Currently, cochlear implants cost between $15,000 – $25,000 without including the $20,000 – $30,000 for screening, surgery, and rehabilitation. In the United States, the price tag can even be over $100,000. These prices make sense in the developed world, where the implants are covered by private and public health insurance companies, but they are not practical for most of the rest of the world without health insurance, which contains about 80% (and growing) of the world population.
Hearing loss affects 360 million people worldwide with a disproportionately high number of them in developing countries because of marriages of blood-related parents. The target market for the simplified cochlear implant described in this paper is the 100 million poorest, profoundly deaf people, none of whom are in the United States. In this paper, I lay out a plan for a trial of 100 implants to be implanted in the first year of production and scaling up from there. Currently the only major company involved in this market is Med-El, and they just give away full-priced implants that were designed for more wealthy consumers. They have done this for less than 500 people.
I examine different methods for lowering costs of cochlear implants without drastically lowering the quality of the product. Specifically, I compare the methods for dealing with the four major conflicts in CI development in developing countries: cost, skill of surgeons, maintenance, and distribution. I also make predictions about the future of low-cost cochlear implants and recommend best areas to improve in the future.
A cochlear implant (CI) is a small electronic device that can provide a sense of hearing to people who are hearing-impaired or profoundly deaf. They are made up of a speech processor to filter unnecessary sounds and amplify necessary ones into a useful signal, a transmitter to convert the signal to electrical impulses, and an array of electrodes to receive the electrical impulses and send them to specific regions of the auditory nerve.
In this proposed device, a four channel CIS processor divides the signal from a microphone into four separate frequency ranges: 2800-5600 Hz, 1400-2800 Hz, 700-1400 Hz, and 100-700 Hz. Interleaved pulses at these frequencies cross the skin of the oval window through transmitting coils. Then the signal is sent to a monopolar array of intra-cochlear electrodes on the medial wall of the scala tympani to stimulate the nerve at the same frequency as the sound would normally be (between 100 and 5600 Hz). This device would cost less than $100 to produce, and I expect it to perform about the same as normally priced implants in speech recognition tests. Along with a longer battery life, fewer moving parts, and less chance of damage in extreme environments, this is the best technology to invest in as cochlear implants become available outside of countries with developed medical systems.
This project is about a cheap, four-channel cochlear implant for use in developing countries. I describe the device in detail, but I also spend a lot of time talking about the systems that will use the device. I made this choice because I think that the system through which a device is implemented is an important part of the design decisions that create the device. For devices meant to be used in countries with stable health-care systems, this can be left out of design documents because the reader will assume them. But, for new contexts, the focus of design decisions broadens. For cochlear implants, this involves analyzing both the electronic itself, and the system that will implement it .
Despite major increases in the quality of cochlear implants in the United States, Europe, and Australia, western biased innovations in this area leave the huge majority of the world without access to cochlear implants. With insurance companies creating a barrier between the manufacturer and consumer, cochlear implant companies have spent large amounts of money for incremental improvements in the quality of their products. Because of this, the entire process from screening , purchasing an implant, and rehabilitation can cost over $100,000 in the United States. This is acceptable in developed countries with stable healthcare systems, but this makes cochlear implants unattainable for the 80% of the world’s population that lives in developing countries, and the more than a third of the world’s population that lives on less than $2 per day. This is further complicated because the income for people with severe hearing loss is even less than for people with normal hearing in the same circumstances .
In order to solve this problem, the most useful thing to do is to examine how people have tried to solve it in the past. A brief summary of attempted solutions is below.
Between the years 2000 and 2005, 52 individuals were given high-end Combi 40+ cochlear implants produced by MED-EL medical electronics in Pakistan. This is a high-end system that is approved for use in the United States. 44 of these patients were children, and there was an 11.5% complication rate, which is much higher than the complication rate in the US.
The need for cochlear implants is especially high in Pakistan because marriage between blood relatives is extremely common. 76% of the children in this cohort were born to blood-related parents.
This test was conducted on a self-financing basis. Since Pakistan has a per capita income of $600, only an extremely small percentage of the country’s population was able to consider this option. Pre-surgical screening consisted of an MRI, and after the surgery, patients had to return for rehabilitation. This meant that all of the patients lived in one of the three major cities in the country. To help with the rehabilitation process, they also had patients meet with existing cochlear implant users, which the researchers concluded was the best way for low-cost follow-up in countries without freely available medical supervision.
The team concluded that cochlear implant programs can be set up successfully, but the implants that were designed for use in America and Australia did not do well in dusty desert conditions. They also concluded that financial considerations will remain the main constraint for the foreseeable future. Fortunately, I think that it is possible for countries to produce their own cochlear implants, which better meet the needs of their populations .
There is a program in China where a Chinese company (Nurotron) manufactures lower cost cochlear implants than the ones available in the United States. They have research offices in the United States, but all manufacturing happens in China, and the implants are made exclusively for people in China. They are known for creating cochlear implants that are simple and have long operating times, which matches up well with a consumer base that cares a lot about keeping maintenance costs low. But, the income levels in China are still higher than most of the world. The Chinese cochlear implant program is one of the most successful in the developing world, but their success still comes primarily from patients with incomes much higher than normal for the country.
Other developing countries that have had small-scale government funded cochlear implant trials include Saudi Arabia, South Africa, Hungary, Brazil, Argentina, Mexico, Cuba , and Columbia, but these have all been severely limited due to lack of funding.
In Egypt, some patients were randomly selected for the trial to be paid for by the government, but some of the people chosen did not have enough money to afford transportation for a follow-up visit.
One last confounding factor is that speech recognition tests are still under development for languages other than English. Many design decisions have been made biasing implants towards distinguishing between sounds common in the English language, and that sounds unique to other languages have been mostly ignored. This would be a topic I would be interested in following up about in a separate paper, but this variable goes beyond the scope of this assignment which is possibly already too broad.
Four Limiting Factors:
Looking at the previous and current programs for the development of cochlear implants, I have reduced all of the complications down to four main limiting factors. Every decision that is made optimizing cochlear implants for developing countries should be made with these factors in mind.
- Money: People cannot afford doctor’s fees, cannot afford to purchase the implant, and cannot afford transportation to the hospital. Patients are also not likely to go back to a doctor if there are complications because they cannot afford to.
- Skill of surgeons: Developing countries have major physician shortages, especially in specialties relating to surgery. Most of the previous trials were performed with surgeons from developed countries, which meant that this variable was usually ignored in the design process, but since most of the surgeons did not want to stay in developing countries long-term, the programs ended when they left. It is important that surgical procedures be simplified as much as possible.
- Maintenance: Patients in developing countries are usually exposed to the elements more than the users in developed countries. This can include rain, but dust causes issues even more. Patients may not live near a medical facility to replace bad parts, and may not be able to do maintenance that others might consider routine like changing a battery. Design decisions should be made so that implants are resilient to extreme environments, and long battery life must be a priority.
- Distribution: Patients have to live near the centers where the cochlear implants are distributed and surgically implanted. Travel is extremely difficult in many countries, since infrastructure quality is often related to the income of a location. This will improve over time as the infrastructure in these countries improves, and once these cochlear implants begin to be mass produced.
Detailed Description of Mechanism:
A cochlear implant (CI) is a small electronic device that can provide a sense of hearing to people who are hearing-impaired or profoundly deaf . They are made up of a speech processor to filter unnecessary sounds and amplify necessary ones into a useful signal, a transmitter to convert the signal to electrical impulses, and an array of electrodes to receive the electrical impulses and send them to specific regions of the auditory nerve. 
One of the most important factors in cochlear implant design is choosing the number of electrodes. The devices on the market in the United States have between 12 and 22 channels,  but people who have as few as four-channel implants receive the same scores in speech-recognition tests. For this project, four electrode seems to be the best, because that is the fewest number of electrodes that has been demonstrated to allow comprehension at the same rate as higher-end models. The next section of this paper goes through each of the parts of the cochlear implant individually.
Since the optimal number of electrodes for this use is four, a four channel speech processor is required.  There are many different processing techniques for speech. The MPEAK Strategy detect when soundwaves cross the soundwave of a fundamental frequency. The timing of these ‘zero-crossings’ becomes a signal that is extracted through envelope signal detectors and outputting high-frequency bands to fixed electrodes. Then these electrodes would be stimulated at slightly different from the fundamental frequency, which the brain would then be able to interpret. Although this strategy is mostly effective, it creates many falsely positive signals, especially when it is in noisy environments.  This requires the patient to have their implant fine-tunes significantly after surgery, which may not be a viable option for many people in developing countries .
Another common approach is the compressed-analog (CA) approach. This was the speech processing system used in the Ineraid device from Symbion, and the UCSF/Storz device . The gain of the signal is controlled and sorted into four frequencies (0.5, 1, 2, and 3.4 kHz). Then the waveforms are set to a standardized magnitude and delivered simultaneously to the four electrodes in analog form. This is a system that has been reported to work very well by multiple sources .
The third common speech processing technique is called the continuous interleaved sampling (CIS) approach. This approach is very similar to the CA approach, but it fixes one major issue. Sometimes, when multiple channels are stimulated at the same time interference can occur which degrades speech comprehension. Using the CIS approach, each of the four channels would deliver their signals briefly and then turn off when the other frequencies are delivered.  This happens fast enough that there is no chance of a patient realizing that engineers took this shortcut. Switching rapidly between simultaneous frequency inputs also allows frequencies between the usual four signals to be produced through simultaneous stimulations . This is the approach that I think is the most promising for cheap cochlear implants in the future because it combines the features of the first two processing systems and can create virtual electrodes, which is cheaper than increasing the number of electrodes.
In any speech processor there are three important steps in the processing: bandpass filters which break the speech into the four frequencies mentioned above, envelope detectors, which find the edges (envelopes) of high-frequency signals, and pulse generators, which create the signal to actually pas down the wire.
Some form of signal transduction is necessary to get the electrical signals from outside of the ear to the inside. The main design decision that has to be made with respect to the signal transduction is whether to follow a percutaneous or transcutaneous pathway. In a transcutaneous connection, radio frequency signals travel from an external coil to an implanted coil. The advantage of this system is that the skin on the scalp can be closed during the surgery , which decreases the chance of infection. In a percutaneous connection, signals are sent to the electrodes through a direct connection. This allows clearer signal and allows defective parts to be replaced without a surgery, but the rates of infection can be very high . It makes sense that this would become even more of an issue in a developing country, where infections are more common than in the United States .
For these reasons, as well as the fact that signal processing technology has been progressing steadily, which allows waveforms to be sent more accurately across membranes than they were in the past, a coil-coupled transcutaneous system seems best .
There are four main design decisions to make with regard to the electrodes in cochlear implants:
- The placement of the electrode: either near the round window of the cochlea, in the scala tympani, or on the surface of the cochlear nucleus . The scala tympani is recommended by the Wilson paper.
- The number of electrodes and the spacing of their contacts. There will be four electrodes spaced apart equally inserted to a depth of 22mm.
- The orientation of electrodes with respect to the tissue that they will be stimulating. The electrodes should go on the medial wall of the scala tympani.
- The configuration of the electrodes. A monopolar electrode array should be used. Monopolar electrodes have longer battery lives , and do not produce significantly worse speech recognition. They are also cheaper.
The information for the recommended design of the electrodes comes almost exclusively from the Wilson’s  and Loizou’s  articles. The only difference between the prototype in the Wilson paper, and the one that I think would be best comes from selection of materials. In the Wilson paper, the contacts were made of a platinum-iridium mixture, but they also say that copper wire can be used too in order to lower the cost of manufacturing even further. The electrodes can be made from copper wire and welded to the lead of the electrodes. It would be important that the ground electrode also be made of the same material, so that no energy is built up between higher and lower conductivity surfaces. The copper wires would also have to be carefully insulated so that no galvanic corrosion occurs. 
This prototype will be evaluated in two phases. The first phase would be to build it in an artificial ear, made out of a similarly resistive material to the human ear. Input signals would be placed into the microphone, and the output signals would be recorded to compare similarity. This would be modeled after the tests that were run in the Wilson paper , so look at it for more details.
The second phase of testing would be the human trials. Since, human trials have also already been done with a very similar setup (except not with copper electrodes) in the Wilson paper , I would model my experimental setup after theirs. Words will be spoken to patients under various conditions and I will compare how many words they hear are actually words that I said. A system of tests can also be conducted with beeps of various frequencies to see what frequencies the patient can actually hear.
I’ll post more here as this product develops.
 M. Bond, S. Mealing, R. Anderson, J. Elston, G. Weiner, R. S. Taylor, M. Hoyle, Z. Liu, A. Price, and K. Stein, “The effectiveness and cost-effectiveness of cochlear implants for severe to profound deafness in children and adults: a systematic review and economic model.,” Health Technol. Assess., vol. 13, no. 44, pp. 1–330, 2009.
 World Health Organization (WHO), “Newborn and infant hearing screening,” WHO Libr. Cat. Data, no. November, pp. 9–10, 2009.
 M. B. Tarabichi, C. Todd, Z. Khan, X. Yang, B. Shehzad, and M. M. Tarabichi, “Deafness in the developing world: the place of cochlear implantation.,” J. Laryngol. Otol., vol. 122, no. 9, pp. 877–880, 2008.
 B. Shield, “Evaluation of the Social and Economic Costs of Hearing Impairment.,” Hear-it, no. October, p. 202, 2006.
 M. I. J. Khan, N. Mukhtar, S. R. Saeed, and R. T. Ramsden, “The Pakistan (Lahore) cochlear implant programme: issues relating to implantation in a developing country.,” J. Laryngol. Otol., vol. 121, no. 8, pp. 745–750, 2007.
 T. Le Roux and C. Laurent, “Open Access Guide To Audiology and Hearing Aids for Otolaryngologists,” Open Access Atlas Otolaryngol., pp. 1–6, 2012.
 G. Giraldo, “Breaking the sound barrier: Cuba’s Cochlear implant program.,” MEDICC Rev., vol. 12, no. 1, pp. 1–4, 2010.
 Nidcd, “NIDCD Fact Sheet: Cochlear Implants,” Heal. (San Fr., 2009).
 “Cochlear Implants,” NIH Publication No. 11-4798, 2014. [Online]. Available: http://www.nidcd.nih.gov/health/hearing/pages/coch.aspx. [Accessed: 02-Feb-2016].
 W. D. Holding, “Cochlear Implant Comparison Chart,” no. April, pp. 1–9, 2013.
 B. S. Wilson, S. Rebscher, F. G. Zeng, R. V. Shannon, G. E. Loeb, D. T. Lawson, and M. Zerbi, “Design for an inexpensive but effective cochlear implant,” Otolaryngol. – Head Neck Surg., vol. 118, no. 2, pp. 235–241, 1998.
 P. C. Loizou, “Mimicking the human ear,” IEEE Signal Processing Magazine, vol. 15, no. 5. pp. 101–130, 1998.
 F.-G. Zeng, “Cochlear Implants in Developing Countries,” CICI Contact, vol. Winter, 1996.
 A. Tjellstrom, B. Papsin, and C. W. R. . Cremers, “Baha 3 Surgery Guide: A Bone Conduction Solution,” pp. 1–21, 2010.
 M. Chatterjee and J. Yu, “A relation between electrode discrimination and amplitude modulation detection by cochlear implant listeners.,” J. Acoust. Soc. Am., vol. 127, no. 1, pp. 415–426, 2010.