Imaging the Knee

The knee is the most frequently imaged lower limb joint and one of the most commonly tested imaging assessment topics in fellowship examinations. Systematic knee imaging rests on four principles: choosing the right views (weight-bearing views for osteoarthritis, the Ottawa rules for trauma), reading the radiograph systematically, knowing when MRI or CT adds value, and interpreting the MRI findings of internal derangement accurately.

Weight-bearing is the first principle. Non-weight-bearing radiographs significantly overestimate the remaining joint space, because the cartilage surfaces are not compressed under load. The error can lead to underestimation of the severity of osteoarthritis and to inappropriate surgical decision-making.

The Ottawa Knee Rules. After an acute knee injury, radiographs are indicated only if any of the following is present:

• Age 55 or older
• Tenderness at the fibula head
• Isolated tenderness of the patella
• Inability to flex the knee to 90 degrees
• Inability to weight-bear for 4 steps, both immediately after the injury and in the emergency department

If none of these is met, the probability of a clinically significant fracture is very low (95% CI 0% to 0.4% in the validation cohort) and radiographs can be safely deferred. In prospective validation the rule was 100% sensitive for clinically important fractures while cutting radiograph use by about 28%.

Reading the radiograph. Work through alignment, bones, cartilage spaces, soft tissues and special views in that order, so that the assessment is complete every time:

• Alignment - mechanical axis, tibial plateau alignment, patellar position (alta or baja), varus or valgus angulation, and joint congruence (dislocation)
• Bones - trace every cortex: femoral condyles, tibial plateaus, patella and fibula head, looking for fractures, lytic lesions and avulsion fragments (Segond, tibial spine)
• Cartilage spaces - joint space width, medial against lateral and patellofemoral, on weight-bearing views
• Soft tissues - suprapatellar effusion, lipohaemarthrosis (fat-fluid level), Pellegrini-Stieda calcification, popliteal cyst and soft-tissue swelling
• Special views - Rosenberg for the posterior condyles, skyline for the patellofemoral joint, and the tunnel view of the intercondylar notch for loose bodies and osteochondritis dissecans

Radiographs first. Plain radiographs are the first-line test for every knee presentation, with a weight-bearing AP and a lateral as the minimum. CT characterises fractures (tibial plateau) and malalignment; MRI is the reference test for menisci, cruciates and cartilage; ultrasound answers questions about effusions, popliteal cysts and guided injection. The scenarios in the table below set out the sequence.

The weight-bearing series. Four views are the key to a comprehensive knee assessment. The standing AP, taken with the beam horizontal, shows the medial and lateral compartment joint space under physiological load. The full-length standing hip-knee-ankle film gives the mechanical axis, and is essential for total knee replacement planning and for assessing a high tibial osteotomy. The Rosenberg view is taken PA with the patient standing and the knees flexed 45 degrees. The skyline view, with the patient seated and the knees flexed 30-45 degrees, shows patellar tracking, tilt, subluxation and the patellofemoral joint space.

Why the Rosenberg view. The standard AP weight-bearing view may miss early to moderate osteoarthritis. Cartilage wear often begins posteriorly on the femoral condyles, and an extension AP projects through the anterior, thicker cartilage, which obscures the posterior wear. Flexing the knees to 45 degrees shifts the beam through the posterior condylar surface, where early cartilage loss occurs; in the original validation against intra-operative findings the view was more accurate, more sensitive and more specific than conventional extension AP views for grade III-IV cartilage loss. It is essential for assessing the severity of osteoarthritis before osteotomy or arthroplasty.

View larger

View larger

Bone marrow oedema (contusion) on fluid-sensitive MRI is a fingerprint of the injury mechanism: the location of the "kissing" bruises tells you which structure was loaded and which ligament to scrutinise. Marrow oedema is also the most durable evidence available. The ligament may be partially imaged or equivocal, but the bruise records where the two bones actually collided, and it persists for weeks after the effusion has gone.

Read the bruise, then hunt the lesion. The contusion pattern is a clue, not the diagnosis. A pivot-shift bruise should send you to scrutinise the ACL and the posteromedial meniscocapsular junction for a ramp lesion; a medial-patellar-facet plus lateral-femoral-condyle bruise should send you to the MPFL and to the lateral trochlear and patellar cartilage for an osteochondral fragment. Bone marrow oedema also has non-traumatic mimics (red-marrow reconversion, insufficiency fracture, SIFK), so always correlate the pattern with the clinical mechanism.

View larger

View larger

Patellar instability is the one knee problem where the radiographic measurements are the diagnosis, because each one names a separate anatomical reason the patella escapes laterally, and each has a different operation attached to it. A candidate who says only "there is patella alta" has answered a quarter of the question.

Patellar height

Measured on the true lateral. Patellar height is measured on a true lateral with the knee flexed about 30 degrees. The ratios in use are not interchangeable:

Trochlear dysplasia

The true lateral radiograph is the diagnostic view, not the skyline. This surprises candidates who assume an axial view must be needed to judge a groove. A true lateral superimposes the two femoral condyles, so any abnormality of the trochlear floor projects as an extra line:

• The crossing sign - the line of the trochlear floor crosses the anterior border of a femoral condyle, meaning the groove has become flat at that point. This is the entry criterion for dysplasia.
• The supratrochlear spur (bump) - a prominence at the proximal trochlea, over which the patella must climb.
• The double contour - the medial trochlear facet projecting behind the lateral, indicating a hypoplastic medial facet.

The Dejour classification (types A to D) is built from these three signs. A is a crossing sign alone with a shallow but symmetrical groove; B adds a spur and a flat trochlea; C adds a double contour with asymmetric facets; D has all of them plus a vertical cliff between the facets. Types B and D are the ones that usually need trochleoplasty rather than a soft-tissue procedure alone.

View larger

Putting the measurements together

Four independent abnormalities. Lateral patellar instability has four separate causes, and a scan should be interrogated for each: trochlear dysplasia on the true lateral, patella alta on a height ratio, excessive lateralisation of the extensor mechanism on the TT-TG or TT-PCL distance, and patellar tilt of over about 20 degrees on axial imaging. They coexist but are independent, and the operation follows whichever ones are abnormal.

View larger

View larger

Ultrasound is cheap, dynamic and radiation-free, and it is the only knee modality that images a structure while the patient moves it. Its weakness is the mirror image of that strength: it is operator-dependent, and sound will not travel through bone, so everything inside the joint is hidden. Knowing precisely which questions it answers, and which it cannot, matters most where MRI is scarce.

The blind spot. Ultrasound is the right first tool for an effusion, a suspected Baker cyst, a superficial tendon problem or a guided injection, and it carries more diagnostic weight where MRI is scarce. It cannot reliably assess the cruciate ligaments, the menisci or the articular cartilage, so a haemarthrosis with a positive Lachman still needs an MRI, not just an ultrasound.

View larger

View larger

View larger

View larger

Knee imaging follows a broadly consistent evidence-based algorithm worldwide: plain radiographs first (weight-bearing for any osteoarthritis question), a validated decision rule to ration radiographs after acute trauma, MRI for internal derangement, and CT for fracture morphology. The differences between societies are mostly of emphasis and access rather than of principle.

Global epidemiology. Symptomatic knee OA affects an increasing share of the adult population worldwide and is a leading cause of disability in the Global Burden of Disease estimates, driven by ageing and rising obesity. ACL injuries occur most commonly in young athletes (pivoting/contact sports), and meniscal tears span a bimodal pattern — traumatic tears in the young and degenerate tears in middle age and older. Tibial plateau fractures cluster bimodally too (high-energy injury in the young, low-energy fragility fractures in older adults).

Registry and audit context. There is no dedicated imaging registry, but national arthroplasty registries (NJR for England/Wales, AJRR in the US, AOANJRR in Australia, the Swedish and Norwegian registries, NZJR) rely on standardised weight-bearing and full-length alignment radiographs for pre-operative planning and outcome audit, reinforcing weight-bearing imaging as the global standard for OA workup.

High- versus limited-resource practice variation. Where MRI is scarce or expensive, ultrasound (effusion, Baker cyst, dynamic patellar tracking, guided injection) and careful clinical examination carry more diagnostic weight, and diagnostic arthroscopy may still be used selectively. In well-resourced settings, pre-operative MRI before arthroscopy is routine and purely diagnostic arthroscopy has largely been abandoned. Decision rules such as the Ottawa Knee Rule are especially valuable in high-volume or resource-limited emergency settings to ration radiography safely.

MRI in degenerate meniscal tears. In middle-aged and older patients with knee pain, MRI frequently shows a degenerate meniscal tear, and such tears are also highly prevalent in asymptomatic knees. Multiple trials show that arthroscopic partial meniscectomy offers little benefit over exercise therapy for degenerate tears without mechanical locking. The controversy is how much weight to give an MRI tear when the same finding is so common in pain-free knees.

Incidental findings and over-imaging. Early MRI can drive over-treatment by surfacing incidental degeneration, marrow oedema or chondral changes of uncertain significance. Decision rules (Ottawa) and a weight-bearing-first radiograph strategy exist precisely to limit low-value imaging, yet practice variation in MRI access and referral thresholds remains wide between health systems.

The anterolateral ligament and the Segond fracture. The structure avulsed in a Segond fracture has been debated: lateral capsule, anterolateral ligament, or the capsulo-osseous iliotibial band. The clinical message is stable. In adults it signals ACL-complex injury and a high rate of meniscal ramp lesions, but in children it can occur with an intact ACL, so MRI, not the radiograph alone, defines management.

Field strength and 3T MRI. Higher field strength (3T) improves signal-to-noise and the depiction of small structures (cartilage, small meniscal and ligament tears), but 1.5T remains diagnostically adequate for most internal derangement. The incremental value of 3T over a well-performed 1.5T study for routine reporting is real but modest, and access varies globally.