A New Era of Prosthesis
A Prosthesis or Prosthetic Implant is an artificial device that can be used for compensating the inconvenience caused by a damaged or missing limb. Body parts may be lost by birth, or by trauma, disease, accident, etc. The first oldest discovered prosthesis is not an artificial arm, leg, or eye patch, but an artificial toe. It was originated from Egypt, most probably between 950–710 B.C.E. In ancient Roman history, General Marcus Sergius is considered to be the first known wearer of an artificial limb. It was a right arm, fashioned from iron.
Today, modern materials like carbon fiber and silicones are playing a vital role in making prosthetic limbs both lighter, stronger, and more durable. Advancement in 3D printing and biometrics has enabled the fabrication of more effective prosthesis. It is very convenient to have prosthetic limbs that are capable of mimicking the sensory capabilities and responsive activities of the original limbs. Until recently, all of these were unimaginable; but there has been some breakthrough.
According to Robert Armiger, the project manager for Amputee Research at Johns Hopkins University’s Applied Physics Lab, “Bio-engineers are looking to create human-machine interfaces embodied by a prosthetic limb that feel like an extension of the body”. FDA (Food and Drug Administration, USA) has a vision outlined for the advancement in invention and testing for AI-based brain-implanted devices which are capable of controlling prosthetic devices attached to the body. The Defense Advanced Research Projects Agency (DARPA) is funding up to tens of millions into these types of new projects for further development in prosthetics.
According to the estimation created by the non-profit Amputee Coalition, there are near about 2 million amputees in the USA and the number of amputees is expected to become nearly double to hit a 3.6 million margin by 2050. The statistics on the global scale are certainly a lot worse than this (although there is no reliable record of amputation at a global scale). Each year, about 1,85,000 lower limb amputations are made with 30% of people suffering from depression and/or anxiety. This is because these artificial limbs can not resemble the original limbs as these can not be used or controlled by the brain like the original.
In the robotic arm being developed at Johns Hopkins APL, there are about 26 joints and load cells in each fingertip which are supposed to detect force and sense the torque applied to each metacarpophalangeal joint. Feedbacks on outside excitation, e.g., pressure, temperature, stress, etc collected by the embedded sensors which work much like nerves in a normal arm are fed to the central system that can simulate the corresponding response in the brain. Also, it can be controlled by the brain in the same manner as a normal arm.
A group of researchers from the University of Utah has developed a brain-controlled prosthetic arm that is capable of simulating 100 different natural touch sensations in the human brain. The project is funded by DARPA and is titled “LUKE Arm”, named after the character “Luke Skywalker” from the Star Wars movie who used to have a prosthetic arm. The Utah researcher group adopted the strategy of implanting an intelligent sensing device in the user’s nervous system and implanting electrodes in electrodes in muscles to transfer signals originally generated by the artificial limb to the brain, enabling the user to have a natural sense of touch. According to one of the developers on the project, Jacob George, “People often think of touch as a single sense, but it’s subdivided into other senses, such as pressure, vibration, temperature, pain, etc. The high resolution of our device allows us to activate these sub-classes of touch in isolation (i.e., pressure without vibration or pain) in a specific part of the hand.” The research team mapped each of the 192 electrodes with each subject’s nervous system to establish sensation correspondence with a real stimulus. For example, if pressure is applied to a specific area, e.g., the tip of the index finger, the user feels that pressure in that exact place of the index finger of the prosthetic arm instead of another area, much like what a normal arm usually does. After the completion of these steps, the user is expected to have natural control over the prosthetic arm, just like the original arm.
At the beginning of 2018, a group of researchers from Imperial College of London and the University of Gottingen together developed an AI-based prosthetic hand. This prosthetic hand is capable of learning with experience with the help of artificial intelligence to improve its interaction with the user’s nervous system and brain signal. Their experiment with their prototypes conducted on five prosthetic-end users has provided satisfactory results. According to the researchers involved in the development of this bionic hand, published in Science Robotics, the new machine-learning-based control provides far better natural, fluid movements than the currently available prosthesis. This AI-enabled bionic hand uses a brain-computer interface (BCI) to interpret the user’s brain signal and sends commands to the embedded system of the prosthetic arm. It has 8 electrodes to fetch weak electrical signals from the user’s scalp. After training the machine learning algorithm from the data collected from the brain signal, the regression model was uploaded to the embedded system of the limb and used to control the prosthetic arm. As in the conventional process, the bionic hand was moved in velocity-controlled mode. The movement of the arm was proportional to the contraction force applied by the user. The bionic hand underwent training with the user sensory data so that the machine learning algorithm could learn the interpretation of each user’s unique electronic brain signals. After the training process, the effect of AI implementation comes into action. A mini-computer is embedded with the bionic limb. The weak electronic signals from the brain are first amplified and then sent to the mini-computer. The mini-computer then runs the machine learning algorithm to interpret the signal, prior to manipulating the bionic hand’s motor movement according to the way the user wants. Thus a prosthetic limb becomes an integral part of the body.
The challenge is much more for the lower-limb amputees. Body movements and balance are dependent on lower-limbs. Bionic legs are trickier to develop compared to bionic hands. An open-source AI-based bionic leg was exhibited at Amazon’s Re: MARS Conference in Las Vegas on June 5, 2019. This leg can do the course of movements: cuts, pivots, turns, and of course running and walking. This prosthesis uses a Raspberry-Pi powered AI-based control system. Much like the same principle used in the bionic hand discussed latter, the AI-based control of this bionic leg uses a combination of muscle contraction signals and sensor data to predict the user’s movement desire and move accordingly, thus successfully mimic normal leg operation.
The concept of prosthetics is not only confined to motor functioning organs, i.e., arms and legs but can also be applied to non-motor functions like vision. At the University of Louvain, a researcher named Claude Veraart has designed a Microsystem-based Visual Prosthesis (MIVP). This prosthesis contains a spiral cuff electrode. This electrode is implanted around the optic nerve situated at the back of the eye. It is connected to the brain via a stimulator which is implanted in a small dent in the skull. Signals from an externally worn camera are fed to the stimulator. The stimulator converts the camera signals into electrical signals that enable the stimulation of the optic nerve according to the surrounding, thus mimicking natural vision inside the brain. Although this prosthesis is currently in its research and development phase, this shows great potential in the field of vision prosthesis.
With the massive advancement in AI technology and development in the fabrication materials like carbon fiber and silicones, successful human augmentation has been possible. Now such prostheses can be built that can mimic natural human limbs, thus giving mankind superpowers, so to say. Although most of these prostheses are in the development phase and still very expensive, it is expected that this human level prosthesis will be widely available and become cheaper to afford very soon.
