What “smart” means here
There is no formal definition of a smart implant. The phrase is used for implants that do more than sit passively in the body. Some contain sensors that measure something, some contain electronics that act on the body, some can be adjusted or updated from outside, some send data to a clinician, and some are simply shaped to match one patient’s anatomy. This article follows these lines in turn, which began long before the word “smart” became fashionable.
The pacemaker: the first active implant
The cardiac pacemaker is the standard starting point. In 1958 in Stockholm, the engineer Rune Elmqvist and the surgeon Åke Senning placed the first fully implanted pacing device in a patient, Arne Larsson. The first unit reportedly failed within hours and was replaced, and Larsson went on to receive many more devices over his life. In the United States around 1960, William Chardack and Wilson Greatbatch developed an implantable pacemaker that was soon used in patients.
Early pacemakers delivered pulses at a fixed rate, and they were limited by their batteries. Progress in cardiac pacing and in pacemaker batteries drove what followed: from short-lived cells to longer-lasting lithium batteries in the 1970s. Programmability followed in the 1970s, as devices began to use radio telemetry to receive new settings through the skin, so that a clinician could adjust the rate or pulse strength without surgery. Later devices could sense the heart’s own rhythm and pace only when needed, and some could respond to body motion by raising the rate during activity.
The same period saw the arrival of the implantable cardioverter-defibrillator, which senses dangerous rhythms and delivers a shock, and a broader class of implantable devices with sensing and therapy built in. The point is that sensing, deciding and acting inside the body became standard in cardiology.
Neurostimulators and other electronic implants
From the 1960s onward, implanted stimulators spread to other conditions. Spinal cord stimulators were used for chronic pain, and deep brain stimulators for movement disorders were approved for tremor in the late 1990s. Cochlear implants were another branch of the family. Each used the pacemaker’s template: a sealed case, a battery, a pulse generator, and a lead running to the target tissue, with settings adjusted through an external programmer.
Newer neurostimulators can sense signals as well as deliver them. Some record electrical activity and change stimulation accordingly, an approach called closed-loop stimulation. Research on brain-computer interfaces is a more experimental extension of the same idea, in which electrodes read from the nervous system and a computer interprets the signals.
Sensors in joint and bone implants
A different strand of work built sensors into orthopedic implants. Since the 1960s and 1970s, researchers have fitted hip and later knee implants with strain gauges to measure forces inside the joint, using wires or radio links to send data out. Most of this work was done on small numbers of volunteers for research on loads during walking, stairs and exercise, and it taught surgeons a great deal about how joints behave. Commercial implants with built-in sensors for routine monitoring are a more recent development, and their value for patients remains under study.
Customizing the shape
The second meaning of “customizable” concerns shape rather than electronics. Standard implants come in a range of sizes, which can be a poor match for some patients, such as those with unusual anatomy or large defects after trauma or tumor removal.
The change began when computed tomography scans made three-dimensional models of bone possible. Since roughly the 1990s, surgeons and engineers have used scan data to design plates and implants for skull, jaw and facial defects, at first by machining and casting. With additive manufacturing, and in particular metal powder printing, implants could be built directly from the design, in titanium or other approved materials, including porous surfaces that bone can grow into. Patient-specific cranial, jaw, pelvic and spinal implants are now made this way at some centers. The same digital pipeline is used for surgical guides that help place standard implants.
Custom implants cost more and take time to design, and evidence for whether they improve outcomes is still being gathered, especially compared with well-designed standard parts.
Connected implants and new risks
By the 2000s, many pacemakers and defibrillators could send data wirelessly to a bedside unit and then over the internet to the clinic, so that doctors could review the device’s performance without a visit. This remote monitoring is a feature of what people now call connected sensors, applied to implants.
Connectivity also brought a new class of risk: security. Researchers demonstrated, in laboratory settings in the late 2000s and after, that some wireless implant interfaces could be attacked. Regulators began to expect manufacturers to address cybersecurity. In 2017, a firmware update was recommended for hundreds of thousands of pacemakers from one maker after vulnerabilities were identified, an event that showed how software had become part of the implant.
Other issues include battery life, data privacy, who owns the data, and what happens to a patient when a company stops supporting its software. As with retinal implants, discontinued support for implanted electronics has already caused hardship for some users.
Where it is heading
Current research explores implants that deliver drugs on demand, that monitor wound healing, or that dissolve after doing their job, as well as ones powered by body motion or heat instead of batteries. Some will no doubt reach patients, while others will remain laboratory ideas. History suggests that progress tends to come from slow improvements in materials, power and reliability, more than from a single design leap, and that every step in adding intelligence to an implant creates new questions about safety, maintenance and consent.