The transducer needs to deliver vibration to the skull with as little loss as possible. Three coupling strategies exist, each with distinct surgical, acoustic, and patient-experience tradeoffs.
Percutaneous bone-anchored hearing aid (BAHA)
The original BAHA, commercialised by Cochlear Bone Anchored Solutions (originally Entific) starting from a 1977 Tjellström clinical implementation in Sweden, is the percutaneous standard. The surgical procedure:
A titanium implant (the fixture) is osseointegrated into the temporal bone behind the ear via a minor outpatient procedure. The fixture is ~4 mm long and 3–5 mm in diameter; it threads into a drilled pilot hole in the cortical bone of the temporal bone.
An abutment screws into the fixture and protrudes through the skin and soft tissue, exposing a connector externally.
The sound processor (containing microphones, DSP, and the vibrator transducer) snaps onto the abutment via a magnetic or snap connector.
The advantages of the percutaneous approach:
Direct bone coupling. The vibrator drives the abutment, which drives the fixture, which drives the bone — no soft tissue in between. The acoustic transmission has minimal frequency-dependent losses.
Maximum output. Percutaneous BAHAs achieve about 70–75 dB HL of equivalent output, the highest of any BCD class. Useful for moderate-mixed losses where high cochlear-reserve thresholds need significant compensation.
Low maintenance. The skin around the abutment requires daily cleaning to prevent peri-abutment infection but otherwise the device is durable.
The disadvantages:
Persistent skin penetration. The abutment is a permanent through-the-skin hardware. Peri-abutment infection is the most common complication, occurring in 10–30% of patients at some point. Soft-tissue overgrowth is also common.
Aesthetic / acceptance. Some patients (especially adolescents) resist the visible hardware.
Surgical risk. Although the procedure is minor, it does involve drilling into bone and the small inherent risks of any surgery.
A passive transcutaneous BCD avoids the percutaneous abutment. The internal hardware:
A titanium fixture in bone, plus
An implanted magnet directly under the fixture.
The external processor contains the transducer plus a magnetic coupling plate that holds the processor against the skin overlying the magnet. The vibration crosses the intact skin from processor to internal magnet (which is fused to the implanted fixture), then through the fixture into bone.
The skin in between is the new failure mode. Skin tissue is acoustically lossy at high frequencies — passive transcutaneous devices show 5–15 dB of attenuation above ~2 kHz compared to equivalent percutaneous devices on the same patient. The maximum effective output is therefore lower. They are clinically suitable for patients with smaller air-bone gaps (mild-to-moderate conductive or mixed losses) where the lower output is sufficient.
Advantages: no through-skin hardware, lower infection rates, more acceptable cosmesis. Disadvantages: lower output, higher cost, and pressure-related skin issues from the holding magnet (~5% of patients experience persistent erythema or breakdown).
Active transcutaneous (Osia, Bonebridge)
The newest BCD class moves the transducer itself under the skin. The implanted hardware:
A small piezo-electric or electromagnetic transducer bonded to the bone.
A receiver coil and electronics package adjacent to the transducer.
An external processor (microphones, DSP, transmitting coil) held magnetically against the skin.
Power and signal flow through the skin via the magnetic coupling, just as in a cochlear-implant external processor. The transducer is directly bonded to the bone — no skin in the vibration path. This combines the low-loss bone coupling of the percutaneous BAHA with the cosmetic and infection advantages of the transcutaneous approach.
Two clinically prominent devices:
Cochlear Osia (launched 2019, FDA-approved 2020): piezo-electric transducer mounted on the temporal bone via a screw fixation, with the Osia 2 processor providing 75 dB output and DNN-based noise reduction.
MED-EL Bonebridge (launched 2012): electromagnetic transducer (Bone Conduction Implant, BCI) placed in a small bone bed drilled into the temporal bone. Lower maximum output than Osia (~65 dB HL) but smaller surgical footprint.
Disadvantages: more invasive surgery than transcutaneous, higher cost than percutaneous BAHA, smaller installed base means fewer audiologists fluent in fitting (though this is rapidly changing).
Softband and Sound-Arc options
For pediatric and pre-surgical patients, BCDs can be worn on a softband (an elastic headband) that holds the processor against the mastoid without any surgical fixation. The softband is the standard pediatric BCD configuration for infants and young children up to age 5, after which most clinical centres proceed with surgical implantation (BAHA, Attract, or Osia).
A more recent variant is the Cochlear Sound-Arc, a flexible plastic frame that wraps under the chin, with the processor held against the mastoid. The Sound-Arc gives slightly better acoustic coupling than the elastic softband and is often preferred for school-age children.
The softband / Sound-Arc options have ~5–15 dB lower output than implanted BCDs (the soft-tissue path attenuates), but they enable non-surgical trials and bridge the pediatric pre-implantation period. Most centres trial a softband for 6–12 months before recommending surgical implantation.
Power and noise-reduction processing
A modern BCD processor is essentially a hearing-aid DSP pipeline (Ch 7) with a transducer output instead of an air-conduction receiver. The same WDRC, directional microphone, noise-reduction, and feedback-cancellation algorithms apply. The fitting workflow is similar to a conventional hearing aid: take audiometric thresholds, apply a prescription algorithm (BCD prescription algorithms exist — they differ from NAL-NL2 / DSL because the bone-conduction route bypasses the air-conduction filter — but the structure is similar), verify with skull-simulator or in-situ measurement (analogous to REM in air-conduction fitting).
Bluetooth audio streaming, smart-phone apps, and accessory remote microphones are all standard in modern BCD processors as of 2024–2026. Power consumption is somewhat higher than a comparable hearing aid because the transducer is less efficient than an air-conduction receiver — typical battery life is 3–5 days vs 5–7 for a conventional hearing aid.
Pertti-Erkki Tjellström, working in Göteborg in the 1970s, was a pioneer of dental-implant osseointegration — the discovery that pure titanium implants can be made to bond directly to bone tissue without an intervening fibrous capsule. The clinical opportunity was straightforward: dental implants were already in routine use. Tjellström’s insight was that the same principle could be applied to acoustic coupling — a titanium implant osseointegrated into the temporal bone could provide a low-loss vibration pathway to the cochlea.
The first BAHA implantation was performed in 1977 in Sweden. The early devices were body-worn (the processor was too large to fit behind the ear). Miniaturisation through the 1980s and 1990s produced behind-the-ear BAHA processors; CE marking in 1995 and FDA clearance in 1996 opened the device to wider clinical use.
The 2000s and 2010s saw progressive improvements: smaller fixtures (4 mm to 3 mm), reduced surgical trauma (single-stage punch technique), improved abutments (longer abutments accommodate thicker scalps without requiring tissue reduction), and transcutaneous variants (Sophono 2009, BAHA Attract 2013, Osia 2019).
The shift in the 2020s has been from “BAHA + alternatives” to “Osia / Bonebridge + percutaneous fallback.” Active transcutaneous devices have become the first-choice option in most contemporary centres because they preserve the acoustic advantages of percutaneous coupling without the through-skin hardware. The percutaneous BAHA remains the first choice when maximum output is needed (severe mixed losses, large air-bone gaps) — but the indications are narrowing.
Next lesson: the candidacy decision tree, the four clinical populations served by BCDs, and a closing reflection on the book.