Robotic Surgery Planning

For Medical Students

What It Shows

Robotic planning turns a computed tomography study into a three dimensional model of the patient's bones, and the plan is made on that model. 5.2.25 covers the 3D reconstruction.

The model is produced by segmentation, which decides where the system believes the bone surface is. 5.2.23 covers the reconstruction.

It matters because the model becomes the reference the system works to, so an error in the model becomes an error in the operation. 5.16.8 covers the 3D reconstructions.

How It Is Acquired

Acquire the study to the system's protocol, because the segmentation was validated on it. 5.2.22 covers the CT acquisition.

Check the segmentation against the source slices rather than accepting the model. 5.2.23 covers the reconstruction.

Check the landmarks the plan depends on, since those are where a segmentation error matters. 2.1.8 covers the imaging anatomy.

Check the alignment against the standing radiograph, which a supine scan cannot give. 5.3.6 covers the long leg alignment films.

Record the dose, since this is a planning study added to the usual radiographs. 5.2.24 covers the radiation dose.

What Normal Looks Like

A segmentation whose surface follows the cortex on the source slices. 5.2.23 covers the reconstruction.

Landmarks identifiable both on the model and on the slices. 2.1.8 covers the imaging anatomy.

An implant plan consistent with the standing alignment. 5.3.7 covers the mechanical axis.

Osteophytes represented as osteophytes rather than as bone surface. 5.3.18 covers osteophytes.

What Abnormal Looks Like

A segmentation that includes an osteophyte as native surface. 5.3.18 covers osteophytes.

A surface that smooths over a defect the slices show. 5.3.19 covers subchondral sclerosis.

Artefact from a previous implant carried into the model as anatomy. 5.2.19 covers the artefacts.

A plan whose alignment target disagrees with the standing film. 5.3.6 covers the long leg alignment films.

A model accepted with no reference to the slices it came from. 5.16.8 covers the 3D reconstructions.

MRCS Level

How It Is Acquired

In a viva the expected answer treats the model as an interpretation.

The expected points are the system protocol, the segmentation check, the landmarks, the standing alignment, and the dose.

Say that the model is derived data rather than the study. 5.2.23 covers the reconstruction.

Say that the plan is made supine for an operation judged in stance. 5.18.5 covers weight-bearing MRI.

Say that the soft tissue balance remains intraoperative. 5.5.9 covers the ligaments.

What Abnormal Looks Like

The reasoning expected is that segmentation uses a density threshold, so a structure near that density is included or excluded by the threshold rather than by anatomy. 5.2.23 covers the reconstruction.

An osteophyte has the density of bone and is anatomically not the surface an implant should sit on, which a threshold cannot distinguish. 5.3.18 covers osteophytes.

Sclerotic bone and a cystic defect sit on either side of the threshold, so the surface near them is where the model is least reliable. 5.3.19 covers subchondral sclerosis.

Metal artefact raises and lowers density around an implant, which the threshold reads as bone and as the absence of bone. 5.2.19 covers the artefacts.

How Good It Is

A plan built from computed tomography is the most precise geometric plan available and is precise about a model rather than about the patient.

Its best contribution is reproducible implant positioning relative to the planned references. 5.16.6 covers arthroplasty planning.

It is limited by segmentation accuracy, by supine acquisition, by the absence of soft tissue information, and by the dose it adds. 5.2.24 covers the radiation dose.

No figures appear for accuracy on this page, under decision 028, and no segmentation tolerances, cut accuracies or dose values appear either: those are published quantities tied to particular systems and this environment cannot check them.

FRCS Level

How Good It Is

At decision level the plan is a document that will be executed, so reviewing it is part of the operation rather than preparation for it. 5.20.1 covers structured reporting.

The landmarks and the alignment target are the two things worth checking against the source. 5.3.7 covers the mechanical axis.

Bone loss and previous implants are where the model is least trustworthy. 5.17.9 covers the revision arthroplasty.

The standing radiograph remains the alignment reference, because the plan is supine. 5.3.6 covers the long leg alignment films.

Intraoperative registration can correct a model error only if the error is noticed. 5.16.6 covers arthroplasty planning.

Where It Misleads

The commonest error is reviewing the model and not the slices. 5.16.8 covers the 3D reconstructions.

A smooth rendering carries an authority the underlying data does not support. 5.2.25 covers the 3D reconstruction.

A plan's numerical precision is read as accuracy against the patient. 5.1.6 covers imaging pitfalls.

Osteophytes included in the model shift the planned resection. 5.3.18 covers osteophytes.

The absence of soft tissue from the plan is read as the soft tissue having been accounted for. 5.5.9 covers the ligaments.

Fellowship Level

Where It Misleads

At fellowship level the plan arrives already made, and the surgeon reviewing it did not acquire the scan or check the segmentation. 5.20.7 covers reporting the post-operative knee.

Each system segments and references differently, so a plan is not portable between systems. 5.16.6 covers arthroplasty planning.

A revision case is where the model is least reliable and where the plan is most relied on. 5.17.9 covers the revision arthroplasty.

The planning dose is additional and recurs across a pathway, which weighs more in younger patients. 5.2.24 covers the radiation dose.

Automated segmentation is improving, and its failures are now the quiet ones rather than the obvious ones. 5.18.8 covers artificial intelligence in imaging.

The honest position is that the precision of these plans is real and is precision about geometry, while the operation is judged on balance and function that none of this measures. 5.5.9 covers the ligaments.