chapter 4:imaging the patellofemoral joint

Tangential (Axial) View

Standard tangential ("axial" or "sunrise") views of the patellofemoral joint provide good, basic information regarding the condition of this joint (15‑18). There are several techniques for taking a tangential patellar radiograph, but those of Laurin et al (17) and Merchant et al (18) have been most helpful in the authors' experience. The Merchant view requires a special cassette holder and frame. His view is widely used in the United States.

Variable anatomy of the femoral trochlea, however, can make interpretation of tan­gential radiographs difficult. The tangential radiograph also suffers from image distor­tion and overlapping shadows that may affect interpretation. Most tangential radiographs are taken of both patellae simultaneously, such that the radiograph beam is directed in a slightly oblique direction across the patellofemoral joints, thereby introducing further possibility of error. Furthermore, subtle patellar tracking abnormalities in the first 20 to 30 degrees of knee flexion are difficult or impossible to detect with tangential radio­graphs of the patella, because it is technically difficult to capture the patella on a ra­diograph cassette in this range of minimal knee flexion. Nonetheless, tangential radi­ography of the patella is helpful in the basic evaluation and understanding of patellofemoral disorders. A well‑controlled, standard tangential radiograph emphasizing early (20 to 45 degrees) knee flexion can provide important baseline information for the examining physician.


Settegast (19), in 1921, recognized the axial/tangential view of the patellofemoral joint. He described a method in which the patient lies prone with the knee acutely flexed. The beam passes parallel to the posterior surface of the patella and perpendicular to the plate on which the knee rests. The disadvantages of this approach include distortion of the patella with the patient prone and lack of adequate visualization of the proximal trochlea.

In 1924, Jaroschy (20) described a different technique with the patient lying supine. The beam passes parallel to the tibia, and the cassette rests on the distal thigh. He later modified the technique to the prone position with the knee flexed 50 to 60 degrees and the plate lying under the patient's distal thigh. Here, the disadvantage is distortion produced by the beam striking the plate at a 45‑degree angle. Also, the high knee flexion angle diminishes sensitivity. Independently in 1941, Wiberg (21) and Knutson (22) described techniques that reversed the direction of the beam. The cassette rests on the anterior tibia and is perpendicular to the beam, which is horizontal. Furmaier and Breit (23), in 1952, improved this technique by elevating the legs to include both knees, placing one foot against the other, thereby controlling rotation and reducing one source of error. Projecting both knees simultaneously allowed useful comparison. This technique gives a good view but, because the plate is resting on the tibia, views of less than 45 degrees of flexion are difficult if not impossible. Positioning with all of these views is often difficult. Brattstrom (24), in studying dysplasia of the trochlea in relationship to recurrent dislocation of the patella, put forth an accurate but complicated technique with which we have no clinical experience.

Merchant et al (18) described a technique similar to those of Wiberg (21) and Knutson (22) by flexing the leg 45 degrees over the end of the table and angling the tube 30 degrees from the horizontal, thus overcoming some of the positioning problems of the tube. They (18) found difficulty, however, in obtaining satisfactory films at low knee flexion angles. This technique, nonetheless, has been well accepted, is reproducible, and pro­vides an excellent overview of patellofemoral congruence. The methods are summarized in Figure 4.9 (a, b).

Ficat et al (25) described a technique in which both knees are taken simultaneously, thus controlling rotation and offering comparison on a single film. This technique was recommended for obtaining tangential views at 30 to 90 degrees of knee flexion. Because the cassette is placed on the thigh, it can be close to the joint and, thus, will minimize distortion. Unfortunately, however, this also increases radiation delivery to the patient. This technique consists of directing the radiograph beam from the subject's feet toward the patellae, with the beam passing through the contact area to project it onto the cas­sette.

When only one tangential (axial) view is desired, a Laurin 20‑degree knee flexion tangential view (17) or a Merchant view (18) provides valuable information. The Laurin view (Fig. 4.10) is difficult to obtain and requires careful instruction of the radiographic technologist, but is helpful in detecting patients with more subtle tracking abnormali­ties as well as those with more severe patellar tilt, subluxation, or dislocation. When patellar articulation up to 45 degrees of knee flexion is desired, the Merchant view is satisfactory. This view is widely accepted, reproducible, and clinically helpful. Because contact pressure on the patella becomes maximal around 60 degrees of knee flexion, and because some patients exhibit progressive patellar tilt, a tangential view at 45 to 60 degrees of knee flexion will provide some useful information. Occasionally, because of obesity or a very prominent tibial tubercle, it will not be possible to obtain a 20‑ or 30‑degree view, or even an adequate Merchant view.

Laurin (17) showed that the patella is well into the trochlea in 97% of normal individuals by 20 degrees of knee flexion. Further flexion will bring some abnormally aligned patellae back into the trochlear sulcus, which deepens as flexion proceeds, and the desired documentation of abnormal alignment in early flexion may be missed. Using CT, we have found that many young patients with patellofemoral pain will show subtle tracking abnormalities with early knee flexion as compared with asymptomatic control volunteers. With increasing knee flexion, significant patellar malalignment will improve in some patients, which suggests that there are patients with patellofemoral pain related to mild tilt or subluxation (as compared with normal controls), but without detectable abnormality of patellar tracking further into knee flexion.

Malghem and Maldague (26) have introduced the axial (30‑degrees) radiograph with lateral rotation of the leg. This may be useful in the assessment of patellar subluxability. Toft (27) has recommended axial radiographs of the patellofemoral joint with the patient weightbearing. Egund (28) has described another patellofemoral axial view using a standard 15‑degree inclination of the lower leg and the patient erect. Turner and Burns (29) have also recommended the erect position for obtaining tangential views.

Despite attention to detail, tangential radiographs have certain limitations that must be recognized. It is difficult to center the beam perfectly tangential to the contact zone in a way that does not separate the superior and inferior borders of the patella. Image distortion, therefore, is common on tangential radiographs. Knee asymmetry or leg length discrepancy may make absolute comparative views impossible. Some radiograph units do not permit lowering of the tube and collimator below the level of the table. This requires some modification of the actual technique of taking radiographs but does not invalidate the basic method. Finally, an accurate reference line for determining patellar tilt is not possible with tangential radiographs because of variable anatomy of the femoral trochlea. Nonetheless, the tangential radiograph is useful for the general evaluation of patients with patellofemoral disorders.


If the beam has been centered well, the joint line is revealed without overlapping bony structures and is examined for height and degree of parallelism of the opposing subchondral plates. The quality of orientation of trabecular and subchondral bone in both the patella and trochlea should be observed and compared. In the clinical evaluation of patients, we have found two measurements to be most helpful: the Laurin lateral patellofemoral angle (patellar tilt angle) (17) and the Merchant congruence an­gle (18).


Laurin et al (17) described their criteria for patellar tilt based on lines drawn along the lateral patellar facet and the anterior margins of the femoral trochlea. As shown in Figure 4.11, the lateral patellofemoral angle should be open laterally. Laurin et al found that 97% of their controls demonstrated a lateral patellofemoral angle open laterally, whereas 60% of patients with patellar subluxation showed parallel lines, and 40% had a medial facing patellofemoral angle. This simple measurement is a good indicator, for screen­ing purposes, of abnormal patellar tilt. Although the lateral patellofemoral angle has been used to assess "subluxation," it is more appropriate to use this measurement to assess tilt and to use other criteria, such as Merchant's congruence angle, to assess sub­luxation. The patellar tilt angle (as distinct from the lateral patellofemoral angle deter­mined on an axial radiograph) determined by CT or MRI is particularly useful in dif­ferentiating tilt from subluxation, because the posterior femoral condyles provide a much more consistent reference line for determining tilt (30). This will become clearer in the section on computerized tomography. The term patellar tilt angle will define this rela­tionship of the lateral facet to the posterior condylar line.

Merchant's Congruence Angle. We have used Merchant's technique (18) for mea­suring the congruence angle and have applied this also to views of the patellofemoral articulation using CT. The patella normally becomes centralized in the trochlea by 10 to 20 degrees of knee flexion. The trochlea cannot be defined well enough at full knee extension to permit accurate appraisal of congruence with the knee extended. Schutzer et al (31) noted that a congruence angle greater than 0 degrees at 10‑degree knee flex­ion should be considered abnormal. Merchant et al (18) had noted in their study that normal congruence is ‑6±11 degrees at 45‑degree knee flexion. In the study by Schutzer et al (31), there were patients with patellofemoral pain who had abnormal congruence at 10‑degree knee flexion, with correction of this subtle abnormality upon fur­ther flexion of the knee. Imai et al (32) pointed out that this is most likely caused by progressive tightening of the medial retinaculum with increasing knee flexion. These distinct but subtle differences (determined by CT) from an asymptomatic control pop­ulation may not be detected on standard tangential radiographs in all patients.

The congruence angle described by Merchant et al (18) is a good indicator of patella centralization and/or subluxation (as opposed to tilt). Merchant's congruence angle may be applied to tangential radiographs at any degree of knee flexion as well as at the 45‑degree knee‑flexion angle described by Merchant originally. We believe that abnormal congruence should be sustained at 10 and 20 degrees of knee flexion to be considered truly abnormal in determining treatment. Certainly, those patients who exhibit abnormal congruence only at 10‑ and 20‑degree knee flexion with normal congruence at 45‑degree knee flexion must be considered less severely afflicted, although their congruence angles are certainly not quite consistent with those of the normal population.

To determine the congruence angle, the sulcus angle, E'TI', as shown in Figure 4.12, is bisected by a neutral reference line, T0. The apex of the median patellar ridge is connected to the lowest point on the sulcus. When this line, RT, is medial to the neutral reference line, the angle is given a negative value; when lateral, a positive value is assigned.

Other Patellofemoral Indices. There are several other measurements that may be used on occasion, but are not often helpful in the analysis of most clinical conditions. These normal patellofemoral indices are based on a 60‑degree knee flexion tangential view and are summarized in Figure 4.13. Of these indices, the sulcus angle may be most helpful in gaining insight into reasons for patellar instability. Serial evaluation of the sulcus at increasing angles of knee flexion using CT has been most helpful in understanding trochlear morphology.