Surgeon Implants World's First 3D-Printed Ossicles, Restoring Patient's Hearing

The three smallest bones in the human body occupy a chamber smaller than a fingertip. When they fail, so does hearing. Professor Mashudu Tshifularo and his surgical team at Steve Biko Academic Hospital in Pretoria walked into an operating theatre on 13 March 2019 carrying custom-printed titanium replacements for all three, and walked out having restored hearing to a 40-year-old man named Thabo Moshiliwa.

It was the world's first full ossicular chain replacement using additive manufacturing. The hammer, anvil, and stirrup, the three bones that transmit vibrations from the eardrum to the cochlea, were printed in biocompatible titanium and implanted via endoscope in under two hours. Moshiliwa had lost his hearing following a middle-ear injury. For a procedure that permanently changed a man's relationship with sound, two hours is fast.

Tshifularo, head of the Department of Otorhinolaryngology at the University of Pretoria's Faculty of Health Sciences, developed the technique during his PhD studies. The workflow starts with a CT scan of the patient's ear, which generates an exact digital model of the damaged anatomy. From there, replacement ossicles are designed to match the patient's individual geometry. "Get the same size bone, position, shape, weight and length and put it exactly where it needs to be, almost like a hip replacement," Tshifularo said. That anatomical match is more demanding than it sounds.

A University of Maryland study benchmarking custom 3D-printed ossicular prostheses found the odds of a surgeon correctly pairing a prosthesis to its intended temporal bone by chance alone were 1 in 1,296. Every ossicle is anatomically unique. The CT-to-CAD-to-print pipeline has to produce a geometrically accurate titanium part in a space measured in millimeters, for surgery measured in minutes.

That is the detail that should stop any desktop printer owner for a moment. The stirrup, the smallest of the three ossicles, is roughly 3mm long. Metal printing at that scale with the tolerances required for surgical clearance means direct metal laser sintering: equipment costing hundreds of thousands of dollars, operating in controlled environments, and passing through extensive post-processing and biocompatibility validation under ISO 10993 standards. That is not a Bambu or a Prusa job, at least not yet.

Parts of Tshifularo's pipeline are closer to the hobbyist bench than they might appear, however. The scan-to-CAD workflow, capturing real anatomy digitally and turning it into a printable model, is the same concept behind consumer photogrammetry apps, smartphone LIDAR, and affordable structured-light scanners that now sell for a few hundred dollars. High-resolution MSLA resin printers available under $500 are hitting XY resolutions below 0.05mm. Custom hearing aid shells, prosthetic finger joints, and surgical planning models are already being produced with consumer-grade resin. The gap between desktop capability and clinical application is real; it is also narrowing.

What this milestone makes unmistakably clear is that printing a functional medical implant is a categorically different discipline from printing a functional prototype, and that distinction has practical lessons for any maker working on precision parts. Tolerance your design to your actual machine output, not the spec sheet. The University of Maryland study showed that CT scan resolution and print accuracy must work in harmony; if one link in that chain drifts, the fit fails. Characterize your printer's real dimensional accuracy with test parts before committing to a precision final build. Then validate with physical test articles first, the way Tshifularo's team used printed anatomical models for pre-surgical rehearsal, rather than discovering a misfit during the critical run.

The hardest boundary is biocompatibility. Implant-grade titanium carries decades of clinical validation. Standard FDM filament does not, and even dental-grade photopolymer resins require specific printer configurations, post-cure protocols, and ISO 10993 testing before any meaningful body contact. Biocompatibility testing is performed according to recognized standards and methodologies that define specimen preparation, conditioning, exposure conditions, and evaluation criteria. Sterilization compatibility is a separate requirement on top of that.

"3D technology is allowing us to do things we never thought we could," Tshifularo said. From a printing standpoint, what his team built in Pretoria is a precision benchmark: the standard the full pipeline has to reach when a human body is holding the part.