Applied Radiology

Dr. Seeger is a Professor and Chief of Musculoskeletal Imaging in the Department of Radiological Sciences at UCLA School of Medicine, Los Angeles, CA.

In the current climate of reduced reimbursement for imaging studies, streamlining image acquisition is essential. Gone are the luxury days when one could leave the patient on the scanner table for an hour, acquiring multiple pulse sequences. One must now consider time: scanner time, technologist time, and physician time. The use of intravenous (IV) contrast increases scan time, and may or may not be reimbursable. Extra imaging also increases the cost of supplies, such as film. To be profitable, we must improve patient throughput.

General principles

Under most circumstances, a complete musculoskeletal magnetic resonance (MR) examination can be accomplished with three pulse sequences. To set the protocols, one needs to be aware of the anatomy to be evaluated and the suspected pathology. If more than three sequences are done, consider what additional information you will get from this scan, as your equipment and personnel are likely not being utilized to maximum capacity.

Joint specific protocols

The protocols outlined below should serve as guidelines only. Variations on the pulse sequences are certainly acceptable without compromising the information that can be extracted from the scan. The important message is the use of a maximum of three sequences. The inclusion of a fat-suppressed T2 and/or inversion-recovery sequence is important regardless of the area under investigation. Usually, the fat-suppressed T2 sequence will show anatomic detail better than an inversion recovery scan will. The latter may, however, be useful for determining the presence of very subtle marrow or soft-tissue changes.

Knee: A complete knee study consists of a sagittal double echo, coronal double echo, and axial fat-suppressed T2 scan. This allows evaluation of the menisci in 3 planes, the cruciate ligaments in 2, and the collateral ligaments in 2. The articular cartilage of all 3 compartments is demonstrated well. The patellar retinaculum is evident on the axial images, and fat suppression provides sensitive identification of subtle marrow changes.

Shoulder: A routine shoulder study consists of a coronal oblique double echo, sagittal oblique fat-suppressed T2, and axial double echo or 2D-gradient echoes scan. The rotator cuff is visible in 2 planes, the biceps tendon can be seen within the intertubercular sulcus and at its anchor, and the labrum is seen circumferentially. The acromioclavicular and glenohumeral joints are displayed well, and acromial morphology can be evaluated in 2 planes.

Ankle: A sagittal T1, axial double echo, and coronal fat-suppressed T2 scan will show all important ankle anatomy. The major medial, lateral, and anterior tendons are seen best on the axial images, and the Achilles tendon is seen well in 2 planes. The lateral ligaments are demonstrated on both the axial and coronal scans. All 3 planes are useful for detecting osteochondral lesions. The sinus tarsi and spring ligament can also be evaluated.

Hip: Unless very high-resolution images of a specific part of the bony pelvis or proximal femur are needed, scans should include both hips, the sacrum, and the pubic symphysis, from the iliac crests to the proximal femoral shafts. Since a large proportion of hip MR studies are looking for fracture, tumor, or avascular necrosis, it is useful to image the entire pelvis to detect lesions that would otherwise be missed. The study can be accomplished with a coronal T1, coronal inversion-recovery, and axial fat-suppressed T2 scan.

Trauma

When searching for radiographically occult traumatic lesions of long bones (tibia, metatarsal, etc.), the scan should be coned to the area of clinical concern to provide high-resolution images. The technologist should put skin markers at the proximal and distal area of pain, and use an appropriate surface coil. A longitudinal T1 and inversion-recovery scan, and an axial fat-suppressed T2 scan will show all pertinent findings, including marrow and soft-tissue edema, periostitis, and cortical infractions.

Tumor/mass

Most centers have established protocols for routine MR imaging of joints and the spine, but when faced with an extremity mass or tumor, excessive steps may be undertaken to compensate for diagnostic uncertainty. It is, therefore, important to establish an approach to MR image acquisition and analysis of nonarticular extremity lesions, and to understand both the strengths and the limitations of MRI.

Most musculoskeletal tumors can be imaged adequately with a maximum of 3 pulse sequences: T1-weighted spin echo, fat-suppressed T2, and inversion recovery. The optimal imaging plane will depend on the location of the lesion.

The T1-weighted scan may be in the axial plane in the case of a soft-tissue mass, and should be confined to the mass itself. If a bone lesion is being evaluated, the T1-weighted scan should be in a longitudinal orientation using a large field-of-view. With bone lesions, the purpose of the T1 scan is to evaluate the entire length of bone involvement, provide measurements for surgery, and detect skip lesions. An oblique orientation should be used, if needed, to include the entire bone with both proximal and distal joints on one large field-of view scan whenever possible.

For soft-tissue and bone tumors, a fat-suppressed T2-weighted scan should be acquired in the axial plane. Fat suppression will differentiate between tumor and normal intermuscular, subcutaneous, or marrow fat. The scan should be coned to the lesion, using a surface coil, if appropriate, and scanning to normal tissue above and below. This sequence will allow localization of any soft-tissue mass around the circumference of the bone and compartmental localization of a soft-tissue mass; it will also determine the relationship of the mass to surrounding muscles and assess any involvement of a joint or major neurovascular structures. Periostitis can also be seen.

While most important information can be obtained from these 2 sequences, the addition of a longitudinal inversion recovery scan may complement the study. For bone lesions, the purpose of this scan is to again assess the extent of marrow involvement and evaluate the extent of periostitis or soft-tissue mass that may extend proximal or distal to the marrow involvement. This may alter surgical margins, extending the area of resection. The scan should be acquired with an oblique orientation if needed, and coned to the lesion.

Infection

Most often, MR imaging for infection is requested to evaluate for osteomyelitis. T1-weighted spin echo, fat-suppressed T2, and inversion-recovery sequences are adequate. The imaging planes will be determined by the location of the clinical pathology, and it is important to be aware of the presence of any skin ulceration or areas of induration. This often requires communication with the technologist after the patient has arrived, as specific information regarding the area of abnormality is usually lacking on the requisition.

Intravenous contrast

Usually, IV contrast administration does not add clinically useful information to the MR diagnosis of musculoskeletal lesions. It cannot be used to differentiate benign masses from malignancy, does not improve the accuracy of tumor staging, cannot differentiate peritumoral edema from tumor infiltration, and, with the rare exception of soft-tissue desmoid, does not improve lesion conspicuity over that available with fat-suppressed T2-weighted scans. The marrow edema of osteomyelitis will enhance nonspecifically. Enhancement characteristics such as intensity and homogeneity reflect nonspecific characteristics of any lesion including vascularity, vascular permeability, and the size of the extracellular fluid compartment. Gadolinium (Gd) may, in fact, partially obscure margins of any focal abnormality, as it may increase the signal of a lesion to equal that of normal fat. If Gd is to be used, post-contrast T1-weighted scans should utilize fat-suppression.

Possible uses for contrast administration include differentiation of viable versus cystic or necrotic tumor to guide percutaneous biopsy, identification or documentation of postoperative seroma, and localization of soft-tissue abscesses. Most fluid collections are, however, evident on nonenhanced scans.

Dynamic image acquisition following IV Gd administration has been suggested for tumor-specific diagnosis, differentiation of tumor from peritumoral edema, and evaluation of tumor response to preoperative chemotherapy. This is a time-consuming and costly technique in terms of personnel, scanner time, contrast material, and post-processing. It has been studied extensively in Europe but has not gained clinical popularity in the U.S., where it is generally considered to be an investigational technique.

Conclusion

Efficiency in MR imaging is now essential. While imaging the musculoskeletal system involves the possibility of innumerable body parts and possible pathologic processes, a diagnostic scan can be accomplished in a time-efficient manner by using only three sequences and omitting the use of IV Gd. In order to sequence any scan properly, one needs to be aware of the specific area of concern and the suspected pathology. If needed, this information can be relayed to the physician by the technologist at the time of scan acquisition. AR