Smart implants, microsensors and bioresorbable materials

What role will integrated micro sensors play in the next generation of smart implants - is it primarily about monitoring, decision support or autonomous therapy adjustment in real time?

Dr. Harald Unterweger: The current focus is clearly on monitoring and decision support. These are also the systems that are currently moving from research towards clinical application. Sensors are used to objectively record healing processes or to monitor the condition of an implant over time.
One specific example is implantable pressure sensors that are attached to osteosynthetic metal implants. They continuously measure the mechanical load in the area of a bone fracture. Conclusions about bone healing can be drawn from the time course of this measurement data - for example, whether the bone is already stabilizing or whether there is a risk of the implant failing. On this basis, doctors can make a more informed decision as to whether readjustment or even a second operation is necessary. Such systems are already in clinical trials and have been used in humans for the first time.
There are similar developments in sensors for the continuous measurement of intraocular pressure, for example. In contrast, I see autonomous systems with sensor-actuator coupling, i.e. implants that intervene therapeutically on their own, as being much further in the future. In addition to technical feasibility, ethical, regulatory and safety-related issues play a key role here.

How are the technical limitations of today's sensor technology - keywords drift, noise and long-term stability - addressed in order to enable reliable long-term measurements in the body?

Unterweger: Basically, measurement technology always works with reference standards, internal or external, which serve as zero or calibration points. Drift effects and interference can be mathematically compensated for or filtered out using such references.
A distinction must be made between the type of interference present. Periodic signals such as a heartbeat can be treated differently to chaotic movements, such as those caused by physical activity. Accordingly, high and low pass filters, threshold values or other signal processing algorithms are used. Which method makes sense depends heavily on what is being measured and how strong the actual signal is in relation to the noise.
The mechanical structure of the sensor also plays a major role. Sensor elements can be encapsulated, mechanically decoupled or protected from external influences by damping structures. Long-term drift can also be compensated for via self-calibration or external calibration pulses, for example via contactless interfaces such as NFC. Ultimately, the specific application is always decisive.

Bioresorbable implants based on magnesium or zinc are regarded as game changers. Which applications are conceivable as the first realistic mass applications?

Unterweger: Bioresorbable implants are no longer just a vision of the future. There are already screws and fixation elements made of magnesium that are being used clinically. The basic idea is that the material degrades over weeks or months and a second operation for explantation is not necessary.
However, this is still difficult for highly stressed applications - for example in the area of the upper or lower leg bones. The reason is simple: a resorbable material loses its mechanical stability over time. If it is subjected to too much stress too soon, there is a risk of mechanical failure. In such cases, classic metal implants are still necessary and have to be removed again later.
In contrast, temporary applications, such as bioresorbable stents, are very promising. These only keep a vessel open for as long as is medically necessary and then dissolve in a controlled manner. Such concepts are already in use, but require very precise control of the degradation rate in order to avoid risks such as particle formation or vascular occlusion.

What role do generative or algorithmic design and additive manufacturing play in the targeted control of the degradation rate and mechanics of bioresorbable implants?

Unterweger: The mechanics and degradation rate can be influenced by several variables: Material composition, microstructure, surface morphology and geometry. For example, composite materials can be used in which different components degrade at different rates. Roughness, porosity or the crystallographic orientation of metals also have a significant influence on degradation behavior.
The underlying mechanisms are essentially known, so a generative approach is not necessarily required to develop functioning designs. However, the system quickly becomes complex because any change to one property can have an impact on other properties - such as mechanical stability or biocompatibility.
AI-supported optimization approaches can be useful here to resolve conflicting goals and find an optimum. In combination with additive manufacturing, this opens up additional degrees of freedom in design.

Energy supply is considered the bottleneck of smart implants. Where do you see the future between batteries, energy harvesting, inductive coupling or ultrasonic energy transmission?

Unterweger: The energy supply is actually one of the limiting factors. For systems with actuators, such as pumps or mechanical components, the energy requirement is considerable. Classic implants such as pacemakers therefore continue to rely on batteries, which are accessible with a relatively simple surgical procedure and can be replaced if necessary.
Passive systems are also possible with pure sensor technology, for example via NFC. The sensor is then only read out when energy is coupled in from outside. Wireless energy transmission is conceivable in principle, but brings with it new security challenges. Every external coupling is also potentially an attack vector.
The trend is therefore clearly towards extremely energy-efficient systems and the further development of miniaturized solid-state batteries. For highly invasive systems, the energy supply will continue to be outsourced to easily accessible locations.

How far along are energy self-sufficient implants that use body movement or other physiological gradients?

Unterweger: I think this is realistic for applications with very low energy requirements. It is certainly conceivable that the movement of the body or electrical signals could be used to continuously supply small amounts of energy. There is intensive research into so-called energy harvesting systems in the context of implants. These include the following types

• mechanical: piezoelectric, triboelectric (movement, vibration)
• thermal: temperature gradient body ↔ environment
• electrical/biochemical: biofuel cells (glucose, oxygen)
• electrophysiological: use of existing field changes (very limited)

But all these system approaches can only cover smaller energy requirements. This is not enough for energy-intensive applications. It is more of a supplement than a complete solution.

Many smart implant concepts rely on AI evaluation of large amounts of data. Does the intelligence take place in the implant itself or in the cloud?

Unterweger: For pure sensor technology, downstream evaluation makes perfect sense, for example to better understand disease progression or optimize therapies. However, I consider a direct cloud connection of implanted systems to be problematic. Many ethical and security issues have not yet been clarified.
Multi-level architectures in which implants do not communicate directly with the cloud make more sense. Cybersecurity concepts must be an integral part of such systems - and are currently still in their infancy.

Finally, a look into the future: how will the everyday life of a patient with a chronic illness change in five years' time?

Unterweger: The changes will be gradual rather than disruptive. Innovations in medical technology take time - also because questions of financing and reimbursement need to be clarified. In the next five to ten years, I expect to see progress on the sensor side in particular: more continuous measurements, better progress data and closer monitoring of therapies.
Therapeutic actuators will follow, but more slowly. The path leads via small, well-proven steps - and that is precisely the right approach in my view.