be interpreted by electrodes. These electrodes are connects to a mechanical motorized prosthetic that can translate the signals into purposeful movements. The future of these bionic limbs are immeasurable at the present time. As we learn more from those who have undergone the surgery and progress with those that will undergo surgery, we push the boundaries of this cutting edge science. Please see the links below for videos and more information on Dr. Kuiken, individuals who have undergone the procedure, and DEKA and the Rehabilitation Institute of Chicago labs.


Biomechatronics: the interdisciplinary study of biology, mechanics, and electronics. Biomechatronics focuses on the interactivity of biological organs, including the brain, muscles and bones with electromechanical devices and systems.
Cutaneous: of, relating to, or affecting the skin
Electromyogram (EMG): a technique for evaluating and recording the electrical activity produced by skeletal muscles
Intercostobrachial Cutaneous Nerve: sensory nerve that supplies the skin of the upper half of the medial and posterior part of the arm. It communicates with the posterior brachial cuntaneous branch of the radial nerve.
Musculocutaneous Nerve: originates from the brachial plexus and lateral cord (C5-C7) and continues to the lateral cunateous nerve of the forearm, innervating the anterior compartment of the arm.
Median Nerve: originates from the brachial plexus and lateral cord (C5-C8) and continues down the arm to the forearm. It innervates the anterior compartment except for two areas, the palm and lumbrical muscles of the hand.
Myoelectric Signal: a motor action potential; an electrical impulse that produces contraction of muscle fibers in the body
Neuromas: a growth or tumor of nerve tissue
Phantom Limb: the sensation that an amputated or missing limb is still attached to the body and is moving appripriately with other body parts.
Prosthetic: an artificial substitute or replacement for a part of the body, may be designed for functional or cosmetic reasons or both.
Radial Nerve: originates from the posterior cord of the brachial plexus and supplies the triceps brachii muscle and twelve other muscles in the posterior osteofascial compartment of the forearm. It also innervates the shoulder, elbow and wrist as well as the overlying skin.
Residual Nerves: left over nerves; those nerves that remain (i.e. after an amputation, the nerves that would continue down that limb)
Supraclavicular Sensory Nerve: originates from C3-C4 and innervates cutaneous receptors above and below the clavicle.
Targeted Muscle Reinnervation (TMR): method developed by Dr. Kuiken to transfer residual nerves to a biomechanically non-function muscle inorder to create EMG which can pick up myoelectric signals to be processed and interpreted by machines to produce voluntary, natural robotic limb motion.
Targeted Sensory Reinnervation (TSR): consequence of the targeted reinnervation, where afferent somatic sensory information of the amputated limb (phantom limb) can be felt in the target muscle area.


Multiple Chioce
1. All of the following are major functional areas cerebellum for descending motor pathway information EXCEPT:
a) Cerebrocerebellum
b) Spinocerebellum
c) Rubrocerebellum
d) Vestibulocerebellum

2.The dorsal column pathways carrying what type of information to higher cortical areas?
a) Cutaneous
b) Subcutaneous
c) Proprioceptive
d) All of the above

3. Which of the following is NOT a nerve typically transferred in TMR of shoulder and transhumeral amputees?
a) Radial Nerve
b) Humeral Nerve
c) Ulnar Nerve
d) Musculocutaneous Nerve

4. Sensory feedback of the amputated limb (phantom limb) can be felt where by an individual who has undergone targeted reinnervation surgery?
a) On the surface of the targeted area
b) Through the robotic device used by the individual
c) Both A & B
d) None of the above

5. The electrodes of the prosthetic are placed where on the individual after recovery from targeted reinnervation surgery?
a) Concentrated on the pectoral muscle and then few on other muscles involved in the desired movements
b) Few on the pectoral muscle and concentrated on the other muscles involved in the desired movements
c) Concentrated on the target muscle and few on other muscles involved in the desired movements
d) Few on the target muscle and concentrated on the other muscles involved in the desired movements

6. A prosthetic can be used for both functional and cosmetic reasons, unlike an orthotic which is only used for cosmetic purposes.

7. The Anterolateral System carries information regarding pain and temperature from the sensory receptors to higher cortical area, such as the thalamus.

8. Targeted Sensory Reinnervation was discovered on accident, meaning researchers were not expecting a sensory component to develop from the TMR surgery.

9. If you were born with a missing limb, or portion of a limb TMR surgery and mechanical prosthesis could be available to you and could be very beneficial.

10. The Pentagon and other government agency are looking into and aiding in the funding of this new technology so that it can be used to benefit those who have lost limbs in battle.

Short Answer
11. Briefly describe how a muscle contracts.
12. What is meant by controlling the motorized prosthetic through thought?
13. What are three aspects of the future of TMR and motorized prosthetics that are currently being considered?

Why is TSR now being involved in the original TMR surgical procedure? Why is sensory feedback important? Is it a vital feature of making the prosthetic component more natural? Why has it not been incorporated into the prosthetic yet? Why is it so difficult to integrate?


1. C
2. D
3. B
4. A
5. C
6. False
7. True
8. True
9. False
10. True


Dr. Kuiken at RIC
This site gives you access to Dr. Todd Kuiken at the Rehabilitation Institute of Chicago. Here you can see further information regarding his studies and patients as well as his current research. You also have access to his lab's website through this page.

Explanation and Demonstration Video
This video link is a wonderful display, explanation and overview of Prototype 1 from 2008. It shows first two individuals, Jesse Sullivan and Claudia Mitchell as they use their new bionic arms. First forty seconds is an introduction that can be skipped, but it is a much watch video!

CBS Bionic Arm Video
This video link show DEKA's involvement in the progression of this study. It is a 60 minutes special from 2009 that focuses on the DEKA Research and Development for veterans. Today, DEKA is working with RIC to develop the best possible arm.

Claudia Mitchell Demonstrates Video
This video link shows Claudia Mitchell, the second to undergo TMR surgery, recently performing precise tasks with great dexterity. This is where the progression of the research is now, incorporating technology from both DEKA and RIC.

Advanced Hand Control Video
This video link shows Amanda Kitts as she demonstrates advance hand coordination and accuracy in the lab.

Center for Bionic Medicine
This site is for the Center for Bionic Medicine at RIC. This will lead you to further information about the TMR process and patients that have undergone the surgery.


Human Neurophysiology. (2007) CNS Clinic. Descending Motor Pathways.

Freudenrich, C.(2010) Discovery Health. How Biomechatronics Works.

Kandell, ER., JH. Schwartz, and TM. Jessell (2000) eds. Principles of Neural Science. 4th ed. New York City:
McGraw-Hill Companies, Inc.

Kuiken, T.A. (2006, Feb) Targeted reinnervation for improved prosthetic function. Phys Med Rehabil Clin N Am., 17(1),1-13

Kuiken, T.A., Dumaniam, G.A., Lipschutz, R.D., Miller, L.A., Stubblefield, K.A. (2004) The use of targeted muscle reinnervation for improved myoelectric prosthesis control in a bilateral shoulder disarticulation amputee. Prosthetic and Orthotics International. 28(3), 245-253

Kuiken, T.A., Li, G., Lock, B.A., et al.(2009, Feb) Targeted muscle reinnervation for real-time myoelectric control of multifunction artificial arms. JAMA, 301(6), 670-671

Kuiken, T.A., Miller, L.A., Lipschutz, R.D., Stubblefield, K.A., Dumanian, G. (2005) Prosthetic command signals following tarted hyper-reinnervation nerve transfer surgery. Eng Med Biol Soc. 7, 7652-7655

Kuiken, T.A., Miller, L.A., Lipschutz, R.D., Lock, B.A., Stubblefied, K.A., et al. (2007) Targeted reinnervation for enhanced prosthetic arm function in women with a proximal amputation: a case study. Lancet, 369(9559), 371-380

McCormick. (2007) Northwestern Engineering, Biomedical Engineering. Todd Kuiken, MD, PhD

Neurophilosophy, Word Press (2006) Jesse Sullivan: The Bionic Man

Rehab Management Journal. (2006) News.

RIC. (2010) Prosthetics and Orthotics Services. Targeted Muscle Reinnervation: Control Your Prosthetic Arm With Thought.

Stubblefield, K.A., Miller, L.A., Lipschutz, R.D., Kuiken, T.A. (2009) Occupational therapy protocol for amputees with targeted muscle reinnervation. JRRD, 46, 481-488.

The Age. (2006) News. Bionic Arms turn science fiction to fact.

Zhou, P., Lowery, M., Dewald, J., Kuiken, T.A. (2005) Towards improved myoelectric prosthesis control: high density surface EMG recording after targeted muscle reinnervation. Eng Med Biol Soc. 5: 5276-5279.