Imaging in patellofemoral instability is essential to the accurate

Transcription

1 REVIEW ARTICLE Imaging in Patellofemoral Instability: An Abnormality-based Approach Paulo Renato Fernandes Saggin, MD,* Jose Idıĺio Saggin, MD,* and David Dejour, MDw Abstract: Imaging in patellofemoral instability confirms the diagnosis and guides treatment. It is essential to the accurate diagnosis of the root causes and their adequate treatment. A structured approach must be used to identify the factors causing instability, and more than 1 imaging modality may provide similar information. This article revises the acute findings and the abnormalities present in the acute and the chronic settings. Key Words: patella, patellofemoral instability, patellar instability, imaging in patellar instability, patellar dislocation (Sports Med Arthrosc Rev 2012;20: ) Imaging in patellofemoral instability is essential to the accurate diagnosis of the root causes and their adequate treatment. In acute cases, imaging is sometimes the only element to provide diagnosis. The 3 main modalities are plain x-ray films, computed tomography (CT), and magnetic resonance imaging (MRI). They must identify the 4 classic factors implied in the genesis of the instability: trochlear dysplasia, patella alta, abnormal tibial tubercle-trochlear groove distance (TT-TG), and patellar tilt (excessive patellar tilt with medial ligamentous disruption), especially in the chronic setting, to guide adequate treatment. ACUTE DISLOCATIONS Imaging of acute dislocations is useful to confirm the diagnosis and define treatment. X-rays are useful to identify gross instability and incongruence. Anteroposterior, lateral, and axial views are complementary and must be performed. Fragments of the patella or the lateral femoral condyle can be identified after acute dislocations and may indicate surgical treatment. CT findings in acute dislocations are similar to the x-ray ones, but with increased accuracy. Smaller osseous fragments can be identified and better measured (Fig. 1). Patellar dislocation may not be suspected before MRI examination in up to 50% of cases. 1 MRI is particularly helpful in acute dislocation recognition and evaluation of associated lesions. The acute findings include the following 2 5 (Fig. 1): Lateral femoral condyle contusion and/or osteochondral lesion; Medial patellar facet contusion and/or osteochondral lesion, sometimes with osteochondral fragment avulsion; From the *Instituto de Ortopedia e Traumatologia de Passo Fundo, Passo Fundo RS Brazil; and wlyon-ortho-clinic, Clinique de la Sauvegarde, Lyon, France. Disclosure: D.D. receives royalties from Tornier, SBM. The other authors declare no conflict of interest. Reprints: David Dejour, MD, Lyon-Ortho-Clinic; Clinique de la Sauvegarde, 8 Avenue Ben Gourion, Lyon, France. Copyright r 2012 by Lippincott Williams & Wilkins Injury of the medial retinaculum at its patellar attachments or midsubstance; Tearing of the distal belly of the vastus medialis obliquus; Injury of the medial patellofemoral ligament at its femoral origin; Patellar tilt and subluxation; Joint effusion. TROCHLEAR DYSPLASIA Trochlear dysplasia is the single most important factor implied in the genesis of patellar instability. 6 The femoral sulcus is not sufficient to provide the osseous restraint capable of avoiding patellar dislocations. Standard lateral x-ray films are the key to dysplasia diagnosis (strict lateral views, with perfect superimposition of the posterior medial and lateral femoral condyles). The crossing sign is typically found in this projection and represents the point where the trochlea becomes flat (the bottom of the groove reaches the height of the facets). 6,7 Additional findings include the double-contour sign and the supratrochlear spur. 8,9 The double contour represents the hypoplasic medial facet found posterior to the lateral one. The supratrochlear spur is found in the superolateral aspect of the trochlea, commonly seen during the surgical exposure of dysplastic trochleae (Fig. 2). Axial x-ray views performed in 45 degrees of knee flexion allow the measurement of the sulcus angle. 10 The normal mean value is 138 degrees (SD ± 6). 11 Angles above 150 degrees are found in trochlear dysplasia. This measurement is impossible in flat or convex trochleae. An important issue when analyzing axial views is that x-rays obtained with higher flexion angles show the lower part of the trochlea, frequently missing the dysplasia present in its upper portion. 12 For this reason, we prefer images obtained at 30 degrees of flexion. CT and MRI have the ability to image the entire trochlea in sequential cuts, from its most proximal part until its distal end. This allows better visualization of the dysplastic upper part. Frequently, dysplasia found in these modalities is missed on x-ray axial views. Bony and cartilaginous trochlear anatomy do not match perfectly what is evident comparing CT and MRI images, 13 but this loses importance in the dysplasia setting, where flat or convex bones will be covered by equally flat or convex cartilage. Carrilon et al 14 investigated the lateral trochlear inclination angle (calculated by means of a line tangential to the subchondral bone of the posterior aspect of the 2 femoral condyles crossed with a line tangential to the subchondral bone of the lateral trochlear facet) in healthy and patellar instability patients. A significant difference between groups was recorded. The mean value in patellar instability patients was 6.17 degrees, whereas in the control group it was 16.9 degrees. Choosing 11 degrees as the threshold Sports Med Arthrosc Rev Volume 20, Number 3, September

2 Saggin et al Sports Med Arthrosc Rev Volume 20, Number 3, September 2012 FIGURE 1. Acute patellar dislocation. A, Axial x-ray showing bone avulsion from the medial patellar aspect. B, Magnetic resonance imaging evidencing medial ligamentous disruption with medial patellofemoral ligament tearing and diffuse edema. C, Computed tomography 3D reconstruction displaying the medial patellar avulsion. value for lateral trochlear inclination, results were excellent in discriminating between the 2 groups, with a sensitivity of 93%, a specificity of 87%, and an accuracy of 90% (Fig. 3). Primarily on the basis of x-ray lateral views, and helped by CT or MRI axial cuts, one can classify trochlear dysplasia into 4 types 8,9,15 (Fig. 4): Type A: The crossing sign is the only of the 3 signs present. On axial views, the trochlea is shallower than normal ones. Type B: The crossing sign and the supratrochlear spur are present. On axial views, the trochlea is flat. Type C: The crossing sign and the double-contour sign are present, but there is no spur. On axial views, the medial facet is hypoplasic. Type D: Combines the 3 signs: crossing sign, supratrochlear spur, and double contour. On axial views, there is a cliff pattern. Lippacher et al 16 analyzed intraobserver and interobserver agreements of radiographic and the MRI-based D. Dejour classification. They concluded that the 4-grade analysis showed fair intraobserver and interobserver agreements, whereas a 2-grade analysis (type A, low grade, vs. types B, C, and D combined, representing high-grade dysplasia) showed good to excellent agreement. They also concluded that the best overall agreement was found for the 2-grade analysis on MRI scans, and that lateral radiographs tended to underestimate the severity of dysplasia compared with axial MRI views. PATELLA ALTA (PATELLAR HEIGHT) Patella alta refers to an abnormally high riding patella that engages the osseous restraint to dislocation (the trochlear groove) later in flexion (in normal knees, the patellar engagement occurs at around 20 degrees of knee flexion), increasing the patellar free arch of movement and facilitating dislocation. As in trochlear dysplasia, x-ray lateral views are the key to the diagnosis of the patellar height. Several methods of measurement (and diagnosis) using the tibia as reference have been described. The 3 main methods are mentioned below (Fig. 5): Caton-Deschamps 17,18 is the ratio between the distance from the lower edge of the patellar articular surface to the anterosuperior angle of the tibia outline (AT), and the FIGURE 2. The 3 findings identifying trochlear dysplasia on lateral x-ray views: crossing sign, double-contour sign, and supratrochlear spur. FIGURE 3. Lateral trochlear inclination: angle between the tangent to the posterior condyles and another tangent to the lateral facet r 2012 Lippincott Williams & Wilkins

3 Sports Med Arthrosc Rev Volume 20, Number 3, September 2012 Imaging in Patellofemoral Instability FIGURE 4. Trochlear dysplasia classification (D. Dejour). 8,15 length of the articular surface of the patella (AP). A ratio (AT/AP) of 0.6 and smaller determines patella infera, and a ratio >1.2 indicates patella alta. Insall-Salvati 19 is the ratio between the length of the patellar tendon (LT) and the longest sagittal diameter of the patella (LP). Insall determined that this ratio (LT/ LP) is normally 1. A ratio <0.8 indicates a patella infera and >1.2 patella alta. Blackburne-Peel 20 is the ratio between the length of the perpendicular line drawn from the tangent to the tibial plateau to the inferior pole of the articular surface of the patella (A) and the length of the articular surface of the patella (B). The normal ratio (A/B) was defined as 0.8. In patella infera, it is <0.5, and in patella alta it is >1.0. Patellar height using the tibia as reference can also be measured on MRI. Miller et al 21 analyzed the patellar FIGURE 5. Patellar height methods of measurement using the proximal tibia as reference. A indicates length of the perpendicular line drawn from the tangent to the tibial plateau to the inferior pole of the articular surface of the patella; AP, length of the articular surface of the patella; AT, anterosuperior angle of the tibia outline; B, length of the articular surface of the patella; LP, longest sagittal diameter of the patella; LT, length of the patellar tendon. r 2012 Lippincott Williams & Wilkins 147

4 Saggin et al Sports Med Arthrosc Rev Volume 20, Number 3, September 2012 height on sagittal MRI of the knee. They applied the Insall- Salvati method to 46 knees, and compared MRI and radiographs. Good to excellent correlation between the values was found, and they concluded that patellar height can be reliably assessed on sagittal MRI using the patellar tendon:patella ratio. On sagittal MRI, patella alta is suggested at values >1.3. Neyret et al 22 measured with radiographies and MRI the patellar tendon length in 42 knees with a history of patellar dislocation and 51 control knees. On MRI images, the mean length was 44 mm in controls and 52 mm in the dislocation group. The distance between the tibial plateau and the point of tendon insertion was also measured and found to be 28 and 29 mm in the control and the dislocation groups, respectively. They concluded that patella alta is caused by a long patellar tendon rather than by its abnormal insertion into the tibia. In addition, they did not find any significant difference between x-ray and MRI tendon length measurements. Another method of measuring patellar height is using the trochlea as the reference. Bernageau et al 23 described a method on lateral x-rays with the knee in extension and the quadriceps contracted. If the inferior edge of the articular surface of the patella (R) is >6 mm above the superior limit of the trochlea (T), there is patella alta, and if R is more than 6 mm beneath T, there is patella baja. Biedert and Albrecht 24 described the patellotrochlear index on sagittal cuts of MRI, performed with the knees in extension, the foot 15 degrees externally rotated, and the quadriceps consciously relaxed. To calculate the index, we must first measure the length of the articular cartilage of the patella [baseline patella (BLp)]. The second measure is the length from the trochlear most superior aspect to the most inferior part of the trochlea facing the patellar articular cartilage (BLt). The ratio BLt/BLp is calculated in percentages, and values above 50% indicate patella baja, whereas values inferior to 12.5% indicate patella alta. Alternatively, axial CT and MRI views provide a clue to patella alta diagnosis when the patella is not found facing the upper part of the trochlea. TIBIAL TUBERCLE-TROCHLEAR GROOVE DISTANCE (AND TORSIONAL MEASURES) TT-TG is a simple way to measure the valgus (lateralizing) forces acting on the patella. TT-TG was described first by Goutallier and Bernageau 25 in 1978 on x-ray axial views at 30 degrees of knee flexion. This distance was able to quantify the coronal alignment of the extensor mechanism, or what is called in clinical evaluation the Q-angle. TT-TG is the distance from the bottom of the most proximal part of the trochlear groove to the proximal part of the tibial tubercle, measured with 2 CT superimposed cuts and expressed in millimeters. 6 Two specific cuts are necessary. The first one is through the proximal trochlea. It is the first cut with cartilage, identified by a slight condensation of the lateral facet and by the shape of the notch, which is rounded and looks like a roman arch. It is called the reference cut. The second cut goes through the proximal part of the tibial tubercle. These 2 cuts are then superimposed. The deepest point of the trochlear groove and the central point of the tibial tubercle are projected on a line tangential to the posterior femoral condyles. The distance between both points is measured. The normal value in a control population is 12 mm; in the population with objective patellar dislocation, FIGURE 6. Tibial tubercle-trochlear groove distance is measured in millimeters. Two axial computed tomography cuts are superimposed. the value is superior to 20 mm in 56% of the cases. 6 Values above 20 mm are considered abnormal (Fig. 6). Another important contribution of CT produced by the superimposition of images is the assessment of torsional deformities, such as femoral anteversion and external tibial torsion. Femoral anteversion is increased in patients with instability (15.6 ± 9 vs ± 8.7 in normal knees), although some overhang of values may exist. Combined with tilt and TT-TG, these constitute the Lyon protocol for CT analysis. 6 Schoettle et al 26 evaluated the reliability of TT-TG on MRI compared with CT scan in 12 knees with patellofemoral instability or anterior knee pain. The mean TT-TG referenced on bony landmarks was 14.4 ± 5.4 mm on CT scans and 13.9 ± 4.5 mm on MRI images. The mean TT-TG referenced on cartilaginous landmarks was 15.3 ± 4.1 mm on CT scans and 13.5 ± 4.6 mm on MRI images. They found excellent interperiod (bony vs. cartilaginous TT-TG) and intermethod (CT vs. MRI measurement) reliabilities: 91% and 86%, respectively. They concluded that TT-TG could be determined reliably on MRI using either cartilage or bony landmarks. PATELLAR TILT (AND SUBLUXATION) Patellar tilt and subluxation refers to the abnormal position of the patella in relation to the trochlear groove. Whereas a tilt refers to increased lateral inclination of the transverse diameter of the patella, subluxation refers primarily to abnormal mediolateral displacement of the patella in relation to the trochlea. Tilt was formerly believed to be one of the leading factors causing dislocations, caused by vastus medialis obliquus insufficiency. Actually, it seems to be the result of a complex interplay of factors, including trochlear and patellar shape and congruence, medial restraint insufficiency, r 2012 Lippincott Williams & Wilkins

5 Sports Med Arthrosc Rev Volume 20, Number 3, September 2012 Imaging in Patellofemoral Instability and lateral retinacular tightness. Whether a cause or a consequence of instability, tilt must be considered for diagnosis and adequate treatment of instability. On the lateral view, the shape of the patella is dependent on its tilt. Normally, the lateral facet is anterior to the crest. Mild tilt occurs when both lines (lateral facet and crest) are superimposed, and severe tilt is when the crest is anterior to the lateral facet. 7 Methods of evaluating tilt and subluxation have been described for x-ray axial views: 1. The congruence angle is measured on x-rays at 45 degrees of knee flexion. After measuring the sulcus angle (used to access trochlear shape), 2 other lines are drawn from its vertex: one bisecting the sulcus angle (reference line) and another to the apex of the patella. The angle between these 2 lines is the congruence angle, which is considered positive if the line to the patellar apex is lateral to the reference line. The average congruence angle is 6 degrees (SD ± 11 degrees), and measures primarily subluxation 11 (Fig. 7). 2. The lateral patellofemoral angle is formed by one line connecting the highest points of the medial and lateral facets of the trochlea and another tangent to the lateral facet of the patella, drawn on 20 degrees of knee flexion axial views. In normal knees, this angle should open laterally (except in 3% in which it is parallel). It demonstrates primarily tilt 27,28 (Fig. 8). 3. The patellofemoral index is the ratio (M/L) between the thickness of the medial joint space (M) and the lateral joint space (L), measured on 20-degree axial views. It should measure 1.6 or less. 27,28 Malghem and Maldague 29 described 1 view obtained at 30 degrees of knee flexion while one examiner pulls the forefoot laterally 30-degree lateral rotation (LR) view. The cassette is held over the patient s thighs, and the x-ray beam is directed cranially. The patellar position (centered or subluxated) is defined according to the congruence angle. In the authors series, the 30-degree LR view was superior to the standard 45-degree axial views in detecting patellar subluxation. In 27 knees operated on for patellar instability, 45-degree routine views depicted subluxation in only 7 cases, whereas 30-degree LR views demonstrated it in all cases. In addition, when both views showed signs of instability, the degree of subluxation was greater in the 30-degree LR view. CT scans allow tilt measurements in complete extension, which increases sensitivity because as the knee FIGURE 7. The congruence angle defined by Merchant. FIGURE 8. The lateral patellofemoral angle (Laurin). flexes, trochlear engagement of the patella reduces or corrects the tilt and subluxation. Another important contribution of CT scans is that they allow tilt measurements to be performed with a constant reference: the posterior femoral condyles (vs. the variable trochlear shape in the instability population observed in x-rays). According to the Lyon Protocol, 6 patellar tilt is the angle formed by the transverse axis of the patella and a tangent to the posterior femoral condyles. It must be measured with and without quadriceps contraction, and this can be accomplished either with 2 superimposed cuts or with a single cut that images both references. Values above 20 degrees are considered abnormal (Fig. 9). Grelsamer et al 30 described their results using an MRI tilt angle similar to that proposed by Dejour and colleagues in the Lyon Protocol, using a line connecting the medial and the lateral borders of the patella and the posterior femoral condyles as reference. Thirty patients with tilt and 51 patients without tilt were evaluated. Patients with significant tilt on physical examination could be expected to have an MRI tilt angle that is 10 degrees or greater, whereas an angle of <10 degrees was associated with the absence of significant tilt on physical examination. As patellar tilt is dependent on the degree of knee flexion (even in normal knees) and quadriceps contraction, the analysis of tilt and subluxation with variable conditions adds important information to the understanding of patellar tracking and to the determination of any abnormalities. Delgado-Martins 31 comparing extension CT images and axial radiographs at 30, 60, and 90 degrees of flexion in normal knees found that in complete extension with the quadriceps relaxed, only 13% of the patellae were centered in the trochlea (the median crest corresponded exactly with the intercondylar groove), whereas this rate increased to 29% at 30 degrees, 63% at 60 degrees, and 96% at 90 degrees of flexion. Schutzer et al 32 also found in healthy subjects a mild degree of lateral shifting and tilting from 0 to 5 degrees of flexion, whereas a central or a medialized patella at 10 degrees of flexion. The study of Martinez et al 33 did not corroborate these findings: 19 of 20 patients had the patella well centered in the trochlear groove in complete extension with the quadriceps relaxed. In the Lyon protocol, tilt is measured with and without quadriceps contraction giving dynamic information of the stability of the patella. In H. Dejour s study, 83% of the objective patellar dislocation group had r 2012 Lippincott Williams & Wilkins 149

6 Saggin et al Sports Med Arthrosc Rev Volume 20, Number 3, September 2012 FIGURE 9. Tilt measurements performed on computed tomography axial cuts with the quadriceps relaxed (A) and contracted (B). patellar tilt superior to 20 degrees compared with 3% in the reference normal group. If instead of using only the relaxed quadriceps measure, a mean is calculated between the measures performed relaxed and in contraction, and the threshold value remains the same, sensitivity and specificity are improved. Ninety percent of the objective patellar dislocation population have presented values superior to 20 degrees, whereas the same remains true only for 3% of controls. 6 Dynamic MRI of the patellofemoral joint has been described to evaluate tracking during early flexion. 34,35 Axial images are acquired sequentially with increments of flexion. These images can be analyzed individually or as a cine-loop display, thus facilitating interpretation and recognition of abnormal tracking. 36 In normal tracking, the ridge of the patella is situated over the center of the trochlea (the groove), and this relation is maintained through increments of knee flexion, as the patella moves distally in the vertical plane. Quantitative assessments have also been described, 37 but despite all the studies produced, no consensus on measurement protocols and abnormal values exist. At the moment, dynamic MRI remains as a promising procedure, but without a well-defined clinical application. REFERENCES 1. Lance E, Deutsch AL, Mink JH. Prior lateral patellar dislocation: MR imaging findings. Radiology. 1993;189: Diederichs G, Issever AS, Scheffler S. MR imaging of patellar instability: injury patterns and assessment of risk factors. Radiographics. 2010;30: Elias DA, White LM, Fithian DC. Acute lateral patellar dislocation at MR imaging: injury patterns of medial patellar soft-tissue restraints and osteochondral injuries of the inferomedial patella. Radiology. 2002;225: Kirsch MD, Fitzgerald SW, Friedman H, et al. Transient lateral patellar dislocation: diagnosis with MR imaging. Am J Roentgenol. 1993;161: Virolainen H, Visuri T, Kuusela T. Acute dislocation of the patella: MR findings. Radiology. 1993;189: Dejour H, Walch G, Nove-Josserand L, et al. Factors of patellar instability: an anatomic radiographic study. Knee Surg Sports Traumatol Arthrosc. 1994;2: Maldague B, Malghem J. Significance of the radiograph of the knee profile in the detection of patellar instability. Preliminary report. Rev Chir Orthop Reparatrice Appar Mot. 1985;71(suppl 2): Dejour D, Reynaud P, Lecoultre B. Douleurs et Instabilité Rotulienne, Essai de Classification. Med Hyg Juillet. 1998; Dejour D, Saggin P. The sulcus deepening trochleoplasty the Lyon s procedure. Int Orthop. 2010;34: Brattstroem H. Shape of the intercondylar groove normally and in recurrent dislocation of the patella. A clinical and X-ray anatomical investigation. Acta Orthop Scand Suppl. 1964; 68(suppl 68): Merchant AC, Mercer RL, Jacobsen RH, et al. Roentgenographic analysis of patellofemoral congruence. J Bone Joint Surg Am. 1974;56: Davies AP, Bayer J, Owen-Johnson S, et al. The optimum knee flexion angle for skyline radiography is thirty degrees. Clin Orthop Relat Res. 2004;423: Sta ubli HU, Du rrenmatt U, Porcellini B, et al. Anatomy and surface geometry of the patellofemoral joint in the axial plane. J Bone Joint Surg Br. 1999;81: Carrillon Y, Abidi H, Dejour D, et al. Patellar instability: assessment on MR images by measuring the lateral trochlear inclination initial experience. Radiology. 2000;216: Dejour D, Le Coultre B. Osteotomies in patello-femoral instabilities. Sports Med Arthrosc. 2007;15: Lippacher S, Dejour D, Elsharkawi M, et al. Observer agreement on the Dejour trochlear dysplasia classification: a comparison of true lateral radiographs and axial magnetic resonance images. Am J Sports Med. 2012;40: Caton J. Method of measuring the height of the patella. Acta Orthop Belg. 1989;55: Caton J, Deschamps G, Chambat P, et al. Patella infera. Apropos of 128 cases. Rev Chir Orthop Reparatrice Appar Mot. 1982;68: Insall J, Salvati E. Patella position in the normal knee joint. Radiology. 1971;101: Blackburne JS, Peel TE. A new method of measuring patellar height. J Bone Joint Surg Br. 1977;59: Miller TT, Staron RB, Feldman F. Patellar height on sagittal MR imaging of the knee. Am J Roentgenol. 1996;167: Neyret P, Robinson AHN, Le Coultre B, et al. Patellar tendon length the factor in patellar instability? Knee. 2002; 9: Bernageau J, Goutallier D, Debeyre J, et al. New exploration technic of the patellofemoral joint. Relaxed axial quadriceps and contracted quadriceps. Rev Chir Orthop Reparatrice Appar Mot. 1975;61(suppl 2): Biedert RM, Albrecht S. The patellotrochlear index: a new index for assessing patellar height. Knee Surg Sports Traumatol Arthrosc. 2006;14: r 2012 Lippincott Williams & Wilkins

7 Sports Med Arthrosc Rev Volume 20, Number 3, September 2012 Imaging in Patellofemoral Instability 25. Goutallier D, Bernageau J, Lecudonnec B. The measurement of the tibial tuberosity. Patella groove distanced technique and results (author s transl). Rev Chir Orthop Reparatrice Appar Mot. 1978;64: Schoettle PB, Zanetti M, Seifert B, et al. The tibial tuberositytrochlear groove distance; a comparative study between CT and MRI scanning. Knee. 2006;13: Laurin CA, Dussault R, Levesque HP. The tangential x-ray investigation of the patellofemoral joint: x-ray technique, diagnostic criteria and their interpretation. Clin Orthop Relat Res. 1979;144: Laurin CA, Lévesque HP, Dussault R, et al. The abnormal lateral patellofemoral angle: a diagnostic roentgenographic sign of recurrent patellar subluxation. J Bone Joint Surg Am. 1978;60: Malghem J, Maldague B. Patellofemoral joint: 30 degrees axial radiograph with lateral rotation of the leg. Radiology. 1989; 170: Grelsamer RP, Weinstein CH, Gould J, et al. Patellar tilt: the physical examination correlates with MR imaging. Knee. 2008;15: Delgado-Martins H. A study of the position of the patella using computerised tomography. J Bone Joint Surg Br. 1979; 61-B: Schutzer SF, Ramsby GR, Fulkerson JP. Computed tomographic classification of patellofemoral pain patients. Orthop Clin North Am. 1986;17: Martinez S, Korobkin M, Fondren FB, et al. Computed tomography of the normal patellofemoral joint. Invest Radiol. 1983;18: Shellock FG, Mink JH, Deutsch AL, et al. Patellar tracking abnormalities: clinical experience with kinematic MR imaging in 130 patients. Radiology. 1989;172: Shellock FG, Mink JH, Fox JM. Patellofemoral joint: kinematic MR imaging to assess tracking abnormalities. Radiology. 1988;168: Brossmann J, Muhle C, Bu ll CC, et al. Evaluation of patellar tracking in patients with suspected patellar malalignment: cine MR imaging vs arthroscopy. Am J Roentgenol. 1994;162: Kujala UM, Osterman K, Kormano M, et al. Patellar motion analyzed by magnetic resonance imaging. Acta Orthop Scand. 1989;60: r 2012 Lippincott Williams & Wilkins 151