Many orthopaedic procedures require implants to be trialled before definitive implantation. Where this is required, the trials are provided in a set with the instrumentation. The most common scenario this is seen in during elective joint replacements. In Scotland (2007) the Scottish Executive (http://www.sehd.scot.nhs.uk/cmo/CMO(2006)13.pdf) recommended and implemented individually packed orthopaedic implants for all orthopaedic sets. The premise for this was to reduce the risk of CJD contamination and fatigue of implants due to constant reprocessing from corrosion. During many trauma procedures determining the correct length of plate or size of implant can be challenging. Trials of trauma implants is no longer common place. Many implants are stored in closed and sealed boxes, preventing the surgeon looking at the implant prior to opening and contaminating the device. As a result many implants are incorrectly opened and either need reprocessed or destroyed due to infection control policy, thus implicating a cost to the NHS. With even the simplest implants costing several hundreds of pounds, this cost is a very significant waste in resources that could be deployed else where. My project was to develop a method to produce in department accurate, cheap and disposable trials for implants often used in trauma, where the original manufacturer do not offer the option of a trial off the shelf. The process had to not involve contaminating or destroying the original implant in the production of a trial.

Several implants which are commonly used within Glasgow Royal Infirmary and do not have trials were identified. These implants were then CT scanned within their sealed and sterile packaging without contamination. Digital 3D surface renders of the models were created using free open source software (OsiriX, MeshLab, NetFabb). These models were then processed in to a suitable format for 3D printing using laser sintering via a cloud 3D printing bureau (Shapeways.com). The implants were produced in polyamide PA220 material or in 316L stainless steel. These materials could be serialized using gamma irradiation or ethylene oxide gas. The steel models were suitable for autoclaving in the local CSSU.

The implants produced were accurate facsimiles of the original implant with dimensions within 0.7mm. The implants were cost effective, an example being a rim mesh was reproduced in polyamide PA220 plastic for £3.50 and in 316L stainless steel for £15. The models were produced within 10 days of scanning. The stainless steel trials were durable and suitable for reprocessing and resterilisation.

The production of durable, low cost and functional implant trials all completed in department was successful. The cost of production of each implant is so low that it would be offset if just one incorrect implant was opened during a single procedure. With some of the implants tested, the trials would have paid for themselves 100 times. This is a simple and cost saving technique that would help reduce department funding and aid patient care.

3D printing and rapid prototyping in surgery is an expanding technology. It is often used for preoperative planning, procedure rehearsal and patient education. There have been recent advances in orthopaedic surgery for the development of patient specific guides and jigs. The logical next step as the technology advances is the production of custom orthopaedic implants.

I aimed to use freely available open source software and online cloud 3D printing services to produce a patient specific orthopaedic implant without requiring the input of a university department, specialised equipment or implant companies.

Using standard CT scan DICOM data, a 3D surface reconstruction was made of a patient's uninjured radial head using open source DICOM viewer OsiriX. This was then manipulated in other open source software packages called Meshlabs and Netfabb to create a mirror image 3D model of the radial head with a stem to produce a prosthesis suitable to replace the contralateral fractured radial head. This was then uploaded and printed in stainless steel via cloud printing service Shapeways.com.

The model produced was an exact replication of the patient's original anatomy, except a mirror image suitable for replacement of the contralateral side. The process did not involve any specialist equipment or input from an academic department or implant company. It took a total of 10 days to produce and cost less than £40.

From this study I was able to show that production of patient specific orthopaedic implants is possible. It also highlights that the technology is accessible to all, and does not require any special equipment or large investment. It can be achieved quickly and for a very small financial outlay. As a proof of concept it has been very successful.

3D printing an additive manufacturing technique, allowing for rapid prototyping in many industries. To date, medical applications have generally been within a research or industry environment, as the costs (expertise, software and equipment) have been prohibitive.

We have established a means by which 3D printing of bones can be achieved quickly, cost-effectively and accurately from standard computer tomography (CT) digital imaging and communications in medicine (DICOM) data.

CT DICOM data of a malunited forearm fracture were manipulated using open-source software (no cost) and a 3D model was produced by selective-laser-sintering. The entire process took 7 days (total cost £77). This process and the resultant model were then assessed for overall accuracy.

This sequential methodology provides ready and economical access to a technology that is valuable for preoperative templating/rehearsal in complex 3D reconstructive cases.

Rapid prototyping (RP), especially useful in surgical specialities involving critical three-dimensional relationships, has recently become cheaper to access both in terms of file processing and commercially available printing resources.

One potential problem has been the accuracy of models generated. We performed computed tomography on a cadaveric human patella followed by data conversion using open source software through to selective-laser-sintering of a polyamide model, to allow comparative morphometric measurements (bone v. model) using vernier calipers. Statistical testing was with Student's t-test.

No significant differences in the dimensional measurements could be demonstrated. These data provide us with optimism as to the accuracy of the technology, and the feasibility of using RP cheaply to generate appropriate models for operative rehearsal of intricate orthopaedic procedures.