Skip to main content
Clinical Orthopaedics and Related Research logoLink to Clinical Orthopaedics and Related Research
. 2022 Apr 14;480(6):1205–1207. doi: 10.1097/CORR.0000000000002213

CORR Insights®: Feasibility of a Wireless Implantable Multi-electrode System for High-bandwidth Prosthetic Interfacing: Animal and Cadaver Study

Pietro Ruggieri 1, Andrea Angelini 1
PMCID: PMC9263472  PMID: 35420548

Where Are We Now?

Amputation has an incidence of about 1.5 per 1000 people and is performed because of trauma, peripheral vascular disease, tumor, infection, and congenital anomalies [10]. The use of an external device to replace the missing part of the body has come a long way from primitive cosmetic prostheses to advanced implants with seamless human–machine integration [2]. This is important in patients with upper limb amputation, considering the high incidence of prosthesis abandonment owing to unsatisfactory functional results and poor quality of life [9]. Myoelectric prostheses are one of the main solutions to achieve functional recovery after upper limb amputation; they have biological signals measured through electrodes (internal wire or needle electrodes, implanted electrodes, or surface electrodes) and are used to control movement of the prosthesis. The basic concept of implanted myoelectric devices for recording intramuscular EMG signals has been widely reported for more than 20 years in patients with spinal cord injury [6-8]. The use of an implanted multi-channel myoelectric device may improve signal source recording and limit noise, but there is a major concern of long-term biocompatibility [4]. A myoelectric prosthesis uses electrodes to measure action potential and control the external device through muscle electricity. The term “crosstalk” defines the interaction between the muscles and the prosthesis control system. The ideal myoelectric control can be achieved by combining multiple EMG signals recorded independently and concurrently from specific muscles with the functions or multiple degrees of freedom to be controlled. Several implantable myoelectric systems for signal detection and prosthesis control are in development. Previous implantable systems relied on an existing telemetry architecture that required a large inductive field to be maintained through a cross-section of the whole limb, facilitated by a large coil that encloses the residual limb. Recently, there have been fascinating advances in myoelectric control, such as the decoding of motor neuron spike trains directly extracted from multi channel EMG [5].

The current study [3] is a clinically oriented paper presenting a new solution to man–machine interface. The authors [3] reported a well-tailored study on the myoelectric implantable recording array system, confirming the implant has full biocompatibility, continued functionality, and mechanical stability after long-term in vivo testing. This study should be considered the beginning of a “proof of concept,” and could lead to examinations of functional differences achieved by increasing the number of electrode sites from five or six to 32. Researchers could use these results to increase the number of movements that can be studied, improving prosthesis control and use at all levels of major upper limb amputation [11].

Based on these discoveries, and considering that patients learn from the actions of the machine and change their behavior for the best results, patient learning could be achieved with prerecorded or actual signals through online or real-life experimentation, allowing external prostheses to function more realistically and accurately.

Where Do We Need To Go?

The paper by Gstoettner et al. [3] raises important questions. Most important among them: What is the appropriate biotechnological interface between the dense neuromuscular data of the biologic signal recording and prosthesis control, and what can be done to alleviate this bottleneck to improve functional outcomes in patients with external prostheses after upper limb amputation? Decoding information transmitted from the brain to the muscles or directly from activated muscles using EMGs is a complicated task. Surgeons are directly involved, maintaining the function of active muscles or through surgical transfer of residual nerves to alternative muscles where surface EMG signals can be recorded (targeted muscle reinnervation). However, large amounts of biologically acquired data risk being redundant, providing useless information and making it difficult to transfer inputs to the prosthesis. How can we deal with the fact that although the anatomy of the upper limb is complex with respect to sensing and movement, most prostheses will be much simpler? Further work is needed to better explore these topics and place them in the context of existing evidence and knowledge about biologic signal transfer to external prostheses in patients who have undergone amputation. The orthopaedic community and biomedical engineers should find ways to support each other in improving the functional outcomes of external prostheses. This is especially true in patients with proximal upper limb amputations, where the clinical need for improved interfacing strategies is the greatest.

Implantable approaches with a high channel count and stable signal acquisition are necessary to improve the functionality of prostheses. Functional restoration is strictly related to the level of upper limb loss: transcarpal, transradial, wrist, or elbow disarticulation; transhumeral and shoulder disarticulation; or forequarter amputation. Currently available prostheses are not compatible with all amputation levels because of their limits in the interfaces adopted for signal recording, prosthesis control, and sensory feedback [1]. Implantable muscle electrodes are less susceptible to signal alterations because of postural changes, and noise reduction allows for much lower detection thresholds.

How Do We Get There?

A full understanding of neural and behavioral changes that result from amputation remains far from reach, even if research in this field has constantly been grown in recent decades. Prosthesis control is based on a bidirectional electric-to-biological signal conversion. Implantable recording systems have potential advantages, although other systems appear promising, such as an electroencephalography brain–computer interface, which is a system that can record neurologic signals that follow activation patterns in the cortex with noninvasive electrodes. Regarding the bottleneck between recorded and output signals, there are two aspects that should be fully analyzed to minimize this gap. First, adaptations from humans to machines have been developed for decades, as described above, through myoelectric signals that try to interpret the patient’s commands and the information of what he or she intends. In fact, biological myoelectric signals from activated muscles are not the only input for controlling the function of a limb. Specific algorithms should be created to simplify the wide spectrum of recorded signals to generate the desired output and allow the use of simpler models with high accuracy. The possible combination of a large multichannel recording system with decoding algorithms for real-time prosthesis control will be helpful. We think a high-density data acquisition system with up to 12 channels of EMG signals seems adequate for the task, even if some authors reported that having more than six channels did not reduce the estimation error and performance did not improve [12]. Second, adaption from machines to humans should be improved by exploring feedback systems that give the user information about what the prosthesis is feeling. An efficient sensory feedback system is the foundation on which effective control of the external prosthesis is based. Currently, academic research is far from clinical use, and there is a need to integrate the patient’s perception of sensors and training into the system to improve prosthesis control and facilitate handling. Without this, the prosthesis will be a simple tool that cannot replace a missing limb’s function. Artificial intelligence and big-data analyses may help researchers analyze the learning parameters that influence the coadaptive human–machine learning process. Although there are few studies in this field, research and prototype models are continuously upgraded, albeit with slow and expensive developments. Still, those researching commercial upper limb external prostheses could take advantage of fast-growing emerging technologies such as three-dimensional printing, smartphones, intelligent robotics, and other consumer and industry sectors. This cooperation will improve the functional outcomes of external prostheses. The desire to make things better should be the foundation of research, and the work of Gstoettner et al. [3] is a viable tool with expected benefits for clinical use.

Footnotes

This CORR Insights® is a commentary on the article “Feasibility of a Wireless Implantable Multi-electrode System for High-bandwidth Prosthetic Interfacing: Animal and Cadaver Study” by Gstoettner and colleagues available at: DOI: 10.1097/CORR.0000000000002135.

The authors certify that there are no funding or commercial associations (consultancies, stock ownership, equity interest, patent/licensing arrangements, etc.) that might pose a conflict of interest in connection with the submitted article related to the authors or any immediate family members.

All ICMJE Conflict of Interest Forms for authors and Clinical Orthopaedics and Related Research® editors and board members are on file with the publication and can be viewed on request.

The opinions expressed are those of the writers, and do not reflect the opinion or policy of CORR® or The Association of Bone and Joint Surgeons®.

References

  • 1.Cordella F, Ciancio AL, Sacchetti R, et al. Literature review on needs of upper limb prosthesis users. Front Neurosci. 2016;10:209. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.El Damaty S, Hazubski S, Otte A. ArtiFacts: creating a 3-D CAD reconstruction of the historical Roman capua leg. Clin Orthop Relat Res. 2021;479:1911-1913. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Gstoettner C, Festin C, Prahm C, et al. Feasibility of a wireless implantable multi-electrode system for high-bandwidth prosthetic interfacing: animal and cadaver study. Clin Orthop Relat Res. 2022;480:1191-1204. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Grill WM, Norman SE, Bellamkonda RV. Implanted neural interfaces: biochallenges and engineered solutions. Annu Rev Biomed Eng. 2009;11:1-24 [DOI] [PubMed] [Google Scholar]
  • 5.Igual C, Pardo LA, Jr, Hahne JM, Igual J. Myoelectric control for upper limb prostheses. Electronics. 2019;8:1244. [Google Scholar]
  • 6.Kilgore KL, Smith B, Campean A, et al. Powering strategies for implanted multi-function neuroprostheses for spinal cord injury. Healthc Technol Lett. 2020;7:81-86. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Makowski N, Campean A, Lambrecht J, et al. Design and testing of stimulation and myoelectric recording modules in an implanted distributed neuroprosthetic system. IEEE Trans Biomed Circuits Syst. 2021;15:281-293. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Memberg WD, Polasek KH, Hart RL, et al. Implanted neuroprosthesis for restoring arm and hand function in people with high level tetraplegia. Arch Phys Med Rehabil. 2014;95:1201-1211. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Østlie K, Lesjø IM, Franklin RJ, Garfelt B, Skjeldal OH, Magnus P. Prosthesis rejection in acquired major upper-limb amputees: a population-based survey. Disabil Rehabil Assist Technol. 2012;7:294-303. [DOI] [PubMed] [Google Scholar]
  • 10.Sabzi AS, Taheri AA. Amputation: a ten-year survey. Trauma Monthly. 2013;18:126-129. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Tenore FV, Ramos A, Fahmy A, Acharya S, Etienne-Cummings R, Thakor NV. Decoding of individuated finger movements using surface electromyography. IEEE Trans Biomed Eng. 2009;56:1427-1434. [DOI] [PubMed] [Google Scholar]
  • 12.Young A, Smith L, Rouse E, Hargrove L. Classification of simultaneous movements using surface EMG pattern recognition. IEEE Trans Biomed Eng. 2013;60:1250-1258. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Clinical Orthopaedics and Related Research are provided here courtesy of The Association of Bone and Joint Surgeons

RESOURCES