Taylor Spatial Frame

The Taylor Spatial Frame (TSF) is a hexapod external fixator, developed by Dr J. Charles Taylor in Memphis, Tennessee. Two rings are joined by six telescopic struts, and graduated strut adjustments, calculated by computer-assisted planning, correct a deformity in all six axes simultaneously.

The Stewart platform. The mechanism is the Stewart-Gough platform, described by D. Stewart in 1965 for flight simulation: six linear actuators connecting two platforms give six degrees of freedom. The same kinematic mechanism is used in flight simulators, robotic surgery and precision positioning systems, and Taylor applied it to orthopaedic external fixation in the 1990s.

Six degrees of freedom. Three are translational and three rotational:

• Translational - anterior-posterior, medial-lateral, and axial (shortening or lengthening)
• Rotational - varus/valgus (coronal angulation), flexion/extension or recurvatum/procurvatum (sagittal angulation), and axial rotation (torsion)

Against the Ilizarov. The hexapod corrects every axis at once, where an Ilizarov frame needs sequential corrections. Its hinge is virtual, calculated by the software, so no physical hinge is placed and the same construct stays on throughout treatment; a deformity that persists is simply re-prescribed, and computer-assisted planning reduces human calculation errors. The price is cost, in struts and software.

Indications. The frame is used across four groups of problem:

• Deformity correction - complex multiplanar deformity (congenital, developmental or post-traumatic), tibial and femoral malunion, deformity from rickets or Blount disease, and angular deformity with rotational and translational components
• Limb lengthening - limb length discrepancy requiring distraction osteogenesis, lengthening combined with deformity correction, congenital short femur and fibular hemimelia
• Fracture management - acute fracture reduction with gradual correction, malunion correction with osteotomy, and nonunion with deformity
• Infection and bone loss - bone transport for segmental defects, infected nonunion, and use combined with the Masquelet technique

Stacked frames. A multisegmental fracture with separate fracture levels needs independently controlled segments. Double- and triple-stacked TSF constructs place one independently adjustable hexapod across each principal fracture line, with intermediate rings sharing fixation between adjacent segments.

View larger

View larger

Contraindications. The absolute contraindications are:

• Active systemic sepsis
• Soft-tissue compromise severe enough to preclude pin or wire placement
• A non-compliant patient, unable to perform the daily adjustments
• Inadequate bone stock for wire or pin fixation

The relative contraindications are:

• Severe peripheral vascular disease
• Uncontrolled diabetes mellitus
• Osteoporosis, which may require a modified technique
• Morbid obesity, because of concerns about frame stability
• Psychological unsuitability

Inverse kinematics. The software is given the desired end position and orientation of the moving ring and calculates the strut lengths required to reach it. That is inverse kinematics; forward kinematics is the opposite, calculating position from joint angles.

The construct. Every TSF is built from the same parts:

• Two rings, the reference ring and the moving ring
• Six telescopic struts, threaded for length adjustment
• Universal joints at each end of every strut, allowing multi-axis rotation at the strut-ring connections

View larger

Why six struts. Six struts give exactly six degrees of freedom, the minimum required for full spatial positioning. It is the mathematical equivalent of six linear equations with six unknowns: a fully determined system, neither over- nor under-constrained.

Reference and moving rings. The reference ring is the "fixed" ring that defines the coordinate system, and every deformity parameter is measured relative to it. It is usually placed on the proximal (stable) segment, by convention the ring closest to the trunk. The moving ring is attached to the distal (deformed) segment and moves to the corrected position as the struts adjust.

The virtual hinge. The centre of rotation of angulation (CORA) is the intersection of the proximal and distal mechanical axes and the ideal pivot for angular correction. Correction about the CORA produces pure angulation without translation; correction about a point away from it produces combined angulation and translation. An Ilizarov frame gets pure angulation from physical hinges placed at the CORA. The TSF software instead calculates strut adjustments that simulate rotation about a hinge at any specified point, so a virtual hinge at the CORA minimises secondary translation, and it can be repositioned during treatment if needed.

Total residual. The total residual is a composite measure of the deformity remaining across all six axes, calculated by the software once the current deformity parameters are entered. It combines the angular (degrees) and translational (mm) components into a single weighted value that approaches zero as correction completes. It is used to monitor progress, to set the endpoint (typically less than 5 mm equivalent) and to identify incomplete corrections that need adjustment.

Mounting parameters. The software has to know where the frame sits on the bone, and it needs 14 mounting parameters, 7 per ring, to do so:

1. Ring internal diameter, in mm
2. Ring-to-ring distance between the reference and moving rings
3. AP frame offset from the bone axis (anterior positive, posterior negative)
4. Lateral (ML) frame offset from the bone axis (medial positive, lateral negative)
5. Rotational orientation, given by the position of the master tab (anterior, lateral, posterior or medial)
6. Strut positions, the numbered mounting holes for each strut
7. Segment and side: proximal or distal, left or right

They are measured on standardised AP and lateral radiographs and recorded on the mounting parameter worksheet. The offsets are the hard part, and have a section of their own below.

Origin and corresponding point. The surgeon chooses an origin, a defined reference point on the reference (fixed) fragment, conventionally at the osteotomy level on the reference-ring side. The corresponding point is the matching point on the moving fragment, the one that should end up superimposed on the origin once the deformity is corrected. The deformity is then described as the translations and angulations needed to move the corresponding point onto the origin.

What the software computes. It takes two sets of inputs: the mounting parameters, which describe the frame at rest, and the deformity parameters, which are the AP and lateral angulation, the AP, lateral and axial translation, and the axial rotation of the corresponding point relative to the origin. Together these give the current position of the rings relative to each other and the desired final position, and from them the software computes the six daily strut lengths. Placing the origin at the CORA puts the virtual hinge there; an origin away from the CORA is still resolved by the software, but the translation it adds must be intended.

Deformity modes. The software works in three modes:

• Chronic mode - gradual correction of an established deformity
• Residual mode - adjustment when the initial correction is incomplete. The surgeon re-defines the current corresponding-point offset (the residual program) and the software issues a fresh strut schedule, with no frame rebuild, unlike an Ilizarov
• Total residual mode - monitoring the combined deformity

The strut schedule. The output is a daily length change for each of the six struts, scheduled over the correction period, typically weeks to months. The rate can be made faster or slower, and the duration depends on the magnitude of the deformity.

Stacked frames. Each hexapod segment of a stacked construct must be imaged perpendicular to its own reference ring, so the software receives valid deformity and mounting measurements.

View larger

What the offset is. The frame offset is the perpendicular distance from the bone (reference) axis to the centre of the reference ring, in the AP and lateral planes. Together with ring diameter and ring-to-ring distance it locates the ring in space relative to the bone, so the software knows where the frame is.

View larger

Why it dominates the error budget. Offset is the hardest parameter to measure and the most error-prone, and the measurement gets less accurate the further the origin sits from the reference ring (Gessmann, in the evidence cards). The software has been told the bone axis is somewhere it is not, so the error reappears at the end of correction as a residual translation of essentially the same magnitude, plus an angular error of arctan(offset error / ring-to-ring distance).

The "1 mm per degree" rule. The often-quoted 1 mm of offset error per degree of residual angulation is that formula evaluated at a ring-to-ring distance of about 57 mm. It holds for a short construct, and the angular error is roughly half that at 100 mm and a third at 150 mm. Take the message (small offset errors matter, and matter more the shorter the construct) rather than the number.

View larger

Measuring it. Draw the bone axis on true AP and lateral films and measure the perpendicular distance to the ring centre. Use calibrated radiographs (a calibration marker of known size) or a calibrated image intensifier; non-calibrated films are the least accurate.

View larger

View larger

The CT alternative. Basing the correction on the actual 3D frame-to-bone position on CT compensates for mounting error more completely: in phantoms, residual error was less than 2 degrees of rotation and less than 0.5 mm of length. Because of cost and radiation it is reserved for complex cases, and routine CT planning is not universally adopted.

When a multiplanar deformity is identified, the exam question is usually "which device or technique?" Justify the choice from the clinical features below rather than reaching reflexively for a hexapod.

The TSF (Smith & Nephew) is one of several hexapod systems worldwide; equivalent platforms include the TrueLok-Hex (Orthofix), the Ortho-SUV frame, and the MAXFRAME (Zimmer Biomet). All share Stewart-platform kinematics and web/desktop planning software, so the principles below are device-agnostic.

Global epidemiology and case mix:

• Hexapod frames are concentrated in tertiary limb-reconstruction units because of the planning, monitoring and patient-compliance demands.
• Commonest indications globally: post-traumatic tibial malunion/nonunion, congenital and developmental deformity (Blount disease, rickets, fibular hemimelia, congenital short femur), and limb-length discrepancy.
• In high-burden, limited-resource regions, neglected fractures and infected nonunions (often post-bonesetter) form a disproportionate share of the workload.

Societies, training and standards (side-by-side):

There is no single high-level guideline mandating hexapod over Ilizarov; selection is guided by deformity complexity, surgeon experience, software access and cost.

Related pages: Hexapod Frame Deformity Correction is the companion page on the same device concept, and carries the Stewart-Gough platform kinematics and the applications survey in more depth; Ilizarov External Fixation is the parent technique this replaced computationally but not biologically - the wire tensioning, ring construction and distraction-osteogenesis biology are unchanged and live there; Deformity Analysis - CORA and MAD is the non-negotiable prerequisite, because a hexapod executes a prescription and cannot rescue a deformity that was analysed wrongly, and the joint-orientation angles and osteotomy rules on that page are what generate the six deformity parameters entered here; Limb Lengthening Principles for latency, rate and rhythm and for the healing index this page's systematic review reports; Distraction Osteogenesis for the regenerate biology that determines how long the frame stays on; Bone Transport Techniques for segmental defects, where the same frame solves a different problem; Blount Disease and Congenital Pseudarthrosis of the Tibia are the paediatric conditions behind much of the carded evidence, and the recurrence-with-growth problem the short follow-up in those series cannot see; Guided Growth for Angular Deformity Correction is the alternative that should be considered first in a child with growth remaining, because a plate and two screws beat six months in a frame whenever the physis will do the work; Malunion and Delayed Union and Nonunion Management for the post-traumatic deformities that are the commonest adult indication; Ankle Arthritis for the joint the supramalleolar osteotomy carded above is trying to preserve; and Limb Length Discrepancy - Epiphysiodesis for the far simpler answer to a modest discrepancy in a growing child.

Hexapod versus Ilizarov. Hexapods give superior 3D control and easier residual re-prescription, but the comparative data are largely level III-IV single-centre series. Systematic review evidence suggests a potentially higher healing index with hexapods, and no high-level trial proves a union-rate advantage.

Acute versus gradual correction. Gradual six-axis correction is more accurate for complex deformity, yet for a simple uniplanar deformity acute one-stage correction avoids months in a frame. The threshold, often cited around 10 degrees of uniplanar varus, is convention, not high-level evidence.

Latency, rate and rhythm of distraction. The classic latency and rate in the Correction Protocol derive from distraction-osteogenesis biology, but the optimal rate varies with age, site and regenerate quality, and is individualised rather than protocolised.

Pin-site care. Reported infection rates are commonly 30 percent or more, mostly superficial, and no single cleaning regimen is proven superior.