mri of the hip MALTA MRI in Practice by nikeborome


									MRI Clinical Applications II                                                            Joseph Castillo

5        MRI of the Hip

This section introduces a case study to address learning outcomes 1 and 2. Technical parameters related to
imaging of the hip are discussed and evaluated. Protocols from Malta and a hospital in the UK are
compared, and a modification suggested. Images related to the study are presented at the end of the section.

Normal MRI Anatomy
The hip is a synovial ball and socket joint between the acetabulum, which is deepened
by the fibrocartilaginous labrum, and the femoral head. The femoral head is covered
with articular cartilage although a small area exists on its surface that is devoid of
cartilage, known as the fovea centralis, to which attaches the ligament of the head of the
femur, known as the ligamentum teres. The substantial iliofemoral ligament together
with the less well developed pubofemoral and ischiofemoral ligaments reinforce the
inelastic fibrous joint capsule.         The gluteus minimus and gluteus medius muscles
represent the main abductors of the hip, and these insert onto the greater trochanter.
The main hip flexor, the iliopsoas tendon passes anterior to the hip joint and inserts onto
the lesser trochanter. (Conway & Totty, 1999).
The appearance of the hip on MRI varies depending on the acquisition sequence and
the skeletal maturity of the patient. Refer to section 1 for a table of bone maturity. On
all Pulse sequences cortical bone always produces a thin black line surrounding the
medullary cavity of the femoral head and acetabulum. In the adult the epiphysis is filled
completely with fatty marrow that results in a high signal on all pulse sequences except
STIR and fat-suppressed sequences in which the epiphyses returns a dark signal. In the
child, the epyphyseal cartilage is homogenous and returns a high signal. The immature
metaphyses contains red marrow and returns an intermediate signal in all sequences.
The physeal scar separating the epiphysis and metaphyses is seen as a black line on all
sequences. Visualizing this lines serves as a benchmark for good quality images (ibid).

Clinical Indications
        Hip arthritis
        Labral tears
        Avascular necrosis
        Transient osteoporosis
        Bursitis

MRI Clinical Applications II                                               Joseph Castillo

       SUFE
       Fractures
       Bone and soft tissue tumours
       Infection

Image optimization in MRI of the hip

Patient positioning, choice of coil and scan parameters affect directly image quality.
These will be discussed separately.

Patient Positioning.

Generally the patient is examined supine and feet first because this position is well
tolerated and also facilitates the positioning of the surface coils (Romagnoli & Torricelli,
1997; Westbrook 1999).

Coil Choice

MRI studies of the hip can be performed with the body or a surface coil. The body coil is
generally used when coverage of large anatomical regions is required, for example in
the assessment of neoplastic or infective processes or as a first approach to scan the
area and then followed by surface coil imaging. In this situation the large FOV (range:
32-48cm) show both hips for comparison. Spatial resolution is a major limitation when
using the body coil.
The surface coil provides detailed information of the hip, such as the acetabular labrum,
articular cartilage or when the patient is a child. It is mandatory in the investigation of
small lesions such as early avascular necrosis or arthropathies. Surface coils, which are
positioned directly above the hip joint, have small FOV about 18-20cm, improving spatial
resolution when using thin slices. This however reduces the SNR and multiple NEX is
required. Another disadvantage is that surface coils do not permit to image both hips at
the same time. A study by Kwok et al. (1999) showed that the Torso phased array coil

MRI Clinical Applications II                                             Joseph Castillo

provided better SNR then the body or pelvic coil. In the same study a purposely-built
four-coil phased array gave a sharper and less noisy depiction of the acetabulum.

Scan Parameters.

The parameters used depend on the unit and pathologic condition and patient build.
Generally scanning in all three planes is usually preferred. The coronal plane clearly
demonstrates the lateral and medial surfaces of the femoral head and the
neck/trochanter region. It is limited in showing the superior and inferior aspects. The
sagittal plane shows better the spherical shape, the AP and SI aspects of the femoral
head and joint cavity. The axial plane is a reference plane for other scans and can also
be used to complement the study especially for labrum studies.

Artifact reducing techniques - Presaturation pulses placed S/I reduce artifacts from the
femoral artery and iliac vessels.    Respiratory compensation is not required in hip
imaging, however in single hip imaging bowel movement may cause artifacts. Placing a
pre-satuartion pulse medial to the affected hip could reduce this. Alternatively, a muscle
relaxant could be given intravenously, but this requires informed consent and is probably
never used.

Metallic hardware in situ such a s hip prosthesis or screws produce significant magnetic
susceptibility artefact. Gradient echo sequence must not be used as gradient reversing
does not compensate for magnetic field inhomogeneities. This artefact can be reduced
by using fast spin echo sequences in conjunction with a broad receive bandwidth.
Repeating the study with a swapped phase/frequency direction may sometimes shift the
artefact away from the region of interest. Oversampling must be considered in this
situation to eliminate aliasing.

Sequences - Spin echoes.

MRI Clinical Applications II                                              Joseph Castillo

The T1W SE (TR500, TE 20) is best for morphological assessment of the osteoarticuar
components and enhances the contrast between bone marrow (High) and cortical bone
(low). In this sequence joint cartilage is not shown well because it has intermediate
signal intensity. Instead T2W SE (TR +2000ms and TE 80-100ms) is used to study the
structure because the signal intensity of the bone marrow and cortical bone are low,
while that of liquid is high producing an arthrography-like effect.
Gradient echoes provide a higher acquisition speed, and high contrast between articular
cartilage and subchondral bone.        T1 weighted contrast is obtained with a long TR
>400ms, a very short TE <15ms and FA >60. T2* is obtained with a short TR<100ms a
short TE <30ms and small FA <15

Fat suppression sequences. Diagnostic imaging may be improved by suppressing the
contrast from fatty tissue, thus enhancing the contrast between bone marrow and
possible intramedullary conditions such as oedema, infection or neoplastic disease. Fat
suppression can be obtained by two methods.               One such method, which is not
completely specific for fat, but a robust sequence to be used on low to mid field
scanners is STIR (Short Tau inversion recovery). The TI, depending on the scanner
strength is set to nullify the signal from fat. As the resultant images have a low SNR, the
image quality may be improved by using high NEX (Andrews, 2000).
Another method frequently used on high field magnets is Chemical saturation
sequences.      The fat suppression is based on the chemical shift, which selectively
suppresses the fat with the selective presaturation of the protons immediately preceding
the start of the actual sequence (Hayes & Balkissoon, 1996).

Paramagnetic contrast - The use of contrast medium is usually limited where the
findings point to possible tumour or infective processes. As noted elsewhere contrast
medium spreads intrarticular after 15-60min IV injection producing a partial
arthrographic effect. Paramagnetic contrast medium can also be injected directly into
the joint cavity to opacify it. This method boosts the diagnostic accuracy of MRI in the
study of early cartilage disorders, but it is rather invasive.

Case study - Painful hip muscle strain (after trauma)

MRI Clinical Applications II                                              Joseph Castillo

Fatigue-type stress fractures result from repetitive submaximal stress on normal bone
resulting in a region of accelerated bone remodeling. At microscopic level, repetitive
overloading leads to increased osteoclastic activity that exceeds the rate of osteoblastic
new bone formation. This results in microtrabecular fractures and eventually leads to a
cortical break. Stress fractures are frequently encountered and account for up to 20%
of all injuries seen in sports medicine clinics (Bencardino & Palmer, 2002).
The diagnosis of stress fractures can be based on clinical history, as they are frequently
occult on initial radiographs. MR imaging, due to its sensitivity to bone marrow changes,
shows a diffuse ill-defined hypointense area on T1-weighted images and markedly
increased signal on fat-suppressed T2-weighted and STIR images (Andrews, 2000).


A 27-year-old boxer in the army was referred for MRI complaining of chronic pain in the
groin region. This pain affected his job as he was unable to jump.

Equipment used and patient consideration.

The patient was examined using GE Signa LX MR/i (GE, Milwaukee) using a Torso
phased array coil.        The patient and I completed the safety-screening form.        The
procedure was explained to the patient and then asked him to change into a hospital
gown.     The patient was positioned supine, feet first, with the pelvic area over the
posterior element of the coil. The legs were positioned straight with the feet parallel.
This ensures that both femoral necks are the same on the resultant images. The legs
were immobilised using Velcro pads. The anterior element of the coil is placed over the
anterior aspect of the pelvis and secured in this position using Velcro straps. The laser
light was centered in the midline at the level of the trochanters.
Earmuffs and a panic buzzer were provided to the patient.

Scanning Protocol

MRI Clinical Applications II                                               Joseph Castillo

The local site protocol for hip imaging includes an 3 plane localizer, Axial proton density,
Coronal Stir, Axial T2-weighted fat suppressed and Coronal T2-weighted fat

        TR       TE       TI    F   Matrix     NEX   ETL   FOV     THK     SAT     Time
        msec     msec           A
Ax      4000     35                 512x256    2     6     34      4/0     S/I     6:45
Cor     4800     80       150       256x224    2     10    34      6/0     S/I     5:32
Ax      6000     85                 256x224    4     20    34      4/1     S/I     5:15
T2 fs                                                                      Fat
Cor     4800     86                 256x224    3     14    34      4/1             4:21

Critical evaluation.

This examination evaluates both hips with T1 and T2 weighted image contrast to look for
occult fractures, avascular necrosis, effusions, masses and soft tissue injuries. The
Torso phased array coil is used instead of the body coil because it provides better
spatial resolution with the same coverage, and because we do not have a small coil to
assess the individual hip. The sacroiliac joint and pubic rami are included in the FOV
because hip pain may often be referred from these sites.
The proton density axial sequence is intended to assess the pelvic area for adenopathy,
SI joint disease and sacral fractures.
The FSTIR in the coronal plane is intended to evaluate signs of marrow oedema
involving the affected hip and elsewhere in the pelvis.         It is particularly useful to
diagnose avascular necrosis, as this condition is usually bilateral. The FOV includes the
SI joints and sacrum.
The Axial T2-weight fat suppressed images provide another perspective of the
pathology if detected on the previous sequences.

Unfortunately, although the Torso phased array could provide acceptable resolution with
a small FOV (range 20x24cm), this was not done. Thus as a modification to the above
protocol it is suggested to add a proton density axial and coronal T2-weighted fat
suppressed. As the structures in the hip joint are comparable to those found in the

MRI Clinical Applications II                                            Joseph Castillo

shoulder joint, spatial resolution would be achieved by using a matrix off 512x256, and
slice thickness of 3mm. This would require multiple NEX up to 4 resulting in longer scan

Using pre-saturation pulses placed S and I to the FOV reduced flow artiacts from
femoral and iliac vessels.
On comparing our site protocol with that used at the UK hospital, the major differences
include less scanning time and minor changes in the parameters. The scanning time is
reduced because three sequences are used instead of four and the TR is lower. The
STIR sequence at the UK hospital results in nicer fat suppression, because the TE used
is intermediate (30) rather than high (80). This also applies to axial T2 fat suppressed
sequence.       The resultant outcome indicates that our site protocol require minor
modifications. This protocol compares to the one suggested by Kaplan et al (2001) for
the evaluation of fractures or AVN. However for labral tears a dedicated protocol with
smaller FOV and thinner slices is required.

         TR      TE       TI   F     Matrix     NEX   ETL   FOV   THK   SAT     Time
         msec    msec          A
Cor      475     Min                 512x224    2           40    5/1   SI      3:33
T1               full
Cor     3375 30         120          256x224    2     10    40    5/1   SI      2:41
AX      3625 68                      384x224    3     16    40    5/1   Fat     2:17
T2 fs
Protocol as used at a UK Hospital.


The Coronal FSTIR and axial T2-weight fat suppression sequences shows a
hyperintense area over the lesser trochanter. The high-resolution axial proton density

MRI Clinical Applications II                                              Joseph Castillo

shows lower intensity areas over the same area. No fracture line was identified. The
resultant images together with the clinical history of the patient indicate a stress injury
(see images end of section).      Follow up studies are advisable in light of no high-
resolution images.

Treatment and prognosis

MRI of the hip in-patient suspected to have occult hip fractures might impact treatment
decisions.    The detection of stress fractures is important as conservative treatment,
such as non-weight bearing of the injured hip, prevent further progression to larger
fractures. Detection of soft tissue trauma in the pelvis and thighs does not require
specific treatment, although it may reasonably aid the physiotherapist during treatment.


The osseous pelvis and hip contain an extensive amount of marrow in which a variety of
processes may occur. Evaluation of this region requires an understanding of normal
maturation and recognition that the marrow contain variable amount of red and yellow
marrow. This variability results in various patterns in MRI ranging from homogenous
signal intensity to patchy signal intensity.     From a technical point of view, although
patient positioning is straight forward anxiety attacks may be infrequent, the choice of
coil and sequence parameters are critical.         Specially designed phased array coils
coupled with powerful and faster gradients will permit smaller FOV and thin slices
necessary to provide further evaluation of labral tears and cartilage.


MRI Clinical Applications II                                                 Joseph Castillo

       Andrews C.L. (2000) Evaluation of the Marrow Space in the Adult Hip, Radiographics

       Bencardino J.T., Palmer W.E. (2002) Imaging of hip disorders in athletes, Radiologic
        Clinics of North America, 40:267-287.

       Conway W.F., Totty W.G. (1999) Hip. In:Magnetic Resonance Imaging, (ed Stark DD &
        Bradley WG Jr). 3 Edn. St.Louis, USA, Mosby.
       Hayes C.W., Balkissoon A.A. (1996) Magnetic Resonance Imaging of the Musculoskeletal
        system II. The Hip, Clinical Orthopaedic, 322:297-309.

       Kaplan PA, Helms CA, Dussault R, Anderson MW & Major NM (2001) Musculoskeletal
        MRI, 2 ed, Philadelphia, USA, W.B.Saunders Company,.

       Kwok W.E., Lo K.K., Seo G., & Totterman S.M.S. (1999) A volume adjustable four-coil
        phased array for high resolution MR imaging of the hip, Magnetic Resonance Materials in
        Physics, Biology and Medicine, 9:59-64.

       Romagnoli R., Torricelli P. (1997) Magnetic Resonance Imaging of the hip, La Radiologia
        Medica 93:150-155.


To top