Future of Knee Imaging

For Medical Students

What It Shows

Three things are changing: the sequences are getting more quantitative, the reading is being assisted by software, and the imaging is moving towards positions and loads that resemble use. 5.18.1 covers quantitative imaging.

Each addresses a real limitation of current practice, and none of them is settled in clinical use. 5.1.7 covers evidence based imaging.

A clinician's useful position is knowing what each would change if it worked.

How It Is Acquired

Quantitative sequences measure a tissue property rather than showing a picture, which turns an appearance into a number. 5.18.2 covers T2 mapping.

Weightbearing and dynamic acquisition image the knee under load or in motion, which addresses the oldest complaint about cross sectional imaging. 5.18.5 covers weight bearing imaging.

Short echo time techniques reach tissues that conventional sequences cannot show, including the deep cartilage and the tendon insertions. 5.18.4 covers ultra short echo time imaging.

Automated reading applies trained software to the images, either to triage or to measure. 5.18.8 covers artificial intelligence in imaging.

The magic angle work sits across several of these, since it is a physical effect that quantitative sequences have to account for and can also exploit. 5.2.20 covers the magic angle effect.

MRCS Level

How It Is Acquired

In a viva this is the final question and the expected answer names a technique and what it would change.

The expected points are the quantitative sequences, the loaded and dynamic acquisitions, the short echo time techniques, the automated reading, and the status of each.

Say that each is at a different stage, and do not present research techniques as available. 5.18.1 covers quantitative imaging.

Say what the technique would change clinically rather than how it works. 5.1.3 covers choosing the investigation.

Say that none of them removes the pitfalls. 5.1.6 covers imaging pitfalls.

How Good It Is

The current position is that these techniques are established as measurements and unestablished as decisions.

Their best contribution so far is in research, where a measured tissue property is a better endpoint than a graded appearance. 5.18.2 covers T2 mapping.

Their clinical contribution is unproven, and availability is the practical limit. 5.1.7 covers evidence based imaging.

No figures appear for any accuracy comparison on this page, under decision 028, and no claim about a technique's performance appears either.

FRCS Level

How Good It Is

At decision level the question is which of these would change a decision, and the answers differ sharply.

A quantitative cartilage measure would change the assessment of early disease, where current imaging is least useful and the decisions are hardest. 5.11.1 covers normal cartilage.

Weightbearing imaging would change alignment and patellofemoral assessment, where the supine study answers a question nobody asked. 5.18.5 covers weight bearing imaging.

Automated triage would change service throughput rather than individual decisions, which is a different kind of benefit and easier to evaluate. 5.18.8 covers artificial intelligence in imaging.

Short echo time work would change the assessment of structures currently reported as normal because they cannot be seen. 5.18.4 covers the UTE imaging.

Where It Misleads

The commonest error is quoting a research technique as if it were available. 5.1.7 covers evidence based imaging.

A measured number is read as more reliable than a graded appearance because it is a number. 5.18.2 covers T2 mapping.

An automated measurement inherits the protocol's variability and reports it with false precision. 5.5.1 covers the standard protocol.

A new technique's artefacts are unfamiliar, so they are missed more often than the ones it replaces. 5.2.19 covers artefacts.

The techniques are evaluated against current imaging rather than against outcomes, which cannot show that they help a patient. 5.1.7 covers evidence based imaging.

Fellowship Level

Where It Misleads

At fellowship level the open questions are specific enough to state, which is more useful than a prediction.

Whether a quantitative cartilage measure predicts progression in an individual is open, and the research measures group differences. 5.11.1 covers normal cartilage.

Whether loaded imaging changes an alignment decision beyond what a weightbearing radiograph already gives is open. 5.3.6 covers long leg alignment films.

Whether automated reading reduces error or relocates it is open, and the second is the historical pattern for every reading aid. 5.18.8 covers artificial intelligence in imaging.

Whether the magic angle effect can be exploited as a measurement rather than only corrected for is the question this site's own programme exists to ask. 5.2.20 covers the magic angle effect.

Whether any of it reaches a district general hospital within a decade is the question that decides whether the rest matters, and it is not a clinical question at all. 5.1.5 covers cost effectiveness.