Rohan Sagar
SFSU - English 670
October 12, 2017
Annotated Bibliography(Draft 3)
Instructor: Sofia de Almeida
Myo-Electric Control for Robotic Prosthetics
It is a bitter truth that accidents keep happening everyday and some of them even lead to people losing their limbs. Hand amputations in particular can drastically affect the victim’s capabilities. There is a wide variety of prosthetics available in the market for victims with such disabilities. However, most of the people prefer a cosmetic alternative over a functional robotic prosthetic. This is because the robotic prosthetics that are available in the market today are either too expensive or too complicated to use. In order to make them inexpensive and intuitive to the user a technique called Myo-electric Control can be used.
Electromyography (EMG) is a technique used for evaluating and recording the electrical activity of skeletal muscles. The device used to obtain the EMG signal is called electromyograph. Whenever a muscle is contracted, electrical activity occurs in the muscle, which can then be recorded by using the electromyograph. Myo-Electric control utilizes electromyography to extract the data from muscles and uses it to control the desired object. Researchers at ICE (Intelligent Computing and Embedded Systems Lab ) in San Francisco State University have designed a robotic prosthetic arm that uses Myo-electric control in an effort to make robotic prosthetics inexpensive. The first part of this two part annotated bibliography that follows talks about the research being done by them.
Even though the myo-electric control is promising, prosthetics can be made better by using small electronic chips embedded inside the skin of a person which collect data directly from the nervous system. For this idea to be practical, the chip that is embedded inside the person should be small. One of the technology that allows for the electronic circuits to become smaller than they are today is the use of nanotubes instead of transistors, hence the second part of this annotated bibliography presents the research being done by the experts in the field of nanotechnology.
The use of an implant can significantly improve the response time and accuracy of a prosthetic. However, there are many challenges to overcome such as battery life extension, the cost of implanting and even the idea of having an electronic implant in the body of a human.
Atzori, M., & Müller, H. (2015). Control capabilities of myoelectric robotic prostheses by hand amputees: a scientific research and market overview. Frontiers in systems neuroscience, 9.
Atzori and Miller, in this article talk about the crucial role that robotic prosthetics can play in the lives of about 41000 people with the major loss of an upper limb (in 2005). They also talk about how to use myoelectricity (electricity generated when muscles are activated ) and pattern recognition algorithms to mimic the function of a real hand or leg. They list detailed specifications of many products that are available in the market today. Finally they say that the prosthetics that exist today may not be as robust and they can be, but recent research being done in this domain is promising.
The paper talks about the need of prosthetics in the present day and thus establishes the need to research them and make them better. It serves as a good source of information regarding the existing market for prosthetics as it states the specifications of most of the commercially available prosthetics. This paper can be considered as a good first paper to read while researching on Myo-Electric control of robotic prosthetics, because it establishes the need to research robotic prosthetics.
Zhang, X., Huang, H., & Yang, Q. (2012, June). Implementing an FPGA system for real-time intent recognition for prosthetic legs. In Design Automation Conference (DAC), 2012 49th ACM/EDAC/IEEE (pp. 169-175). IEEE.
This paper presents the design and implementation of a system that uses an FPGA (Field programmable gate array) and a Microcontroller unit to recognize and replicate the intended movement in a prosthetic limb. It goes into detail about why the microcontroller and FPGA are used where they are used. The microcontroller system uses its ADC’s (Analog to Digital Converters) to get the analog data from the electromyographic (EMG) sensors and convert it to digital data. The digital data is sent to an FPGA which does all the processing and pattern recognition using an SPI (Serial Peripheral Interface) bus. The paper presents the data pertaining to the system performance by providing the delays for systems using 7 EMG sensors and 12 EMG sensors which are 0.23 and 0.25ms respectively.
The article explains a system that can be used to control a robotic prosthetic limb thus provides insights on how to build a system that is capable of myo-electric control. By giving an example it explains the architecture of a myo-electric control system. This paper is useful to the research done on Myo-electric control because it talks about how to practically build an embedded system that can perform pattern recognition and data acquisition at the same time.
Farina, D., Jiang, N., Rehbaum, H., Holobar, A., Graimann, B., Dietl, H., & Aszmann, O. C. (2014). The extraction of neural information from the surface EMG for the control of upper-limb prostheses: emerging avenues and challenges. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 22(4), 797-809.
In this paper the authors talk about the connection between neural drive information and electromyogram (EMG) signal. They talk about how neural drive can be interpreted from the EMG signal using a mathematical relationship between them. While talking about neural drive, they say that it would be easier to get the neural drive information directly from the spine but that would need a surgery and so the only option today, is to rely on EMG signals. The authors also talk about the factors that influence the surface EMG signal acquisition such as changes in electrode skin potential due to sweat and relative movement of muscles with respect to electrode.
This paper helps in understanding the characteristics of EMG signal with respect to muscle activation. By analyzing these characteristics of an EMG signal one can extract the action intended by the person. This paper gives detailed information in terms of neural drive and EMG signal characteristics and thus acts as a foundation for EMG pattern recognition.
Englehart, K., & Hudgins, B. (2003). A robust, real-time control scheme for multifunction myoelectric control. IEEE Transactions on Biomedical Engineering, 50(7), 848-854.
In this paper Englehart and Hudgins talk about some basic necessities for a good functional robotic prosthetic such as accuracy, interface and response time. They also talk about the history of myo-electric robotic prosthetic control in the 1970s. However, the technology that existed then didn’t support the complex computational requirements of pattern recognition algorithms. The advancement in semiconductor technologies in the current day makes it possible to perform such calculations. They talk about a technique that can be used to extract the intent of the user, called as pattern recognition using sliding window technique (taking a small window of values and performing pattern recognition continuously instead of waiting to collect huge amounts of data and process it all at once). They explain this technique in detail and also present the findings of the experiment conducted on a workstation that implemented the above mentioned technique using a software called MATLAB. They found out that a system with 1Ghz Pentium 3 processor takes roughly 16ms to process the information using the sliding window technique.
This paper gives key information on how the Electrographic data can be analyzed to facilitate pattern recognition and myo-electric control. The technique mentioned in this paper is efficient in terms of processor resource utilization and reducing response time. This paper serves as a guide on how to frame data acquisition and analysis algorithms efficiently.
Mitra, S., & Acharya, T. (2007). Gesture recognition: A survey. IEEE Transactions on Systems, Man, and Cybernetics, Part C (Applications and Reviews), 37(3), 311-324.
This paper talks about gestures in general and gives some introduction to gesture recognition. It explains the different tools that can be used for gesture recognition, such as The Hidden Markov’s Model,The Particle Filtering, The Condensation algorithm and The Finate State machine approach. It also talks about all the different gestures, such as hand gestures, face and head gestures and facial recognition. It explains how gestures can be recognized as a face or hand gestures. This is done by using a camera and image processing techniques. The paper talks about how all the tools mentioned above for gesture recognition can be used in each of these gestures to recognize it.
This paper gives a good overview of gestures and explains how gesture recognition is achieved. However, it is only partially useful to my current research as the paper talks about gesture recognition using a camera and image processing techniques, whereas we are interested in gesture recognition using electromyographic data and pattern recognition.
Bachtold, A., Hadley, P., Nakanishi, T., & Dekker, C. (2001). Logic circuits with carbon nanotube transistors. Science, 294(5545), 1317-1320.
The authors of this article talk about miniaturizing electronics with a technology known as Carbon Nanotube Transistors, instead of the Field Effect Transistors that are being used today.They also talk about the properties of Carbon Nanotube Transistor Circuits in detail. Namely Gate voltage, Source to Drain Voltage, Doping level. They mention the materials and the processes that are currently in use to make Field Effect Transistors and explains in detail, a 3 stage process for manufacturing Carbon Nanotube Transistors. They talk about fabrication of a NOR gate, a Flip-flop and a Ring oscillator circuit using Carbon Nanotube Transistors instead of Field Effect Transistors and present the data that they recorded while examining the circuit for its standard parameter values. Finally they say that apart from the usual properties of a regular transistor, these transistors have some additional features, but there is still a long way to go in perfecting carbon nanotube transistors for use in industry.
This paper provides detailed information on the use of carbon nanotubes in electronic circuits instead of regular semiconductor transistors. By using these to manufacture electronic chips we can make devices small enough to be inserted into the body by the use of a needle. Once the device is inside the body it can bond with the nerves and relay information to the prosthetic. This paper puts a new spin on my research by opening up the possibility to get neural information directly from the spinal cord.