Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

FROM COCHLEAR IMPLANTS TO BRAIN-COMPUTER INTERFACES

FROM COCHLEAR IMPLANTS TO BRAIN-COMPUTER INTERFACES In this article two groups of technologies based on connecting a medical device to the human brain are presented. The first group exploits the existing nerves, like the cochlear implant where ear prosthesis is connected to the auditory nerve. Another group is based on a direct connection between an electronic device and the human brain and it is called Brain-Computer Interfaces. The article contains the description of these technologies, points out their current capabilities and limitations and the main barriers to further development. The authors indicate possible directions of future expansion of the discussed technologies. KEYWORDS: ICT Implants, Cochlear Implants, Brain-Computer Interfaces. 1. Introduction In this article we describe technologies that use communication between the implant or external electronic device and the human brain. The first application where this kind of technology have been widely used is hearing restoration. Prosthetics of hearing, as everybody knows, include many methods and many technologies. Typical hearing loss can be repaired by simple hearing aids, which are often simple sound amplification devices. Such electrical (sound-gaining) hearing aids were first developed in early years of the 20th century and this technology is very popular until now. R. Tadeusiewicz and P. Rotter Because of increasing amount of people with hearing disability (elderly people or people working in noisy environment), behind-the-ear aid or in-theear aids became very common. But if deafness is very profound or occurs from birth hearing restoration process cannot be done using typical external apparatus for the deaf and we must think about more advanced device, implemented directly in human body, precisely speaking ­ in the human head. Such IT device is now available, and it is named cochlear implant. The method of hearing restoration based on cochlear implant is nowadays a standardized medical procedure. According to the U.S. Food and Drug Administration, about 219,000 people worldwide have received implants as of December 2010 and this number is increasing fast. In the next section we present foundations of this technology and medical procedures used for this purpose. We will describe the general structure of the considered implant, its functioning, medical applications and the results. Cochlear implant send the acoustic signal to the nervous system, more precisely to the auditory nerve but it is not regarded as Brain-Computer Interface (BCI), as electrical signals generated by the implant are sent to the brain indirectly. BCI is hardware and software for direct communication between the brain and external devices. Information can be passed from the brain to an external device, for example to enable control of computer or robot arm with thought or in opposite direction, e.g. from the artificial eye to the brain. The performance of BCI highly depends on invasiveness: In non-invasive BCI, no implant is used. The most common noninvasive method is based on EEG signal and other methods include for example magnetoencephalography and magnetic resonance imaging. Non-invasive BCI can pass information only in one direction (from the brain). Electromagnetic signals are blurred by the skull so non-invasive BCI cannot achieve the performance comparable with those based on implanted electrodes [1]. On the other hand they do not require a surgery, do not cause side effects and they are easy to put on and to take off. Invasive BCI includes electrodes implanted directly into the grey matter, which requires a complicated and risky surgery. This approach raises a number of health and ethical concerns. However invasive BCI allows for targeting precisely the desired area of the brain, both to read information and to transmit signal to the gray matter. Implanted electrodes enable applications which are unreachable for non-invasive methods due to their insufficient performance [2]. In partially-invasive BCI electrodes are implanted into the skull but outside of the brain. The operation is easier and brings less medical risks than in the case of invasive BCI but the performance is lower. The choice of the method (invasive or non-invasive) depends on the application. When high performance is required regardless high costs and risks, invasive methods are applied, especially if signal should be transmitted to the brain (e.g. for sight restoration). Non-invasive methods can be used rather for simple tasks in a range of fields, from entertainment to medical applications, like for example to provide quadriplegics with simple, but not very effective, way of communication. 2. Cochlear implants As was mentioned above, cochlear implants are very popular. In fact they are currently in massive production and there are a number of reports about their medical use. Moreover, methods of the cochlear implants application become precisely defined medical procedures. The fact that first successful applications of Information and Communication Technology (ICT) implants in the human nervous system are related to perception systems is definitely not a random effect. People suffering lack of some perception skills, e.g. the blind or the deaf, are very unhappy despite their general health status may be very good. Blindness or deafness is a kind of disability which affects both everyday functioning of a handicapped person and communication with other people, therefore we can observe increasing effort in the biomedical engineering area towards development of the perception systems prosthesis. The cochlear implant is probably the most popular and the most widely used type of implant that contacts with the human nervous system and sends information to the human brain. 2.1 General outline When describing the cochlear implant, we must start from the general outline of the whole system. Elements of such "artificial ear" are shown in Figure 1. Figure 1. General scheme of the cochlear implant ­ a kind of artificial ear (Source: http://www.ucsfbenioffchildrens.org/treat ments/cochlear_implant/index.html, Accessed 20 April 2012) R. Tadeusiewicz and P. Rotter In the figure we can see both biological and artificial (technological) elements. All these elements must work together. An implanted part of the system is located inside the patient body, under the skin covering the skull behind the ear, and must be connected to the internal ear biological structures through a small canal in the skull bones. Signal from the implant is transmitted to the cochlea, a biological sound receptor which is part of the internal ear ­ and because of that the whole system is named cochlear implant. The implanted part located under the skin must be connected not only to biological structures forming the internal ear, but also to the parts of the device that are located outside the skull. This communication is achieved through live tissues by wireless technologies. Let see, how it works. 2.2 Functioning of the cochlear implant Although the general outline of considered prosthesis is relatively simple and evident, now we must discuss some details explaining how the whole system works. Discussion will be easier when we take into account nature of the deafness. In healthy ear the following steps of sound signal processing can be taken into account (see Figure 2): sounds acquisition by the pinna and the ear canal, sounds registration by the eardrum and the middle ear structures, sounds reception by the inner ear (cochlea and Corti organ), and transmission of neural information characterizing sound to the brain by the auditory nerve. Figure 2. Scheme of the hearing process useful for explaining nature of deafness and role of the cochlear implant. (Source: http://www.earinc.com/images/about/how.jpg Accessed 24 April 2012) Deafness appears when one or more of the steps mentioned above is badly damaged. Sounds acquisition is almost never damaged, but all consecutive processes can be disturbed as result of many illnesses, accidents and also congenital defects. If a damage appears before the auditory nerve, the hearing process can be reconstructed by using discussed in this chapter IT device: cochlear implant. The general structure of the cochlear implant is designed as a technological analogy (and alternative in case of damage) of natural healthy human ear. Let see these analogies. In Figure 3 we can compare healthy human hearing system (see Figure 3 ­ top left part) and artificial ear, in this case ­ cochlear implant (see Figure 3 ­ bottom right part). Figure 3. Relations between natural hearing system and the cochlear implant. Detailed description in text. (Figure prepared by the author using images taken from the following websites: http://www.advancedbionics.com/userfiles/Image/componentsimage1.jpg and http://www.tanhearingcentre.com/EarAnatomy.jpg Accessed 19 March 2011) Let see step by step operation of both compared systems. The first step is sound registration. In healthy human ear this step is performed by the following anatomical structures: the pinna, the ear canal and the ear drum. The same function in the cochlear implant are performed by a microphone with an amplifier and speech processor. The structure and functions of elements of the cochlear implant will be described in detail in the next section, here we only describe functional analogies. R. Tadeusiewicz and P. Rotter The next step of the sound signal processing in the human ear is transmission of sound and the resulting mechanical vibrations to the hearing receptor, which should deliver sound signal to the brain. In the biological ear this function is performed by the middle ear structures ­ three small bones named the malleus, the incus, and the stapes, which are by the way the smallest bones within the human body. They are necessary because of big difference in acoustic impedance between the air transmitting sounds to the outer ear and fluid (endolymph) in the membranous labyrinth of the inner ear. Analogical elements in the cochlear implant structure are more complicated and must be divided into two parts: transmitter and receiver. There is necessity for wireless communication through the skin between artificial outer microphone and the implanted stimulator exciting Corti organ (natural element for sound analysis, registration, and neural signal generation), which is the most important part in low-level hearing system. The last part of the natural ear is mentioned above Corti organ located in the membranous labyrinth of the inner ear. It is the main natural hearing receptor. In fact it is rather complicated sounds analysis system, located in the cochlea. It has two functions: evaluation of sound spectral structure and conversion of the registered signal from mechanical form (waves of pressure in fluid contained in the membranous labyrinth of the inner ear) to the electrical form ­ nervous impulses going to the brain through the auditory nerve. In the cochlear implant instead of the natural biological receptor and sound analyzer we must use a special technological device. Because of its size and power consumption this part of the cochlear implant structure is located outside the human body ­ in previously discussed part, named speech processor. The implanted part contains only electronic device that prepares electrical signals and introduces it into the human nervous system (shortly speaking ­ to the brain). This part of the system is build in form of stimulating electrode. In fact it is not one electrode, but an array of up to 24 electrodes. This compound set of electrodes runs through the cochlea. From the cochlea signal of artificial stimulation goes through the auditory nerve system to the brain. Therefore this technology is useful in most types of deafness caused by pathologies localized in external and middle ear and in some pathologies localized in the inner ear (when neural part of the cochlea is not damaged). This technology cannot be used when hearing problem is related to neural structures. Now we discuss some details related to the cochlear implant structure and functioning. 2.3 Signal processing in the cochlear implant Some elements of the considered cochlear implant are simple and their structure and functioning are evident. Looking at Figure 4 we can see elements, which are known from many other applications of contemporary electronics. Nothing original can be found in a microphone and sound signal amplifier located in part A in Figure 4, not unique is wireless short distance communication device labelled as part B, rather typical is also electrical pulse generator working as inner ear stimulator (part C in Figure 4). S p Figure 4. Main parts of cochlear implant (Source: http://www.dartmouthengineer.com/wp-content/uploads/2009/08/cochlear_implant.jpg e Accessed 24 April 2012) e Original and unique for the cochlear implant are undoubtedly two c elements pointed with arrows in Figure 4 ­ and further discussion will be focused on this two special and unique solutions. First we present general h outline of the cochlear implant functioning. The cochlear implant operation in based on the fact that in natural hearing p process very important stage is conversion of sounds to so-called "tonotopic organization" of nervous signals going to the brain. Because of complex r mechanical properties of so-called basilar membrane, which is a part of the inner ear located inside the cochlea, sounds at different frequencies causes o excitation of hearing receptor cells in the Corti organ (so called hair cells) located in different positions inside the cochlea. There is ac precise mapping between sound frequencies and locations of the excitation area inside the Corti organ. When the sound includes strong high frequency components, it e excites receptor cells located near the base of the cochlea. On the contrary, low frequency sounds excite receptor cells in the upper (apical) part of the ss cochlea named helicotrema. o Excitation of receptor cells in the Corti organ creates an electrical signal that is registered by bipolar neural cells (spironeurons) that belong to the structure r named spiral ganglion, whose fibers (axons) form the auditory nerve going to the brain. Different spironeurons receive electrical signals from different hearing receptor cells located in the Corti organ and tonotopic organization is also observed in the auditory nerve. Sounds of different frequencies cause propagation of neural signals (impulses named spikes) in different fibres inside auditory nerve. Signals observed inside the brain are clustered by R. Tadeusiewicz and P. Rotter frequency and in this form they are interpreted ­ this is the nature of hearing sense. Taking into account the rules described above, the designers of the cochlear implant must first divide registered sound signals into segments assigned to particular frequency bands. This signal processing is performed in the speech processor. This name is related to the fact that the scheme of sound signal segmentation is adopted for best speech signal representation, not for even signal representation on whole sounds diapason. Such signal segmentation has one advantage and one disadvantage: deaf person equipped with the cochlear implant can hear and easy understand speech of other people, but unfortunately cannot enjoy all sounds, for example music. Such a person can hear music, but cannot savour it. Now we discuss structure of those parts of the cochlear implants that are unique for this type of apparatus. Figure 5. General diagram of speech processor with typical waveforms of signals in selected points of speech processor structure General diagram of speech processor with typical waveforms of signals in selected points of speech processor structure is presented in Figure 5. Speech processor first prepares signal from the microphone by means of so called "preemphasis filtering". A preemphasis filter is used for attenuation of strong and not very informative components of speech below 1.2 kHz. Preemphasis filter is followed by multiple processing channels, where sound energy is measured in particular frequency bands. Because the next step of signal transmission (internal ear electrical stimulation) cannot be performed very fast, therefore after signal segmentation the output of every band-pass filter in the speech processor is subjected to full-wave rectification and radical lowpass filtering. The envelope signals extracted from the processed bandpass filters are additionally compressed with a nonlinear mapping function. This transformation must be performed prior to the stimulation because of big difference between wide dynamic range of sound in the environment (up to about 100 dB) and narrow dynamic range of electrically evoked hearing (about 10 dB). Such prepared signals (so called auditory signals envelopes) are used for the control of the stimulation process. Compressed envelope signals are transmitted by the wireless communication system to the internal part of the cochlear implant system, located under the skin. General diagram of this part of the cochlear implant system is shown in Figure 6. Signal received by the implanted receiver is used to control the intra-cochlear stimulation process. The output of each channel (related to the corresponding bandpass filter in speech processor) is used for modulation with a special carrier signal: carrier signal, produced by special generator in the form of bipolar rectangular wave, is multiplied by the envelope signal from particular frequency band ­ and the resulting signal is used for stimulation of the inner ear structures. Outputs of multipliers (modulators) are connected to intracochlear electrodes. Assuming that the bandpass 1 is of highest frequency and bandpass n is of the lowest frequency ­ the electrode for bandpass 1 is located at the bottom part of the electrode array structure and next electrodes at consecutive more distant parts of the array, when the electrode for bandpass n is placed on the top of the array. Figure 6. General diagram of internal part of the implant R. Tadeusiewicz and P. Rotter Typical carrier waveform used for modulation of the signal before it is sent to the inner ear is shown in Figure 7 and typical waveforms of signals in selected parts of the structure of the internal part of the implant are shown in Figure 8. Every bandpass channel is directed to a single electrode, with lowto-high channels assigned to apical-to-basal electrodes, to mimic at least the order, if not the precise locations, of biological frequency mapping in the cochlea. Figure 7. Typical carrier waveform used for inner ear stimulation Figure 8. Typical waveforms of signals in selected parts of the structure of the internal part of the cochlear implant. The most important and the most subtle element of discussed structure of the cochlear implant is the intracochlear electrode array, shown in Figure 9. Figure 9. Intracochlear electrode array. (Source: http://bheldner.swissblog.ch/files/ 2009/02/flexeas_electrode.jpg, Accessed 19 March 2011) In real working position the electrode array structure is coiled inside the cochlea internal canal named scala tympani (Figure 10). Note small size of the considered structure! Figure 10. Electrodes array structure is coiled inside the cochlea (Source: http://www.freepatentsonline.com/7184843-0-large.jpg, Accessed 21 April 2012) Full visualization of the internal part of the cochlear implant located inside the head of a deaf patient is presented in Figure 11. R. Tadeusiewicz and P. Rotter Figure 11. Visualization of internal part of cochlear implant. (Source: http://www.chha-nl.nl.ca/05_Cochlear%20Implant%20Works%20Illustration.jpg, Accessed 19 March 2011) 2.4 Implantation Last but not least we can illustrate the method of implantation of the structures described above in the human body. Figures presented below are adopted from paper [3]. Figure 12. Introduction the electrode array to the cochlea (scala tympani): (A) drawing, (B) photography taken during the surgery. (Source: [3]) Figure 13. Fixation of the device in the well in temporal bone: (A) drawing, (B) photography taken during the surgery. (Source: [3]) 3. Brain-Computer Interfaces Cochlear implants are an example of well-established technology, where a medical device is connected to the human brain indirectly, in this case ­ though the auditory nerve. Below we present an emerging technology which is based on direct connection between an electronic device and the brain. 3.1 BCI for sight restoration The quality of cameras became sufficient to deliver a high quality image a long time ago so in theory digital cameras could replace biological eyes and even enhance human sight. However it is much more difficult to establish a connection between a camera and the brain than to construct the sensor. Experiments with camera-brain interface began in early 1970s (Dobelle, Mladejovsky 1974). A grid of electrodes implanted to the patient's visual cortex was used to produce phosphenes ­ flashes of light seen by a blind patient at different positions. Until the beginning of XXI century progress was made mostly thanks to the development of tools for image processing. By upgrading signal processing hardware and software Dobelle [5] achieved good results using electrodes implanted over 20 years earlier: the patients were able to omit obstacles and even read 15-cm letters from about 2 m. Several years later the first commercial BCI implant, constructed by William Dobelle, was used for the first time. R. Tadeusiewicz and P. Rotter Figure 14. A patient with Dobelle implant: image captured by the camera mounted in binoculars is processed and sent directly the brain. Source: http://www.jwen.com/rp/articles/dobelle.html, accessed 20 April 2012. The system shown in Figure 14 is similar to first BCI implants: the image from the camera mounted in binoculars is transmitted to aportable computer carried on the patient's belt. Image processing removes the background and noise and then the image is passed to electrodes implanted in the visual cortex. Rising numbers of electrodes allows to achieve better resolution. In 2000 Dobelle said that "it is unlikely that patients will be able to drive an automobile in the foreseeable future" [5]. Two years later a patient with 2nd generation Dobelle implant was able to drive slowly on parking area shortly after operation [6], although there is still a very long way before people with artificial sight can drive in normal traffic. Figure 15. A patient with second generation Dobelle implant cruising the parking after operation. Note that quality of BCI-based artificial eye is still too low (tunnel vision, starry night effect) to enable a patient to drive in normal traffic. Source: http://www.wired.com/wired/archive/10.09/vision_pr.html, Accessed 20 April 2012. BCI enabled blind people to walk carefully in the city or indoors but it is too early to speak about vision restoration. Recently progress has slowed; the major problems are that the image is perceived merely as a set of bright points on a black background (starry night effect) and the field of view is very narrow (tunnel vision). In other words, the perceived image is of extremely low resolution. The image seen by the patient does not facilitate colour vision or depth perception. The cost is much higher than any other technology designed to help the blind. Finally, there are important health concerns, which we address in section 4. 3.2 BCI for control of artificial limbs Simple control of external devices by intentional activity of the brain is possible using basic, non-invasive methods. In experiments carried out in mid-1990s at the University of Tübingen paralysed patients were trained to convert their brain activity to computer commands: intended variations of slow cortical potentials in their EEG was changed into binary signal so they could step-by-step select letters and write messages [7]. This example and many similar experiments demonstrate that non-invasive methods can be applied to transfer small amount of information, like binary signals. However their performance is insufficient for more sophisticated tasks, while invasive methods offer nowadays much higher quality of communication. Several years ago there was no doubt that if a patients can make eye movements or minimal movements of the thumb, muscular-based communication is more efficient than BCI (Neumann, Kübler 2003) but recently BCI-based methods have become an interesting alternative. In 2005 the first surgery was carried out which enabled a quadriplegic to control the artificial hand [9]. On the other hand precision of control is still unsatisfactory, even after long and frustrating training. It is therefore questionable whether currently benefits justify health risks (inflammation, implant rejection, swelling), not mentioning high costs of surgery. Currently there is no massive deployment of this technology, although research has already demonstrated its high potential. BCI-based prostheses are much more promising for patients who lost their limbs but are not paralysed. Especially technology of mind-controlled robot arms based on recently developed TMR (Targeted Muscle Reinnervation) has already proved high performance. Christian Kandlbauer was the first man with mind-controlled prosthesis who got driving license. His story is an example of success of TMR technology but also a warning: some time later he died after he crashed his car1. It is not clear if the bionic arm was the reason of his accident but certainly apart from very high potential of TMR technology, long-term experience is needed for its assessment. http://www.techeye.net/science/man-with-mind-controlled-prosthetic-arm-dies-in-car-crash R. Tadeusiewicz and P. Rotter Figure 17. Christian Kandlbauer and his mind-controlled left arm (right arm was a traditional prosthesis). Source: http://www.amsvans.com/blog/3034-europes-first-man-with-mind-controlled-bionic-armdies/, Accessed 20 April 2012. 3.3 Non-medical applications Invasive and semi-invasive methods require a complicated surgery and causes medical risks, so consequently they are accepted for medical applications only. The performance of non-invasive BCIs is currently too low for practical applications but advanced processing of EEG data may change this situation. For example artificial neural network may significantly improve classification of EEG signals [10]. Recently US Defense Advanced Research Projects Agency (DARPA) started project Silent talk, aiming at basic communication between soldiers in the battlefield through an "artificial telepathy". Before being vocalized, intended speech generate word-specific EEG pattern and classification of these signals could enable silent communication [11].2 In another DARPA project, Cognitive Technology Threat Warning System, one of the goals is to detect a threat feeling based on EEG signals generated by subconscious mind faster than a soldier can realize it. 3 Potential of BCIs has been recently noticed by computer game producers. Numerous games were developed with simple EEG-based interfaces which measure concentration of the player of even enable some basic control through the brain activity, like in NeuroBoy. 4 Transfer rate of BCI is too low Similar non-invasive system based on analysis of Motor Unit Actions Potentials (sensor is placed over muscles) instead of BCI has been already developed (http://www.theaudeo.com) http://www.wired.com/dangerroom/2007/04/soldierportable/ http://www.neurosky.com/mindset/neuroboy.html to be the only way of user interaction with the computer in actions games [12], but on the other hand using BCI may be attractive itself: for example in the game Mindball the goal is to move a ball up the sloping surface by being relaxed. It seems however that in the next years BCI-based games will remain a niche application. For massive development of BCI-based games a considerable progress of non-invasive BCI performance is needed. 4 Conclusions and future prospects Cochlear implants technology is very effective method of hearing restoration for absolutely deaf persons. For patients with badly damaged but existing hearing sometimes we can observe disadvantages related to two way sound perception: by natural hearing using for example bone sound transmission to the cochlea and by the cochlear implant. For these patients it is necessary to develop improved cochlear implants systems, which take into account existing leftovers of the natural hearing process and improve hearing in these frequency bands where the physiological elements of the ear are insufficient. This multi-sensor methodology can improve cochlear implant technology and it is going to be deployed in the future. Next cochlear implants development direction should be related to advanced signal processing in the speech processor. Now the sound received by the microphone is only divided into selected frequency bands and is subject of rather simple demodulation and non-linear compression. For enhancement of speech perception, especially in noisy environment, more advanced methods of sound signal processing should be developed, more similar to the signal analysis performed by the natural cochlea of healthy and efficient ear. Such processes apart from frequency distribution extraction require advanced form of non-linear signal processing. In fact such processes are not known enough yet and further development of cochlear implant technology depends on progress in psychoacoustic. While prosthesis of the hearing system for the deaf is well developed and in fact ­ very popular, the prosthesis of the eye, like any other application of BCI, is still at the stage of prototype rather than massive deployment. There are a number of medical risks related to invasive BCI, most of all infections, which were the reason why the wire connections to the brain of the best known blind volunteer, Jerry, had to be removed [13]. The results achieved in the past decade are very interesting from a scientific point of view but at the time being it is questionable whether the implant justifies high risk and very high cost. Coulombe, Sawan, Gervais in paper [14] proposed invasive BCI with reduced infection risk, where the electrodes are connected to the computer wirelessly. But, especially in case of vision prosthesis, other R. Tadeusiewicz and P. Rotter technologies may offer comparable (if not better) mobility for the patients at much lower cost and without medical risks ­ for example vOICe technology which changes the image into the sound.5 There are alternative solutions also for BCI that read brain activity, for example automatic analysis of eye movements or The Audeo ­ a non-invasive device which recognizes unpronounced speech by analysis of motor potentials which are too weak to vocalize intended sound [15]. On the other hand, in the future, we can expect increasing capabilities of BCI and this fact already raises ethical questions, for example: can this technology be used for enhancement of abilities of healthy people [16]? People with BCI implants might have additional benefits ­ for example artificial limbs are stronger and less vulnerable so if precisely controlled, they could be more capable than arms and legs. Cameras can see more details from the distance than the human eye; computer or TV connected directly to the brain interface [5] could offer a perfect virtual reality. Although currently BCI is very far from achieving the performance comparable with natural senses or limbs, in the future many people may be tempted to become a cyborg. Another ethical in context of quadriplegics is informed consent of the patient [17]. Finally, there is a risk of damage caused by mind-controlled device and a question about liability. A case of Christian Kandlbauer mentioned in section 3.2 raises not only technical but also legal issues. The border between a person (including even personality) and machine become very thin. The vision of reading people's thoughts and fillings, with or without their consent, is still science fiction nowadays 6 but may become reality in the future. For example the authors of the report issued by US National Research Council ask questions in context of BCI: "How can we disrupt the enemy's motivation to fight?", "How can we make people trust us more?" and finally "Is there a way to make the enemy obey our commands?" [18]. It cannot be excluded that, with further development of BCI, direct influence on people's thoughts may become technically possible in the future. Finally, combination of reading mind and feeding it with information can enable artificial telepathy, i.e. direct brain-to-brain communication. Although the above scenario is related to technology of the future rather than the one available today, such visions make BCI seen as a controversial technology. After several years of rapid deployment of BCI and recent years of apparent stagnation, it is difficult to foresee development of BCI in next years. Nowadays we can speak about success of technologies similar to BCI, which connect the brain with external world indirectly, through existing nerves or muscles ­ like cochlear implants, Targeted Muscle Reinnervation or Motor Unit Actions Potentials. "Pure" BCI technologies, where electrodes http://www.artificialvision.com Except from instruments for simple emotion detection, like a polygraph (lie detector) interact directly with the brain have proved to be feasible but a number of technical obstacles make them impractical today due to low performance, substantial health risks and high costs. We believe that all mentioned difficulties can be overcome in the future. As William Dobelle said, "Braille, the long cane and the guide dog are doomed to obsolescence" but we do not think it will happen in this decade. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Bio-Algorithms and Med-Systems de Gruyter

FROM COCHLEAR IMPLANTS TO BRAIN-COMPUTER INTERFACES

Download PDF
20 pages

Loading next page...
 
/lp/de-gruyter/from-cochlear-implants-to-brain-computer-interfaces-AllAhK6f00
Publisher
de Gruyter
Copyright
Copyright © 2012 by the
ISSN
1895-9091
eISSN
1896-530X
DOI
10.2478/bams-2012-0022
Publisher site
See Article on Publisher Site

Abstract

In this article two groups of technologies based on connecting a medical device to the human brain are presented. The first group exploits the existing nerves, like the cochlear implant where ear prosthesis is connected to the auditory nerve. Another group is based on a direct connection between an electronic device and the human brain and it is called Brain-Computer Interfaces. The article contains the description of these technologies, points out their current capabilities and limitations and the main barriers to further development. The authors indicate possible directions of future expansion of the discussed technologies. KEYWORDS: ICT Implants, Cochlear Implants, Brain-Computer Interfaces. 1. Introduction In this article we describe technologies that use communication between the implant or external electronic device and the human brain. The first application where this kind of technology have been widely used is hearing restoration. Prosthetics of hearing, as everybody knows, include many methods and many technologies. Typical hearing loss can be repaired by simple hearing aids, which are often simple sound amplification devices. Such electrical (sound-gaining) hearing aids were first developed in early years of the 20th century and this technology is very popular until now. R. Tadeusiewicz and P. Rotter Because of increasing amount of people with hearing disability (elderly people or people working in noisy environment), behind-the-ear aid or in-theear aids became very common. But if deafness is very profound or occurs from birth hearing restoration process cannot be done using typical external apparatus for the deaf and we must think about more advanced device, implemented directly in human body, precisely speaking ­ in the human head. Such IT device is now available, and it is named cochlear implant. The method of hearing restoration based on cochlear implant is nowadays a standardized medical procedure. According to the U.S. Food and Drug Administration, about 219,000 people worldwide have received implants as of December 2010 and this number is increasing fast. In the next section we present foundations of this technology and medical procedures used for this purpose. We will describe the general structure of the considered implant, its functioning, medical applications and the results. Cochlear implant send the acoustic signal to the nervous system, more precisely to the auditory nerve but it is not regarded as Brain-Computer Interface (BCI), as electrical signals generated by the implant are sent to the brain indirectly. BCI is hardware and software for direct communication between the brain and external devices. Information can be passed from the brain to an external device, for example to enable control of computer or robot arm with thought or in opposite direction, e.g. from the artificial eye to the brain. The performance of BCI highly depends on invasiveness: In non-invasive BCI, no implant is used. The most common noninvasive method is based on EEG signal and other methods include for example magnetoencephalography and magnetic resonance imaging. Non-invasive BCI can pass information only in one direction (from the brain). Electromagnetic signals are blurred by the skull so non-invasive BCI cannot achieve the performance comparable with those based on implanted electrodes [1]. On the other hand they do not require a surgery, do not cause side effects and they are easy to put on and to take off. Invasive BCI includes electrodes implanted directly into the grey matter, which requires a complicated and risky surgery. This approach raises a number of health and ethical concerns. However invasive BCI allows for targeting precisely the desired area of the brain, both to read information and to transmit signal to the gray matter. Implanted electrodes enable applications which are unreachable for non-invasive methods due to their insufficient performance [2]. In partially-invasive BCI electrodes are implanted into the skull but outside of the brain. The operation is easier and brings less medical risks than in the case of invasive BCI but the performance is lower. The choice of the method (invasive or non-invasive) depends on the application. When high performance is required regardless high costs and risks, invasive methods are applied, especially if signal should be transmitted to the brain (e.g. for sight restoration). Non-invasive methods can be used rather for simple tasks in a range of fields, from entertainment to medical applications, like for example to provide quadriplegics with simple, but not very effective, way of communication. 2. Cochlear implants As was mentioned above, cochlear implants are very popular. In fact they are currently in massive production and there are a number of reports about their medical use. Moreover, methods of the cochlear implants application become precisely defined medical procedures. The fact that first successful applications of Information and Communication Technology (ICT) implants in the human nervous system are related to perception systems is definitely not a random effect. People suffering lack of some perception skills, e.g. the blind or the deaf, are very unhappy despite their general health status may be very good. Blindness or deafness is a kind of disability which affects both everyday functioning of a handicapped person and communication with other people, therefore we can observe increasing effort in the biomedical engineering area towards development of the perception systems prosthesis. The cochlear implant is probably the most popular and the most widely used type of implant that contacts with the human nervous system and sends information to the human brain. 2.1 General outline When describing the cochlear implant, we must start from the general outline of the whole system. Elements of such "artificial ear" are shown in Figure 1. Figure 1. General scheme of the cochlear implant ­ a kind of artificial ear (Source: http://www.ucsfbenioffchildrens.org/treat ments/cochlear_implant/index.html, Accessed 20 April 2012) R. Tadeusiewicz and P. Rotter In the figure we can see both biological and artificial (technological) elements. All these elements must work together. An implanted part of the system is located inside the patient body, under the skin covering the skull behind the ear, and must be connected to the internal ear biological structures through a small canal in the skull bones. Signal from the implant is transmitted to the cochlea, a biological sound receptor which is part of the internal ear ­ and because of that the whole system is named cochlear implant. The implanted part located under the skin must be connected not only to biological structures forming the internal ear, but also to the parts of the device that are located outside the skull. This communication is achieved through live tissues by wireless technologies. Let see, how it works. 2.2 Functioning of the cochlear implant Although the general outline of considered prosthesis is relatively simple and evident, now we must discuss some details explaining how the whole system works. Discussion will be easier when we take into account nature of the deafness. In healthy ear the following steps of sound signal processing can be taken into account (see Figure 2): sounds acquisition by the pinna and the ear canal, sounds registration by the eardrum and the middle ear structures, sounds reception by the inner ear (cochlea and Corti organ), and transmission of neural information characterizing sound to the brain by the auditory nerve. Figure 2. Scheme of the hearing process useful for explaining nature of deafness and role of the cochlear implant. (Source: http://www.earinc.com/images/about/how.jpg Accessed 24 April 2012) Deafness appears when one or more of the steps mentioned above is badly damaged. Sounds acquisition is almost never damaged, but all consecutive processes can be disturbed as result of many illnesses, accidents and also congenital defects. If a damage appears before the auditory nerve, the hearing process can be reconstructed by using discussed in this chapter IT device: cochlear implant. The general structure of the cochlear implant is designed as a technological analogy (and alternative in case of damage) of natural healthy human ear. Let see these analogies. In Figure 3 we can compare healthy human hearing system (see Figure 3 ­ top left part) and artificial ear, in this case ­ cochlear implant (see Figure 3 ­ bottom right part). Figure 3. Relations between natural hearing system and the cochlear implant. Detailed description in text. (Figure prepared by the author using images taken from the following websites: http://www.advancedbionics.com/userfiles/Image/componentsimage1.jpg and http://www.tanhearingcentre.com/EarAnatomy.jpg Accessed 19 March 2011) Let see step by step operation of both compared systems. The first step is sound registration. In healthy human ear this step is performed by the following anatomical structures: the pinna, the ear canal and the ear drum. The same function in the cochlear implant are performed by a microphone with an amplifier and speech processor. The structure and functions of elements of the cochlear implant will be described in detail in the next section, here we only describe functional analogies. R. Tadeusiewicz and P. Rotter The next step of the sound signal processing in the human ear is transmission of sound and the resulting mechanical vibrations to the hearing receptor, which should deliver sound signal to the brain. In the biological ear this function is performed by the middle ear structures ­ three small bones named the malleus, the incus, and the stapes, which are by the way the smallest bones within the human body. They are necessary because of big difference in acoustic impedance between the air transmitting sounds to the outer ear and fluid (endolymph) in the membranous labyrinth of the inner ear. Analogical elements in the cochlear implant structure are more complicated and must be divided into two parts: transmitter and receiver. There is necessity for wireless communication through the skin between artificial outer microphone and the implanted stimulator exciting Corti organ (natural element for sound analysis, registration, and neural signal generation), which is the most important part in low-level hearing system. The last part of the natural ear is mentioned above Corti organ located in the membranous labyrinth of the inner ear. It is the main natural hearing receptor. In fact it is rather complicated sounds analysis system, located in the cochlea. It has two functions: evaluation of sound spectral structure and conversion of the registered signal from mechanical form (waves of pressure in fluid contained in the membranous labyrinth of the inner ear) to the electrical form ­ nervous impulses going to the brain through the auditory nerve. In the cochlear implant instead of the natural biological receptor and sound analyzer we must use a special technological device. Because of its size and power consumption this part of the cochlear implant structure is located outside the human body ­ in previously discussed part, named speech processor. The implanted part contains only electronic device that prepares electrical signals and introduces it into the human nervous system (shortly speaking ­ to the brain). This part of the system is build in form of stimulating electrode. In fact it is not one electrode, but an array of up to 24 electrodes. This compound set of electrodes runs through the cochlea. From the cochlea signal of artificial stimulation goes through the auditory nerve system to the brain. Therefore this technology is useful in most types of deafness caused by pathologies localized in external and middle ear and in some pathologies localized in the inner ear (when neural part of the cochlea is not damaged). This technology cannot be used when hearing problem is related to neural structures. Now we discuss some details related to the cochlear implant structure and functioning. 2.3 Signal processing in the cochlear implant Some elements of the considered cochlear implant are simple and their structure and functioning are evident. Looking at Figure 4 we can see elements, which are known from many other applications of contemporary electronics. Nothing original can be found in a microphone and sound signal amplifier located in part A in Figure 4, not unique is wireless short distance communication device labelled as part B, rather typical is also electrical pulse generator working as inner ear stimulator (part C in Figure 4). S p Figure 4. Main parts of cochlear implant (Source: http://www.dartmouthengineer.com/wp-content/uploads/2009/08/cochlear_implant.jpg e Accessed 24 April 2012) e Original and unique for the cochlear implant are undoubtedly two c elements pointed with arrows in Figure 4 ­ and further discussion will be focused on this two special and unique solutions. First we present general h outline of the cochlear implant functioning. The cochlear implant operation in based on the fact that in natural hearing p process very important stage is conversion of sounds to so-called "tonotopic organization" of nervous signals going to the brain. Because of complex r mechanical properties of so-called basilar membrane, which is a part of the inner ear located inside the cochlea, sounds at different frequencies causes o excitation of hearing receptor cells in the Corti organ (so called hair cells) located in different positions inside the cochlea. There is ac precise mapping between sound frequencies and locations of the excitation area inside the Corti organ. When the sound includes strong high frequency components, it e excites receptor cells located near the base of the cochlea. On the contrary, low frequency sounds excite receptor cells in the upper (apical) part of the ss cochlea named helicotrema. o Excitation of receptor cells in the Corti organ creates an electrical signal that is registered by bipolar neural cells (spironeurons) that belong to the structure r named spiral ganglion, whose fibers (axons) form the auditory nerve going to the brain. Different spironeurons receive electrical signals from different hearing receptor cells located in the Corti organ and tonotopic organization is also observed in the auditory nerve. Sounds of different frequencies cause propagation of neural signals (impulses named spikes) in different fibres inside auditory nerve. Signals observed inside the brain are clustered by R. Tadeusiewicz and P. Rotter frequency and in this form they are interpreted ­ this is the nature of hearing sense. Taking into account the rules described above, the designers of the cochlear implant must first divide registered sound signals into segments assigned to particular frequency bands. This signal processing is performed in the speech processor. This name is related to the fact that the scheme of sound signal segmentation is adopted for best speech signal representation, not for even signal representation on whole sounds diapason. Such signal segmentation has one advantage and one disadvantage: deaf person equipped with the cochlear implant can hear and easy understand speech of other people, but unfortunately cannot enjoy all sounds, for example music. Such a person can hear music, but cannot savour it. Now we discuss structure of those parts of the cochlear implants that are unique for this type of apparatus. Figure 5. General diagram of speech processor with typical waveforms of signals in selected points of speech processor structure General diagram of speech processor with typical waveforms of signals in selected points of speech processor structure is presented in Figure 5. Speech processor first prepares signal from the microphone by means of so called "preemphasis filtering". A preemphasis filter is used for attenuation of strong and not very informative components of speech below 1.2 kHz. Preemphasis filter is followed by multiple processing channels, where sound energy is measured in particular frequency bands. Because the next step of signal transmission (internal ear electrical stimulation) cannot be performed very fast, therefore after signal segmentation the output of every band-pass filter in the speech processor is subjected to full-wave rectification and radical lowpass filtering. The envelope signals extracted from the processed bandpass filters are additionally compressed with a nonlinear mapping function. This transformation must be performed prior to the stimulation because of big difference between wide dynamic range of sound in the environment (up to about 100 dB) and narrow dynamic range of electrically evoked hearing (about 10 dB). Such prepared signals (so called auditory signals envelopes) are used for the control of the stimulation process. Compressed envelope signals are transmitted by the wireless communication system to the internal part of the cochlear implant system, located under the skin. General diagram of this part of the cochlear implant system is shown in Figure 6. Signal received by the implanted receiver is used to control the intra-cochlear stimulation process. The output of each channel (related to the corresponding bandpass filter in speech processor) is used for modulation with a special carrier signal: carrier signal, produced by special generator in the form of bipolar rectangular wave, is multiplied by the envelope signal from particular frequency band ­ and the resulting signal is used for stimulation of the inner ear structures. Outputs of multipliers (modulators) are connected to intracochlear electrodes. Assuming that the bandpass 1 is of highest frequency and bandpass n is of the lowest frequency ­ the electrode for bandpass 1 is located at the bottom part of the electrode array structure and next electrodes at consecutive more distant parts of the array, when the electrode for bandpass n is placed on the top of the array. Figure 6. General diagram of internal part of the implant R. Tadeusiewicz and P. Rotter Typical carrier waveform used for modulation of the signal before it is sent to the inner ear is shown in Figure 7 and typical waveforms of signals in selected parts of the structure of the internal part of the implant are shown in Figure 8. Every bandpass channel is directed to a single electrode, with lowto-high channels assigned to apical-to-basal electrodes, to mimic at least the order, if not the precise locations, of biological frequency mapping in the cochlea. Figure 7. Typical carrier waveform used for inner ear stimulation Figure 8. Typical waveforms of signals in selected parts of the structure of the internal part of the cochlear implant. The most important and the most subtle element of discussed structure of the cochlear implant is the intracochlear electrode array, shown in Figure 9. Figure 9. Intracochlear electrode array. (Source: http://bheldner.swissblog.ch/files/ 2009/02/flexeas_electrode.jpg, Accessed 19 March 2011) In real working position the electrode array structure is coiled inside the cochlea internal canal named scala tympani (Figure 10). Note small size of the considered structure! Figure 10. Electrodes array structure is coiled inside the cochlea (Source: http://www.freepatentsonline.com/7184843-0-large.jpg, Accessed 21 April 2012) Full visualization of the internal part of the cochlear implant located inside the head of a deaf patient is presented in Figure 11. R. Tadeusiewicz and P. Rotter Figure 11. Visualization of internal part of cochlear implant. (Source: http://www.chha-nl.nl.ca/05_Cochlear%20Implant%20Works%20Illustration.jpg, Accessed 19 March 2011) 2.4 Implantation Last but not least we can illustrate the method of implantation of the structures described above in the human body. Figures presented below are adopted from paper [3]. Figure 12. Introduction the electrode array to the cochlea (scala tympani): (A) drawing, (B) photography taken during the surgery. (Source: [3]) Figure 13. Fixation of the device in the well in temporal bone: (A) drawing, (B) photography taken during the surgery. (Source: [3]) 3. Brain-Computer Interfaces Cochlear implants are an example of well-established technology, where a medical device is connected to the human brain indirectly, in this case ­ though the auditory nerve. Below we present an emerging technology which is based on direct connection between an electronic device and the brain. 3.1 BCI for sight restoration The quality of cameras became sufficient to deliver a high quality image a long time ago so in theory digital cameras could replace biological eyes and even enhance human sight. However it is much more difficult to establish a connection between a camera and the brain than to construct the sensor. Experiments with camera-brain interface began in early 1970s (Dobelle, Mladejovsky 1974). A grid of electrodes implanted to the patient's visual cortex was used to produce phosphenes ­ flashes of light seen by a blind patient at different positions. Until the beginning of XXI century progress was made mostly thanks to the development of tools for image processing. By upgrading signal processing hardware and software Dobelle [5] achieved good results using electrodes implanted over 20 years earlier: the patients were able to omit obstacles and even read 15-cm letters from about 2 m. Several years later the first commercial BCI implant, constructed by William Dobelle, was used for the first time. R. Tadeusiewicz and P. Rotter Figure 14. A patient with Dobelle implant: image captured by the camera mounted in binoculars is processed and sent directly the brain. Source: http://www.jwen.com/rp/articles/dobelle.html, accessed 20 April 2012. The system shown in Figure 14 is similar to first BCI implants: the image from the camera mounted in binoculars is transmitted to aportable computer carried on the patient's belt. Image processing removes the background and noise and then the image is passed to electrodes implanted in the visual cortex. Rising numbers of electrodes allows to achieve better resolution. In 2000 Dobelle said that "it is unlikely that patients will be able to drive an automobile in the foreseeable future" [5]. Two years later a patient with 2nd generation Dobelle implant was able to drive slowly on parking area shortly after operation [6], although there is still a very long way before people with artificial sight can drive in normal traffic. Figure 15. A patient with second generation Dobelle implant cruising the parking after operation. Note that quality of BCI-based artificial eye is still too low (tunnel vision, starry night effect) to enable a patient to drive in normal traffic. Source: http://www.wired.com/wired/archive/10.09/vision_pr.html, Accessed 20 April 2012. BCI enabled blind people to walk carefully in the city or indoors but it is too early to speak about vision restoration. Recently progress has slowed; the major problems are that the image is perceived merely as a set of bright points on a black background (starry night effect) and the field of view is very narrow (tunnel vision). In other words, the perceived image is of extremely low resolution. The image seen by the patient does not facilitate colour vision or depth perception. The cost is much higher than any other technology designed to help the blind. Finally, there are important health concerns, which we address in section 4. 3.2 BCI for control of artificial limbs Simple control of external devices by intentional activity of the brain is possible using basic, non-invasive methods. In experiments carried out in mid-1990s at the University of Tübingen paralysed patients were trained to convert their brain activity to computer commands: intended variations of slow cortical potentials in their EEG was changed into binary signal so they could step-by-step select letters and write messages [7]. This example and many similar experiments demonstrate that non-invasive methods can be applied to transfer small amount of information, like binary signals. However their performance is insufficient for more sophisticated tasks, while invasive methods offer nowadays much higher quality of communication. Several years ago there was no doubt that if a patients can make eye movements or minimal movements of the thumb, muscular-based communication is more efficient than BCI (Neumann, Kübler 2003) but recently BCI-based methods have become an interesting alternative. In 2005 the first surgery was carried out which enabled a quadriplegic to control the artificial hand [9]. On the other hand precision of control is still unsatisfactory, even after long and frustrating training. It is therefore questionable whether currently benefits justify health risks (inflammation, implant rejection, swelling), not mentioning high costs of surgery. Currently there is no massive deployment of this technology, although research has already demonstrated its high potential. BCI-based prostheses are much more promising for patients who lost their limbs but are not paralysed. Especially technology of mind-controlled robot arms based on recently developed TMR (Targeted Muscle Reinnervation) has already proved high performance. Christian Kandlbauer was the first man with mind-controlled prosthesis who got driving license. His story is an example of success of TMR technology but also a warning: some time later he died after he crashed his car1. It is not clear if the bionic arm was the reason of his accident but certainly apart from very high potential of TMR technology, long-term experience is needed for its assessment. http://www.techeye.net/science/man-with-mind-controlled-prosthetic-arm-dies-in-car-crash R. Tadeusiewicz and P. Rotter Figure 17. Christian Kandlbauer and his mind-controlled left arm (right arm was a traditional prosthesis). Source: http://www.amsvans.com/blog/3034-europes-first-man-with-mind-controlled-bionic-armdies/, Accessed 20 April 2012. 3.3 Non-medical applications Invasive and semi-invasive methods require a complicated surgery and causes medical risks, so consequently they are accepted for medical applications only. The performance of non-invasive BCIs is currently too low for practical applications but advanced processing of EEG data may change this situation. For example artificial neural network may significantly improve classification of EEG signals [10]. Recently US Defense Advanced Research Projects Agency (DARPA) started project Silent talk, aiming at basic communication between soldiers in the battlefield through an "artificial telepathy". Before being vocalized, intended speech generate word-specific EEG pattern and classification of these signals could enable silent communication [11].2 In another DARPA project, Cognitive Technology Threat Warning System, one of the goals is to detect a threat feeling based on EEG signals generated by subconscious mind faster than a soldier can realize it. 3 Potential of BCIs has been recently noticed by computer game producers. Numerous games were developed with simple EEG-based interfaces which measure concentration of the player of even enable some basic control through the brain activity, like in NeuroBoy. 4 Transfer rate of BCI is too low Similar non-invasive system based on analysis of Motor Unit Actions Potentials (sensor is placed over muscles) instead of BCI has been already developed (http://www.theaudeo.com) http://www.wired.com/dangerroom/2007/04/soldierportable/ http://www.neurosky.com/mindset/neuroboy.html to be the only way of user interaction with the computer in actions games [12], but on the other hand using BCI may be attractive itself: for example in the game Mindball the goal is to move a ball up the sloping surface by being relaxed. It seems however that in the next years BCI-based games will remain a niche application. For massive development of BCI-based games a considerable progress of non-invasive BCI performance is needed. 4 Conclusions and future prospects Cochlear implants technology is very effective method of hearing restoration for absolutely deaf persons. For patients with badly damaged but existing hearing sometimes we can observe disadvantages related to two way sound perception: by natural hearing using for example bone sound transmission to the cochlea and by the cochlear implant. For these patients it is necessary to develop improved cochlear implants systems, which take into account existing leftovers of the natural hearing process and improve hearing in these frequency bands where the physiological elements of the ear are insufficient. This multi-sensor methodology can improve cochlear implant technology and it is going to be deployed in the future. Next cochlear implants development direction should be related to advanced signal processing in the speech processor. Now the sound received by the microphone is only divided into selected frequency bands and is subject of rather simple demodulation and non-linear compression. For enhancement of speech perception, especially in noisy environment, more advanced methods of sound signal processing should be developed, more similar to the signal analysis performed by the natural cochlea of healthy and efficient ear. Such processes apart from frequency distribution extraction require advanced form of non-linear signal processing. In fact such processes are not known enough yet and further development of cochlear implant technology depends on progress in psychoacoustic. While prosthesis of the hearing system for the deaf is well developed and in fact ­ very popular, the prosthesis of the eye, like any other application of BCI, is still at the stage of prototype rather than massive deployment. There are a number of medical risks related to invasive BCI, most of all infections, which were the reason why the wire connections to the brain of the best known blind volunteer, Jerry, had to be removed [13]. The results achieved in the past decade are very interesting from a scientific point of view but at the time being it is questionable whether the implant justifies high risk and very high cost. Coulombe, Sawan, Gervais in paper [14] proposed invasive BCI with reduced infection risk, where the electrodes are connected to the computer wirelessly. But, especially in case of vision prosthesis, other R. Tadeusiewicz and P. Rotter technologies may offer comparable (if not better) mobility for the patients at much lower cost and without medical risks ­ for example vOICe technology which changes the image into the sound.5 There are alternative solutions also for BCI that read brain activity, for example automatic analysis of eye movements or The Audeo ­ a non-invasive device which recognizes unpronounced speech by analysis of motor potentials which are too weak to vocalize intended sound [15]. On the other hand, in the future, we can expect increasing capabilities of BCI and this fact already raises ethical questions, for example: can this technology be used for enhancement of abilities of healthy people [16]? People with BCI implants might have additional benefits ­ for example artificial limbs are stronger and less vulnerable so if precisely controlled, they could be more capable than arms and legs. Cameras can see more details from the distance than the human eye; computer or TV connected directly to the brain interface [5] could offer a perfect virtual reality. Although currently BCI is very far from achieving the performance comparable with natural senses or limbs, in the future many people may be tempted to become a cyborg. Another ethical in context of quadriplegics is informed consent of the patient [17]. Finally, there is a risk of damage caused by mind-controlled device and a question about liability. A case of Christian Kandlbauer mentioned in section 3.2 raises not only technical but also legal issues. The border between a person (including even personality) and machine become very thin. The vision of reading people's thoughts and fillings, with or without their consent, is still science fiction nowadays 6 but may become reality in the future. For example the authors of the report issued by US National Research Council ask questions in context of BCI: "How can we disrupt the enemy's motivation to fight?", "How can we make people trust us more?" and finally "Is there a way to make the enemy obey our commands?" [18]. It cannot be excluded that, with further development of BCI, direct influence on people's thoughts may become technically possible in the future. Finally, combination of reading mind and feeding it with information can enable artificial telepathy, i.e. direct brain-to-brain communication. Although the above scenario is related to technology of the future rather than the one available today, such visions make BCI seen as a controversial technology. After several years of rapid deployment of BCI and recent years of apparent stagnation, it is difficult to foresee development of BCI in next years. Nowadays we can speak about success of technologies similar to BCI, which connect the brain with external world indirectly, through existing nerves or muscles ­ like cochlear implants, Targeted Muscle Reinnervation or Motor Unit Actions Potentials. "Pure" BCI technologies, where electrodes http://www.artificialvision.com Except from instruments for simple emotion detection, like a polygraph (lie detector) interact directly with the brain have proved to be feasible but a number of technical obstacles make them impractical today due to low performance, substantial health risks and high costs. We believe that all mentioned difficulties can be overcome in the future. As William Dobelle said, "Braille, the long cane and the guide dog are doomed to obsolescence" but we do not think it will happen in this decade.

Journal

Bio-Algorithms and Med-Systems – de Gruyter

Published: Jan 1, 2012

There are no references for this article.