Before beginning discussion of the labrum, although widely known, it should be emphasized that labral pathology has a high prevalence in the asymptomatic population and is frequently seen on imaging done for reasons other than hip pain. For example, in an MRI study of 45 patients without hip pain, Register and colleagues found that 69 % of hips had labral tears [2].

Many studies have been performed evaluating the accuracy of detecting labral tears using MRI with and without contrast (gadolinium). The majority of studies have been performed with the use of joint distention with arthrography. One study examined the detection of tears with the use of intravenous gadolinium [14]. The sensitivities and specificities in tear detection with MR arthrography (MRA) range from 69–100 % and 0–100 %, respectively [15]. The sensitivities and specificities in tear detection with conventional MRI range from 0–97 % to 33–100 %, respectively [15]. The wide range of the sensitivity and specificity for both reflect a great heterogeneity of the studies, particularly the fewer conventional MRI studies. It is difficult to draw a definite conclusion as to whether MR arthrography (MRA) or conventional MRI in the evaluation of labral tears.

On MRI, the normal labrum is homogeneously low in signal. With aging, intrasubstance degeneration is common, manifested as intermediate to even high signal contained within the labrum. A triangular shape in all quadrants is far and away the most common shape. It occasionally thin and elongated and not infrequently somewhat rounded in shape with aging. Intralabral ossification is frequently seen, often reflecting pincer FAI (Fig. 3).

There has been one grading scheme proposed using MRI and one using arthroscopy [16, 17]. A study directly comparing the schema found no significant correlation between the two [18]. Different descriptive terms have been and are currently used to describe tears. It is most important for the radiologist and referring surgeon to use the same terminology.

On MRI, labral tears may be intermediate in signal. More often, tears are high in signal, similar to joint fluid (on both conventional MRI and MR arthrography (MRA)). Peripheral detachments (i.e., peripheral longitudinal tears) are seen when linear fluid or contrast insinuates between the labrum and rim at the chondrolabral junction (Fig. 4). Radial (or flap) tears are present when fluid or contrast extends into the substance of the joint side of the labrum, entering discretely away from the chondrolabral junction (Fig. 5). Sometimes multiple small radial tears are present, also called a radial fibrillated tear (Fig. 6). A complex tear is the preferred term when signal is branching and extending in multiple directions or multiple tear types coexist (Fig. 7). Occasionally, bucket-handle tears can be seen with frank displacement of a portion of the labrum (Fig. 8). Abnormal shape, particularly blunting, of the labrum is often a clue to a displaced flap. While the assessment of tear stability is very difficult on MRI without frank labral displacement, instability can be suggested if the labrum is thickened and distorted [18].

Normal labral sulci, simulating peripheral detachments, are potential pitfalls on MRI. Sulci are normal variant clefts at the chondrolabral junction. They can be located at one position on the clockface (e.g., 8 o’clock) or extend over several positions (e.g., 8–9 o’clock). On MRI, they are well-defined linear areas of high signal located at the chondrolabral junction (Figs. 9 and 10). Importantly, they are partial-thickness clefts, distinguishing them from full-thickness peripheral detachments. There is no extension of the signal into the adjacent substance of the labrum. The adjacent labrum is usually normal in signal.

Several studies have found very conflicting results as to the frequency of sulci at all different locations [19–22]. In the author’s experience, sulci are frequent in all quadrants (particularly around 8 o’clock and 11 o’clock) except for anterosuperiorly. Given the majority of these studies have found normal sulci in the anterosuperior quadrant, it would be impossible to distinguish a focal partial-thickness peripheral detachment from a sulcus in this location. Fortunately, for MRI diagnosis, tears most often extend over multiple hours on the clockface in this quadrant (Fig. 11). Additionally, isolated tears in quadrants other than the anterosuperior quadrant are infrequent.

The suitability for tear repair or even refixation can be assessed by noting the underlying quality of the labrum. As mentioned previously, the normal labrum is homogeneously low in signal. A degenerated labrum is usually diffusely intermediate to slightly high in signal. Occasionally, a frank round fluid signal focus can be seen in the labral substance, consistent with focal mucoid/myxoid degeneration.

A paralabral cyst is pathognomonic for a labral tear, often seen in the setting of underlying extensive labral degeneration (Fig. 12). The vast majority of the cysts have the same signal intensity as the joint fluid. On occasion, they may have different signal than joint fluid, likely due to mucinous or proteinaceous contents. They are well circumscribed and may be multilobulated or multiseptated. While usually small, they may be large and elongated. Even rarer, large paralabral cysts have the potential to impinge on adjacent neurovascular structures. A connection/communication of these cysts with the labrum should be seen on MRI; the communication may be very small and only definitively visualized on one image of the entire exam. Nonvisualization of a discrete communication should raise the possibility of a cystic mass.

In using MRI for labral tear detection, it is useful to recognize the strengths and limitations for the standard imaging planes used in the evaluation. Coronal images are best to evaluate the superior labrum. Oblique axial images are best to evaluate the anterior and posterior labrum. Sagittal images are best to evaluate the anterior labrum around the 3 o’clock position. However, not infrequently anterior tears can be seen, occasionally to better advantage, on the coronal images (Fig. 13). Additionally, superior tears can be seen on the oblique axial images. Due to the obliquity of the dome, the labrum is more difficult to evaluate accurately anterosuperiorly (particularly around 1 o’clock) and posterosuperiorly (particularly around 11 o’clock) in all standard imaging planes; the labrum is not crisply visualized due to the curvature and volume averaging in these locations. Given this relative limitation, as the vast majority of tears are located in the anterosuperior quadrant, particular scrutiny of all imaging planes should be made of this quadrant. Labral tears, including those in the anterosuperior quadrant, can be seen on low-resolution and/or large field of imaging studies (Figs. 14, 15, and 16).

Radial MR imaging has thus been utilized to evaluate the labrum. Radial imaging is performed by using a plane perpendicular to the acetabular rim, obtained from the oblique axial and coronal sequences. Slices in this plane are then obtained at a constant interval (usually 15° intervals) over 360°. This technique lessens the problem of volume averaging in standard imaging planes as the labrum can be imaged in a more direct, perpendicular fashion at all clockface locations. Somewhat surprisingly, however, the use of the sequence has been shown in one study to not improve the accuracy of labral tear detection [23].

Certain labral tear sites are associated more frequently with certain conditions. For example, given the otherwise relative rarity, the presence of a posterior labral tear suggests prior posterior subluxation or dislocation. An isolated tear at or close to 3 o’clock may reflect iliopsoas impingement, particularly in the absence of FAI pathomorphology or its sequelae (Fig. 17) [24–26]. However, the labrum may appear entirely normal at MRI in iliopsoas impingement; no findings suggestive of the focal injection and inflammation seen at arthroscopy are detected with MRI. No specific findings related to the iliopsoas tendon have been found in the impingement patients [24, 25].

The contribution of ligamentum teres tears to pain is difficult to accurately pinpoint as most patients have other coexistent pain-generating pathology (e.g., labral tears, chondral lesions) [42–44]. Pain due to isolated tears without other intra-articular pathology can, however, occur [45]. Its contribution to hip stability is controversial [46]. If the ligament has a significant contribution to stability, it would be expected and make sense that patients having previously undergone surgical hip dislocation with usual sacrifice of the ligament would have subsequent instability; however, to the author’s knowledge, no such literature of instability following surgical dislocation exists.

The ligament is usually composed of two bands, although it has been found to instead be composed of three bands [47]. Congenital absence of the ligament has been described [48]. The normal ligamentum teres is homogeneously low in signal on all sequences, extending from the fovea/bare area distally to insert on the transverse acetabular ligament and adjacent pubis and ischium. It is more ovoid in shape proximally at the fovea; its midportions and distal portions are flatter and somewhat pyramidal in shape.

Like all ligaments, intrasubstance degeneration is common with aging, most frequently observed proximally, manifested as intermediate and high signal within its substance on fluid-sensitive sequences. Given the mucoid/myxoid and fatty degeneration, intrasubstance intermediate to high signal may be seen within the ligament on T1-weighted sequences. Additionally, occasionally osseous metaplasia can be seen in the ligament.

Partial tears of the ligament are fairly common (up to 46 %) and ruptures less so (4–15 %) [42, 45, 49, 50]. Partial tears are often difficult to differentiate from degenerative change of the ligament on MRI. The sensitivity and specificity in partial tear detection are subpar [51].

As most partial tears occur proximally at or near the fovea, this location should be most highly scrutinized. The ligament has normal striations of high to slightly high signal on fluid-sensitive sequences, reflecting interstitial fluid between its layers of collagen, throughout its course that usually become more conspicuous with aging; this signal is also located between its discrete bands. Both the intact degenerated ligament and partially torn ligament can have frayed margins and be attenuated or thickened on MRI (Fig. 33). Partial tears may be small, very thin, and focal. For definitive diagnosis of a partial tear, fluid signal should be seen extending to at least one margin of the ligament (Fig. 34).

However, partial tears can simply manifest with intrasubstance signal. In an MR arthrographic study, 10 (out of 12) known partial tears only had linear intrasubstance fluid signal, with or without irregular contour of the adjacent periphery of the ligament; intact ligaments also had similar findings [51]. Hence, both false-positive and false-negative diagnoses on MRI are not unexpected. Further complicating matters of diagnosis, as Botser et al. state in their series delineating partial tears into less than or more than 50 % in thickness, “low-grade partial-thickness tears … may have been disregarded in previous literature” [42].

The oblique axial (or straight axial) plane is usually the best plane to detect partial tears as the ligament is seen in cross section. The coronal plane may also be useful, although the normal striations within and thin separation between the bundles can simulate tears. The sagittal plane is frequently unhelpful given relatively poor visualization of the ligament and volume averaging as this is the plane that the majority of the ligament traverses. In addition, the so-called ligamental plica can often be seen just medial to the ligament, particularly after arthrography and in the setting of a joint effusion; although rarely confusing due to its clear separation from the ligament and its elongated course, it should not be mistaken for a partial tear (Fig. 35) [52].

Rupture of the ligamentum teres is most often due to prior hip dislocation or in the setting of advanced degeneration [53]. Ruptures can occasionally be due to a twisting injury [45]. Rupture detection is straightforward. Frank discontinuity is present, usually at the fovea or in its midportion (Fig. 36). The ligament distal to the site is frequently attenuated and wavy. While occasionally reactive marrow edema, or a frank intraosseous ganglion, is present at its foveal insertion, almost always in the setting of ligament degeneration, parafoveal marrow edema can signify the presence of an acute rupture/avulsion in the setting of recent trauma.

No description of the MRI appearance of ligamentum teres reconstruction or tears of reconstructions in the literature is known to the author. Likely, similar to the appearance of the reconstructed or augmented labrum, heterogeneous signal is to be expected within the reconstructed ligament on MRI. Tears of the reconstructed ligament most likely will be those of any ligament, namely, fluid signal extending into the ligament with or without an irregular contour or frank displacement or absence of a portion of the ligament.

The capsule of the hip includes four distinct thickenings: the iliofemoral, ischiofemoral, and pubofemoral ligaments and the zona orbicularis. The iliofemoral ligament is a stout thickening of the anterior and anterolateral capsule; it is readily visualized on all three usual planes of imaging on MRI. The ischiofemoral ligament is a posteroinferior thickening of the capsule; it is also visualized well on all three standard planes of imaging. The pubofemoral ligament, located caudally, is the most difficult of the capsular ligaments to image fully, best noted on the sagittal and axial sequences. The zona orbicularis, present as a circular sling-like structure about the femoral neck, can be seen in all three planes; it is better visualized in certain locations in different planes (e.g., due to volume averaging, the anterior and posterior portions of the ligament are better seen on the oblique axial sequence than on the coronal sequence).

In an MR arthrogram study of 30 patients, equally divided between men and women, men had a significantly thicker capsule only anteriorly [54]. However, a defined normal thickness of the capsular ligaments is not known. It is also not known what, if any, change in thickness of the ligaments occurs with capsular distention (e.g., after arthrography). However, although nonspecific, as in the glenohumeral joint, subjective capsular laxity suggests capsular instability; this subjective assessment most likely is more accurate in the eyes of an experienced radiologist who has seen many hip joints imaged and has a rough baseline as to the typical appearance of the capsule. In the author’s experience, the native capsule in patients with capsular laxity/insufficiency has more prominent outward curvature, particularly evident near the insertions (Figs. 37 and 38). In the author’s experience, unless a process is present predisposing the patient to capsular laxity (e.g., Ehlers-Danlos, Marfan, Down syndrome), the insufficient capsule usually subjectively appears normal in thickness.

The iliofemoral and/or ischiofemoral ligament can be disrupted during hip dislocation, as can, occasionally an avulsion fracture at the ligament’s insertion. On MRI, the findings are not subtle and usually manifest as a large gap in the injured ligament, often with joint fluid extravasating into the adjacent soft tissues (Fig. 39).

Of note, the iliofemoral ligament can appear markedly irregular and usually abnormally thickened due to iatrogenic contrast injection during arthrography (Fig. 40). While usually in its midportion, the abnormal appearance can be relatively diffuse. This spurious finding on MRI can usually be detected as the gadolinium-containing mixture is high in signal on the fat-suppressed T1-weighted sequences (unlike normal synovial fluid); the high signal on these sequences will be seen in the ligament at the site of abnormality, often extending into the adjacent soft tissues along the injection track.