A flat-printed implant self-folds into its final shape once inside the body
The process starts backwards from how surgeons normally think: the device is printed first in the complex geometry it must ultimately perform, not in a shape convenient for insertion.
Patient imaging (CT) is converted into a 3D model of the target anatomy — for example, an infant airway segment. A shape-memory polymer such as polycaprolactone (PCL) or a crosslinked PLA-based copolymer is extruded layer by layer using fused-deposition or micro-extrusion printing, building the splint, patch, or mesh directly in its intended, permanent, functional form.
Because shape-memory polymers "remember" whichever geometry they are in when their crosslinked network is fixed, printing the permanent shape first means every later deformation is temporary and reversible — the network topology, not the momentary form, defines the device's true identity.
Material choice sets the transition temperature, degradation rate, and stiffness. Thermoset PCL networks trigger near body temperature with a sharp melting transition; crosslinked PLA copolymers and polyurethanes offer broader design windows for stiffer load-bearing implants.
Above Ttrans the polymer chains regain mobility; the printed part is deformed — rolled, flattened, or compressed — and then cooled under constraint so the network freezes into this new, temporary configuration.
The device is warmed above Ttrans, mechanically deformed into a low-profile temporary shape, then cooled to below Ttrans while still under load. Cooling vitrifies (or recrystallizes) the network in the new configuration, and the external load can be removed — the compact shape now holds on its own.
Shape-fixity ratio quantifies how much of the applied deformation is actually retained after unloading — typically 90–98% for well-designed SMPs. A high fixity ratio means the compact form stays stable during storage, sterilization, and delivery, with no premature unfolding.
Crucially, programming does not change the permanent network — the crosslinks (chemical or physical) that define the printed geometry are undisturbed. Only the polymer chains between crosslink points are stretched and temporarily frozen, storing elastic strain energy that will later drive spontaneous recovery.
Because the temporary shape is small and flexible, the device can be delivered through a catheter, endoscope, or a much smaller incision than the permanent geometry would ever allow.
A flat or tightly rolled implant can pass through vessels, airways, or trocar ports that its bulky final geometry never could. This is the central clinical payoff of 4D printing: complex, patient-specific final shapes delivered through minimally invasive routes.
Throughout transport, sterilization, and insertion the device must remain below Ttrans so the temporary shape does not prematurely recover. Cold-chain handling or simply keeping the device below body temperature is usually sufficient given typical Ttrans values in the 37–45°C range.
Surgeons position the compact device precisely at the target site — around an airway segment, across a defect, or against a tissue surface — before any warming occurs, since recovery cannot easily be reversed once triggered.
Once in contact with body-temperature tissue (or fluid, for hydration-triggered polymers), the stored elastic strain is released and the network begins pulling itself back toward its printed permanent shape.
Unlike a mechanical stent that a surgeon expands with a balloon, a shape-memory implant unfolds itself — the trigger is simply the ambient warmth or moisture already present in the body, so no batteries, motors, or additional procedures are needed.
As temperature rises through Ttrans, chain segments regain mobility and entropic elasticity drives the network back toward its lowest-energy, permanent conformation. Recovery speed depends on how far above Ttrans the tissue is, polymer thickness, and network crosslink density.
Morrison et al. (Science Translational Medicine, 2015) implanted 3D/4D-printed, bioresorbable PCL splints in infants with severe tracheobronchomalacia. The splints were designed to expand with airway growth and were engineered to gradually resorb over roughly 2–3 years, avoiding a second surgery to remove them — a landmark demonstration that a device could be sized for growth rather than for the day of implantation.
True
The device now sits in the exact functional shape it was originally printed in — supporting, splinting, expanding, or covering tissue exactly as designed, with the delivery step already behind it.
Three numbers governed whether the story above worked: Ttrans placed safely between storage/room temperature and 37°C body temperature; a high shape-fixity ratio so the compact form survived delivery intact; and a high shape-recovery ratio so the final geometry matched the original CAD design closely enough to function correctly.
The same principle now extends to self-expanding cardiac patches that unfurl against the heart wall, self-folding surgical meshes that wrap around organs, and self-deploying microneedle arrays that flatten for transport and spring into an array on skin contact.
For bioresorbable devices, function does not end at recovery — the material continues to hydrolyze slowly over months to years, transferring mechanical load to healing or growing tissue and ultimately disappearing without a retrieval procedure.