3D Printing from Imaging: Surgical Planning

View larger

What it is. Three-dimensional printing is a downstream use of imaging, not a separate imaging modality. The quality of the printed model depends first on acquisition, then segmentation, then the clinical relevance of the model that is produced. Good orthopaedic printing starts with the imaging dataset, not the printer, and with the surgical question, which is defined before anything is segmented.

Where it earns its place. The strongest orthopaedic indications are complex articular trauma, deformity planning, implant templating, and patient-specific education where spatial understanding genuinely changes decision-making.

View larger

The steps run in sequence from the scanner to the operating theatre:

1. Acquire. CT, or MRI for soft-tissue assessment in complex cases, with minimal artefact. The data must be in DICOM format for processing.
2. Transfer. The DICOM data go to the processing software, anonymised and by secure transfer.
3. Segment. The structures of interest are separated: threshold-based bone segmentation, manual refinement for accuracy, then quality-control checks.
4. Build the model. A surface mesh is generated and saved as an STL file. This DICOM-to-STL conversion is the key processing step.
5. Design and plan. Virtual surgical planning, guide design (cutting planes, cutting slots, drill trajectories) and implant templating, with engineering input for instruments and sterilisation considered at the design stage.
6. Print. The printer translates the STL into G-code and builds the construct from its parameters. The material (PLA, nylon, titanium) is chosen for the purpose, and the printer is calibrated and quality-controlled.
7. Post-process. Cleaning, surface finishing for theatre use and validated sterilisation, with regulatory compliance for implants.
8. Use. A pre-operative team briefing with the model, intraoperative guide placement, and post-operative validation.

View larger

Segmentation. Segmentation identifies and separates specific anatomical structures from the imaging data, and its quality directly affects the accuracy of the final print. For bone, thresholding on Hounsfield units, typically above 150-200 HU, can segment cortical bone automatically. Manual refinement is often needed for pathological areas, fracture fragments and tumour boundaries.

View larger

View larger

The review by Wong and colleagues names four core orthopaedic uses of additive manufacturing: anatomical models, patient-specific instruments (PSI), custom implants, and bone and cartilage tissue engineering.

Anatomical ModelsPatient-Specific InstrumentsCustom ImplantsScaffolds and Bioprinting

Pre-operative planning

Benefit

3D visualisation of pathology

Examples

Complex fractures, tumour resection

Template/trialling

Benefit

Pre-bend plates, trial implants

Examples

Pelvic fractures, spine deformity

Patient education

Benefit

Tangible explanation of surgery

Examples

Joint replacement, deformity correction

Surgical training

Benefit

Practice complex procedures

Examples

Resident education, rare procedures

Intraoperative reference

Benefit

Anatomical guide in OR

Examples

Tumour margins, fracture reduction

Anatomical Model Applications

Application	Benefit	Examples
Pre-operative planning	3D visualisation of pathology	Complex fractures, tumour resection
Template/trialling	Pre-bend plates, trial implants	Pelvic fractures, spine deformity
Patient education	Tangible explanation of surgery	Joint replacement, deformity correction
Surgical training	Practice complex procedures	Resident education, rare procedures
Intraoperative reference	Anatomical guide in OR	Tumour margins, fracture reduction

The trauma evidence. In complex acetabular fractures, a comparative study found that pre-operative 3D models significantly reduced operative time, blood loss and fluoroscopy (Ivanov 2022). Boudissa (2021) found shorter operative time and less blood loss with patient-specific virtual planning, in which no physical model is printed. The print cost is often offset by reduced theatre time.

Models are particularly valuable in pelvic and acetabular trauma, complex spine deformity and tumour surgery.

Total knee arthroplasty

PSI Type

Cutting blocks

Benefit

Potentially improved alignment

Total hip arthroplasty

PSI Type

Acetabular guides

Benefit

Cup positioning accuracy

High tibial osteotomy

PSI Type

Cutting/hinge guides

Benefit

Precise correction angle

Tumour resection

PSI Type

Cutting guides

Benefit

Accurate margin achievement

Spinal fusion

PSI Type

Screw trajectory guides

Benefit

Safe pedicle screw placement

Corrective osteotomy

PSI Type

Combined cut/reduction guides

Benefit

3D deformity correction

PSI Applications

Procedure	PSI Type	Benefit
Total knee arthroplasty	Cutting blocks	Potentially improved alignment
Total hip arthroplasty	Acetabular guides	Cup positioning accuracy
High tibial osteotomy	Cutting/hinge guides	Precise correction angle
Tumour resection	Cutting guides	Accurate margin achievement
Spinal fusion	Screw trajectory guides	Safe pedicle screw placement
Corrective osteotomy	Combined cut/reduction guides	3D deformity correction

PSI in knee arthroplasty. Patient-specific cutting guides for TKA are designed from pre-operative CT or MRI. Early studies suggested improved alignment, but recent meta-analyses show variable results: PSI may reduce outliers in mechanical-axis alignment, while overall clinical outcomes are similar to conventional instrumentation. It reduces instrument trays and may streamline theatre workflow.

Why the alignment gain does not reach the patient. Conventional TKA instrumentation already aligns most routine knees, so there is little freehand error left for a guide to remove. Coronal alignment in TKA is also only a surrogate: tightening it changes a number, not a knee.

Clinical Warning

PSI does NOT improve outcomes in routine total knee arthroplasty: the Level I meta-analysis of 43 RCTs found no clinically relevant benefit. Reserve PSI and 3D guides for cases where conventional referencing fails or cannot be used at all (severe extra-articular deformity, retained hardware, an obliterated canal, tumour margins), where they may still help.

View larger

Large bone defects

Implant Type

Custom porous cages

Advantage

Fills defect, allows ingrowth

Tumour reconstruction

Implant Type

Custom prosthesis

Advantage

Matches resection geometry

Revision arthroplasty

Implant Type

Custom acetabular component

Advantage

Addresses bone loss

Limb salvage

Implant Type

Intercalary prosthesis

Advantage

Preserves joint function

Spinal tumour

Implant Type

Custom vertebral body

Advantage

Anatomical reconstruction

Custom Implant Applications

Indication	Implant Type	Advantage
Large bone defects	Custom porous cages	Fills defect, allows ingrowth
Tumour reconstruction	Custom prosthesis	Matches resection geometry
Revision arthroplasty	Custom acetabular component	Addresses bone loss
Limb salvage	Intercalary prosthesis	Preserves joint function
Spinal tumour	Custom vertebral body	Anatomical reconstruction

View larger

Regulation. Custom implantable devices are subject to regulatory oversight worldwide: the FDA in the US, the MDR and notified bodies in Europe, and equivalent national regulators elsewhere. The manufacturing process must be validated, materials must be biocompatible, and sterility must be ensured. Most custom implants are made by specialised companies with regulatory approval, and point-of-care printing of implants requires an accredited quality management system (commonly ISO 13485).

Printed scaffolds. The fourth use is bone and cartilage tissue engineering, with 3D printing as the fabrication route. A biodegradable scaffold with a controlled porous architecture is printed from polymers (such as PCL or PLGA), ceramics (such as tricalcium phosphate or hydroxyapatite) or bioactive glass. It is then seeded with cells (e.g. mesenchymal stem cells) and/or growth factors to regenerate bone or cartilage.

Bioprinting goes further, depositing living cells suspended in a bio-ink, a hydrogel such as alginate, gelatin or collagen, layer by layer to build a living construct.

Where it stands. Tissue engineering remains largely experimental / pre-clinical. The barriers are vascularising large constructs, maintaining cell viability, achieving load-bearing strength and gaining regulatory approval. Its biology belongs to its own topic; what matters here is that 3D printing is the manufacturing method that makes patient-specific scaffolds possible.

View larger

Materials. Material choice follows the application: PLA or ABS plastics for anatomical models, biocompatible resins for surgical guides, and titanium alloys (Ti6Al4V) for implantable devices.

The porous lattice. Metal additive manufacturing, by laser sintering or electron-beam melting, can build a porous or trabecular lattice inside a titanium implant that a solid cast or machined implant cannot, and the porosity can be varied regionally within one part. The lattice is designed to do two things:

• Reduce stress shielding. It lowers the implant's effective elastic modulus toward that of bone (solid titanium around 110 GPa, cortical bone roughly 15-20 GPa), so more load is transferred to the surrounding bone and less is shielded, reducing peri-implant bone resorption.
• Promote osseointegration. The interconnected pores, optimally roughly 300-700 micrometres, allow bone to grow into the implant, giving durable biological fixation on the same principle as a trabecular-metal ingrowth surface.

How far the case goes. Both advantages are engineering rationale supported by bench and animal data, not by clinical comparison. No trial has shown that a printed lattice implant outperforms a conventional one in patients.

View larger

View larger

Accuracy. Studies show printed anatomical models typically achieve dimensional accuracy within 0.5-1 mm of the source CT data, so sub-millimetre fidelity is achievable. Accuracy depends on segmentation quality, printer resolution and post-processing, and degrades with metal artefact. The requirement for implantable devices is a deviation under 1 mm, yet there is no universal validation or verification standard.

Before clinical use. Surgical guides and implants must be verified against the original imaging data. Documentation must give full traceability from imaging to implant, and sterilisation calls for gamma irradiation or autoclave-compatible materials.

View larger

Point-of-care printing. Hospital-based printing allows rapid turnaround and lower costs for anatomical models, and models and guides are increasingly printed in-house. It requires significant infrastructure and expertise, and the in-hospital quality-management burden is high. For implantable devices, regulatory requirements generally necessitate partnership with certified manufacturers.

Examiners often probe whether a candidate can select the appropriate technology rather than reach for a 3D print by reflex. Printing competes with these alternatives for the same clinical problems:

The decision rule. Print when a physical model or a custom cut or drill trajectory genuinely changes the operative plan and there is time to manufacture. For routine arthroplasty alignment, navigation or robotics outperform a static guide because they allow intraoperative correction. For most standard cases, no advanced technology is needed at all.

When a printed guide helps. A guide is a one-shot pre-operative plan with no intraoperative feedback: where the plan may need to change on the table, navigation adapts and a guide does not. It earns its place where the unaided hand has a real error rate to remove, and where accuracy is itself the outcome, as with a pedicle screw, whose breach is the complication. The evidence section reconciles the two Level I meta-analyses on exactly these two conditions.

Related pages: CT Imaging Principles governs the input that determines everything downstream - sub-millimetre slices, bone algorithm, and the artefact from existing metalwork that segmentation cannot recover from; MRI Imaging Principles is the alternative when soft tissue or cartilage must be in the model. Robotic and Navigated Arthroplasty and Navigation and Robotics in Spine Surgery are the competing guidance technologies, and the comparison matters: a printed guide is a one-shot preoperative plan with no intraoperative feedback, while navigation adapts - which is why the pedicle-screw meta-analysis carded here compares templates with freehand rather than with navigation. Apply the same two questions to those pages that this one applies to printing. TKA Alignment Options and TKA Alignment Philosophies explain why alignment accuracy in the knee is a surrogate rather than an outcome, which is the reason patient-specific instrumentation failed there, and The Painful Total Knee Replacement shows what actually drives dissatisfaction. On the applications side, Acetabular Fractures and Both-Column Acetabular Fractures are where printed models have the best trauma evidence, Adolescent Idiopathic Scoliosis is the deformity in which the drill-guide meta-analysis was performed, and Custom Triflange Acetabular Components with Megaprosthesis in Non-Oncologic Salvage and Limb Salvage Surgery Principles are the reconstructions in which a custom printed implant is genuinely the only option. Titanium Alloys and Implant Fracture Biomechanics cover the material and the fatigue behaviour of a printed porous lattice, and Gene Therapy and Tissue Engineering carries the scaffold and bioprinting application - the fourth category, and still the least clinically realised of the four.

Do 3D models help complex trauma? The evidence is mostly small single-centre series with few multicentre RCTs, and benefit to outcome, as opposed to process, is less proven.

Custom tumour and revision implants. They allow reconstruction that is otherwise impossible, but long-term survivorship data are immature, and failure modes and registry capture are incomplete.

View larger

Landmark and Supporting Studies

Key Points

• Trauma planning: Comparative evidence supports reduced operative time, blood loss, and fluoroscopy with pre-operative 3D models in complex acetabular fractures.
• Deformity guides: Meta-analytic Level I evidence shows 3D-printed drill guides improve pedicle screw accuracy and lower radiation exposure.
• PSI in routine TKA: High-quality Level I evidence shows no clinically relevant outcome benefit — value is confined to selected, atypical anatomy.
• Custom implants: Geometry and stiffness matching are real advantages, offset by cost, multi-week lead time, and loss of intraoperative flexibility.

Reconciling the two Level I findings - the question an examiner will ask. This page carries two meta-analyses of the same technology reaching opposite conclusions: 3D-printed drill guides clearly improve pedicle screw accuracy in paediatric scoliosis (OR 2.96), while patient-specific instrumentation confers no clinically relevant benefit in routine primary TKA across 43 randomised trials. The technology did not change; the task did. A patient-specific guide helps only where BOTH conditions hold.

1. The unaided task must be unreliable. Freehand pedicle screw placement in a rotated, dysplastic adolescent spine has a real and consequential error rate. Conventional TKA instrumentation, by contrast, already delivers acceptable alignment in the large majority of routine knees - there is little error left for a guide to remove.
2. The accuracy gained must BE the outcome, not a surrogate for it. A malpositioned pedicle screw is itself the complication - breach, neurological injury, revision. Coronal alignment in TKA is a surrogate: the patient-reported result is dominated by soft-tissue balance, rotational positioning and expectation, so tightening alignment by a degree or two changes a number without changing a knee.

Apply the same two questions to any new planning or guidance technology - navigation, robotics, augmented reality. Where the unaided task is already reliable, or where the thing being made more accurate is only a surrogate, the technology will measure better and the patient will not notice.