The MRI pelvis protocol encompasses a set of MRI sequences for the routine assessment of the pelvis.

Note: This article aims to frame a general concept of an MRI protocol for the assessment of the pelvis in the setting of suspected musculoskeletal pathology. Protocol specifics will vary depending on MRI scanner type, specific hardware and software, radiologist and perhaps referrer preference, patient factors e.g. implants, specific indications and time constraints.

Indications

Typical indications include the following: different causes of groin pain, bilateral hip pain or pain in the buttock and the following:

• bilateral osteonecrosis of the hip 

• groin pain in athletes

• ischiofemoral impingement

• proximal hamstring tendinopathy

• tumours

• assessment for metastatic disease

• lumbosacral plexus imaging

1.5 vs 3 tesla

Musculoskeletal examinations are generally done on both 1.5 and 3 tesla and this is also the case for the pelvis. They profit from the improved spatial and contrast resolution of 3 tesla. Postoperative examinations in patients with metallic implants, however, should be done on 1.5 tesla with metal artifact reduction sequence. 

Patient positioning

An MRI of the pelvis is conducted with the patient in the supine position.

Technical parameters

Coil

Multi-phased array coils are recommended.

• anterior surface coil

Scan geometry

• in-plane spatial resolution: ≤0.7 x 0.7 mm

• field of view (FOV):  320-360

• slice thickness: ≤4 mm

Planning

• coronal images:             

  • angulation: strictly coronal to the body axis 

  • volume: skin to skin

  • slice thickness: ≤4 mm

• axial images:

  • angulation: strictly axial to the body axis 

  • volume: from the iliac crest to the lesser trochanter

  • slice thickness: ≤4 mm

• sagittal images*:

  • angulation: strictly sagittal to the body axis 

  • volume: depends on the clinical question

  • slice thickness: ≤4 mm

• coronal oblique images*:

  • angulation: along the axis of the symphysis

  • volume: includes the whole symphysis

  • slice thickness: ≤3 mm

Sequences

The mainstay in musculoskeletal imaging are water-sensitive sequences, this can be achieved with STIR, T2-weighted fat-saturated images or with intermediate-weighted images.

At least one T1-weighted sequence should be included to ease the assessment and interpretation of bone marrow and/or soft tissue lesions.

Standard sequences

Frequently an MRI of the pelvis does not require any contrast media. A typical MRI of the pelvis might look like as follows:

• T2-weighted or intermediate-weighted (fat-saturated)

  • purpose: bone and/or soft-tissue characterisation, detailed anatomy, including ligament and tendon anatomy as well as an adductor and rectus abdominis muscle insertions

  • technique: T2 FS fast spin echo / IM fast spin echo 

  • planes: coronal, axial, coronal oblique* with decreased FOV

• T1-weighted

  • purpose: bone and/or soft-tissue characterisation, the depiction of inguinal hernias

  • technique:  T1 fast spin echo

  • planes: coronal, axial* (option e.g. in case of lumbosacral plexus imaging or inguinal hernias), sagittal* (option in tumours)

Optional sequences

• T2-weighted

  • purpose: bone and soft tissue characterisation - tumours

  • technique: T2 fast spin echo

  • planes: axial * (option in tumours)

• 3D imaging

  • purpose: for lumbosacral plexus imaging

  • technique:  3D GRE

  • acquisition plane: coronal

Some indications might benefit from the application of contrast media as e.g. inflammatory conditions or tumours.

• T1-weighted C+ (fat-saturated)

  • purpose: inflammatory conditions, tumours

  • technique:  T1 fast spin echo

  • planes: coronal, axial, sagittal* depending on the pathology

(*) indicates optional planes

Practical points

• the protocol can and should be tailored to the specific indication or clinical question

• a typical native protocol will consist of 4-5 sequences

• an alternative to the T2-weighted or intermediate-weighted (fat-saturated) image axial stack would be a T2-weighted Dixon variant, which includes a fluid-sensitive fat-saturated and a non-fat-saturated image stack