A cochlear implant is a small electronic device that lets many deaf people hear, by skipping the damaged part of the inner ear and sending tiny electrical pulses straight to the hearing nerve. By the end of 2019, about 737,000 had been implanted around the world. It was the first medical device to bring back a human sense in a big way, and it took decades of doubt to get there. Here is how it works, what it does and does not fix, and how a new gene therapy, approved in the U.S. in April 2026, may let a small group of deaf children hear without one.
How normal hearing works
Sound is a wave of pressure in the air. The ear canal funnels it to the eardrum, and three tiny bones pass the vibration on to the inner ear. There it reaches the cochlea, a fluid-filled tube coiled up like a snail shell.
Inside the cochlea sit thousands of sensory hair cells. When the fluid moves, they bend and turn that motion into electrical signals, which travel along the hearing nerve (also called the auditory nerve) to the brain.
The cochlea is laid out a little like a piano keyboard. High-pitched sounds, like a baby’s cry, stir the hair cells near the entrance of the spiral. Low sounds, like a rumbling truck, stir the ones deep in the middle. The brain reads which spot is active to work out the pitch, and it also uses the timing of the signals.
In most people with severe hearing loss, the problem is that the hair cells are damaged or missing. Often the nerve that they used to feed is still there and still works. That gap is what a cochlear implant fills.
What the implant actually does
A hearing aid is basically an amplifier. It makes sound louder so a damaged ear can pick it up. A cochlear implant does something different. It skips the damaged hair cells and stimulates the hearing nerve directly with electricity.
The U.S. National Institute on Deafness and Other Communication Disorders (NIDCD) lists four main parts. A microphone picks up sound. A speech processor sorts and arranges that sound. A transmitter and a receiver-stimulator turn the result into electrical pulses. An electrode array carries those pulses to different parts of the hearing nerve.
The microphone and processor are worn outside, usually behind the ear. A round coil, held on the side of the head by a magnet, sends power and signals by radio through the skin. Under the skin, the implanted receiver picks them up and sends pulses down a thin, flexible lead that a surgeon threads into the spiral of the cochlea.
Each electrode along that lead sits at a different spot in the cochlea. The processor splits incoming sound into frequency bands and sends each band to the electrode in the matching spot, high notes near the entrance and low notes deeper in. The device uses the cochlea’s own piano layout. But it has only a couple of dozen electrodes at most (Clark’s original Nucleus design had 22) to stand in for thousands of hair cells. That is why implant hearing is useful, but rougher than natural hearing.
A shock in the head
The idea that electricity can make a person hear is older than the light bulb. Around 1800, the Italian physicist Alessandro Volta, who had just invented the electric battery, put a metal rod in each of his ears and connected them to the battery. He described a jolt in the head, followed by a noise like a kind of crackling or bubbling, as if some dough or thick material was boiling. He found the experience unpleasant and did not want to repeat it.
The first real attempt to stimulate the hearing nerve directly came in Paris in 1957. Two French doctors, André Djourno and Charles Eyriès, placed an electrode on what was left of the nerve in a deaf patient who was having surgery. The patient could hear the pulses and tell loud from soft, but could not understand natural speech, and the device failed after a few weeks.
News of that attempt reached William House, an ear doctor in Los Angeles. In 1961, House and neurosurgeon John Doyle placed electrodes in two patients. House’s single-electrode device let people hear sounds around them and helped them lip-read, but speech came through badly distorted. A longer-lasting version went on sale in the early 1970s.
Many experts thought the whole idea was hopeless. The signals the brain needs to hear speech seemed far too complex to copy with a few crude pulses. The Australian surgeon Graeme Clark later recalled hearing it said that “direct stimulation of the auditory nerve fibres with resultant perception of speech is not feasible.”
Many channels, and a seashell on the beach
One wire could not do the job, because a single electrode cannot use the cochlea’s piano layout. Two teams set out, separately, to build implants with many electrodes, each aimed at a different spot.
In Vienna, the electrical engineer Ingeborg Hochmair and her future husband, Erwin Hochmair, built a multichannel implant that was first placed in a patient in December 1977. In Melbourne, Clark, whose father was a deaf pharmacist, had spent years on the same goal. His team was so short of money that young graduates stood on the streets of Melbourne shaking tins to raise funds.
One of Clark’s hardest problems was mechanical: how to slide a bundle of wires around the tightening spiral of the cochlea without tearing anything. The idea came to him at the beach. He used a spiral seashell as a model of the cochlea and blades of grass as the electrodes. Grass is floppy at the tip and gets stiffer toward the base, and that is what let it follow the spiral. His electrode bundle was built with the same graded stiffness.
Clark’s first patient, in August 1978, was Rod Saunders, who had lost his hearing at age 46. Before the operation, Saunders told him, “I would like to be able to hear something again. It’s a nightmare being deaf.” By December, Saunders could identify some spoken words he had not been told in advance. In 1985, the U.S. Food and Drug Administration (FDA) approved Clark’s multichannel device, sold by the company Cochlear as the Nucleus, for adults who could hear before they went deaf. It was the first multichannel implant the FDA approved. In 1990, the approval was extended to children aged 2 to 17.
The software that made speech clear
Even with many electrodes, early implants helped most users mainly alongside lip-reading. The next leap was not in the hardware but in the way the sound was coded.
Blake Wilson, an electrical engineer at the Research Triangle Institute in North Carolina, worked on this for years with funding from the U.S. National Institutes of Health (NIH). In 1991, his team reported in the journal Nature a method called continuous interleaved sampling, or CIS. The trick is that no two electrodes fire at exactly the same moment. The pulses take turns, very quickly, so they do not blur into each other inside the cochlea.
According to the Lasker Foundation, CIS let most implant users understand words and sentences without any visual cues for the first time. It also spread fast, because the NIH-funded work was placed in the public domain, so every manufacturer could use it. It became the basis for sound-processing methods that are widely used today. In 2013, Clark, Ingeborg Hochmair, and Wilson shared the Lasker-DeBakey Clinical Medical Research Award, one of medicine’s top prizes, for the modern cochlear implant.
What it sounds like, and why it takes practice
An implant does not give back normal hearing. The NIDCD says it gives a useful representation of sounds and can help a person understand speech, but that hearing through it is different and takes time to learn or relearn. Over weeks and months of practice, the brain learns to make sense of the new patterns.
Results vary widely from person to person. Many users can talk on the phone and follow a conversation in a quiet room. The hard parts are noisy places and music. In his Lasker acceptance remarks, Wilson said that even top performers struggle in “noisy restaurants or workplaces,” and that music is “less than satisfying for most patients.” Pitch is coarse, so tone languages such as Mandarin, where pitch changes the meaning of a word, are also harder.
The surgery is usually done under general anesthesia, and many people go home the same day. Complications are uncommon but real. They include infection, dizziness, and, rarely, damage to the facial nerve. Because implants carry a small risk of bacterial meningitis, the U.S. Centers for Disease Control and Prevention (CDC) recommends that they get vaccinated against pneumococcal bacteria.
Why timing matters for children
For a child born deaf, the clock matters. The brain sets itself up for spoken language in the first few years of life, and it needs sound to do that. The NIDCD says that children who get an implant early, followed by intensive therapy, often hear and speak better than those implanted later. Since 2020, the FDA has approved implants for eligible children from 9 months of age, and some of those children develop language at a pace close to their hearing peers.
Implants have also been part of a long argument. Many Deaf people, especially those whose first language is a sign language, see deafness as a culture and an identity rather than something that needs fixing. Some have criticized implanting young children as a threat to sign language and Deaf communities. More than 90 percent of deaf children are born to hearing parents, who often do not sign. Many families and schools now try a both-and approach, with an implant and sign language together.
Hundreds of thousands helped, millions more who could be
The NIDCD, citing FDA figures reported by manufacturers, says that by December 2019 about 736,900 cochlear implants had been registered worldwide. In the United States, about 118,100 had gone to adults and 65,000 to children. In 2022, Medicare in the U.S. widened its coverage, so more older adults with hearing aids that no longer help enough can qualify.
That still leaves a big gap. When he accepted his Lasker award in 2013, Wilson estimated that as many as 25 million people worldwide could benefit from an implant, and that only 1 or 2 percent of them had one. Cost is a barrier in many countries, mostly the cost of the surgeons, audiologists, and long-term care rather than the device itself.
Hearing loss in general is very common. The World Health Organization (WHO) says more than 430 million people need help for disabling hearing loss, and that over 1 billion young adults are at risk of permanent, avoidable hearing loss from unsafe listening, such as loud music through earbuds or at concerts. Once noise has killed off hair cells, they do not grow back, which is why protecting the ears matters long before an implant ever comes up.
A gene therapy that lets some children skip the implant
Implants work around missing or broken hair cells. A new treatment goes after one specific cause of deafness at its root.
About half of hearing loss present at birth is caused by genes. One of those genes, OTOF, carries the instructions for a protein called otoferlin. Inner hair cells need otoferlin to pass their signal to the nerve. Children with two broken copies of the gene are born profoundly deaf even though their hair cells are there. According to the FDA, OTOF changes account for 2 to 8 percent of inherited hearing loss that is not part of a wider syndrome. Until recently, these children were usually offered a cochlear implant.
The drug company Regeneron developed a gene therapy, first called DB-OTO, that uses a harmless virus to deliver working copies of the gene. A surgeon infuses it directly into the cochlea, in an operation similar to implant surgery. In the CHORD trial, reported in the New England Journal of Medicine, 9 of the first 12 children met the main goal, being able to hear sounds at 70 decibels or softer by 24 weeks, a level that generally avoids the need for an implant. Six could hear soft speech without any device, and three reached normal hearing. The company reported that the first child treated, dosed at 10 months old, later picked out words such as “mommy,” “cookies,” and “airplane” at conversational volume, with no visual cues.
On April 23, 2026, the FDA gave the therapy accelerated approval under the name Otarmeni (lunsotogene parvec-cwha). It is the first gene therapy approved for genetic hearing loss. The FDA based its decision on a trial of 24 children aged 10 months to 16 years, and it said 80 percent of the 20 children who could be evaluated had better hearing. Common side effects included middle ear infection, nausea, dizziness, and pain from the procedure. Regeneron has said it will provide the therapy for free in the U.S.
It has real limits. It only helps people with this one gene problem, whose outer hair cells still work, and who have not already had an implant in that ear. The FDA still wants proof that the gains last and that they lead to better speech and quality of life. For most deaf people, whose hearing loss has other causes, the cochlear implant remains the main tool.
A model for other senses
The cochlear implant has become the model other sensory implants are measured against. Researchers working on vision often point to it. The team behind a tiny wireless eye chip that helped people with macular degeneration read again compared the months of training their patients needed to the way people learn to hear with a cochlear implant. In both cases the device gives the brain a rough new signal, and the brain does much of the work of turning it into something useful.
Volta’s boiling soup was only noise. Two centuries later, a couple of dozen electrodes and some clever timing let hundreds of thousands of people hear voices, and a gene therapy now lets a small group of children hear with their own ears.
Further reading and handy gear
Rebuilt: My Journey Back to the Hearing World — Michael Chorost’s award-winning memoir of getting a cochlear implant in 2001 and learning to hear again through a computer, told with humor and plenty of science.
Loop Experience 2 Earplugs — Reusable high-fidelity earplugs for concerts, sporting events, and other loud places, with about 17 dB of noise reduction and four sizes of silicone tips.