A Systematic Approach to Diagnostic Imaging - MRI Breast
Breast MRI with and without contrast has several indications:
Screening for high risk patients
Evaluation of extent of disease for known cancer
Follow up of post treatment/neoadjuvant therapy.
Overall approach
Learn the clinical context
Assess the breast density and background parenchymal enhancement
Check for non-breast and incidental pathologies.
Use the T2 and early/first post-contrast images to detect pathology.
Further characterize with vascular kinetics and other sequences.
A separate BI-RADS score is assigned to each breast. A general checklist is below.
Check the history/indication
a. Specifically, the breast biopsy results, surgical history, risk factors/family history.
b. Assess lactation status, time in menstrual cycle (ideally second week of menstrual cycle).
c. Where relevant, check for hormone therapy, medications with sex-hormone or lactation effects.
d. Know if patient has implants or reconstruction, and what type.
Check priors, including mammograms, other modalities.
Note the technique, limitations/artifacts, post-processing/CAD use.
a. Assess adequate contrast delivery by looking at heart, vessel, and liver enhancement.
b. Make sure that at least some vessels in the breast enhance on subtraction images (slow flow on any T1 weighted image can produce signal within vessels)
c. NOTE: if the study is non-diagnostic due to extravasation/partial contrast injection, you cannot repeat the same day (kinetic information compromised by injected contrast).
Assess breast composition on the T1 (non-fat saturated) images
a. Categories: almost entirely fat, scattered fibroglandular tissue, heterogeneous fibroglandular tissue, or extreme fibroglandular tissue.
b. Correlate with prior mammograms or other studies as necessary
Assess the background parenchymal enhancement (BPE).
a. Use the first post contrast images.
b. Categories: minimal, mild, moderate, or marked.
c. BPE can change based on when in the menstrual cycle the patient is imaged.
Examine the non-breast anatomy using the post-contrast and T2 images.
a. Look at the lower neck, chest wall, lungs, heart, and mediastinum.
b. Assess the visualized vasculature on post-contrast images.
c. Look at the visualized upper abdomen, particularly the liver.
d. Assess lymph nodes in the axilla, internal mammary distribution, and mediastinum.
e. Correlate all findings across T1 non-fat-sat, T2, and post-contrast images.
Lymph nodes (axial post-contrast, T1 non-fat-sat, and T2/STIR)
Anatomic Basins & Staging:
Intramammary: Staged as Level I axillary nodes.
Axillary Levels I–III: Lateral/inferior to pectoralis minor (Level I)
Posterior/interpectoral Rotter's nodes (Level II)
Medial/subclavicular (Level III).
Extra-Axillary: Internal mammary chain (specify intercostal space) and supraclavicular fossa.
Morphologic Criteria (Priority over Absolute Size):
Normal/Benign: Reniform shape, thin uniform cortex, preserved central fatty hilum on non-fat-sat T1.
Abnormal/Suspicious: Eccentric or focal cortical thickening, hilum effacement or replacement, rounded shape, irregular margins, and matting or perinodal fat infiltration (extracapsular extension).
Kinetics & Signal: Rapid initial uptake with delayed washout (Type III kinetics), heterogeneous enhancement, or central necrosis (T2-bright center with thick rim enhancement).
Assess each breast on the post-contrast images and T2.
a. Rotating MIPs may be helpful for quick detection of lesions and assessment of overall extent of disease.
b. Rely mostly on the early (first) post-contrast phase.
i. Scan through the breast in at least two planes (axial and sagittal or coronal).
ii. It can be useful to segment each breast into medial and lateral areas when scrolling through the anatomy on the axials.
iii. Also search each axilla and the internal mammary lymph node distribution.
iv. Dynamic post-contrast images can help separate out abnormality from background parenchymal enhancement. Pathology is expected to enhancement earlier than the background tissue.
c. Between the first and last (early and delayed) post-contrast images, you should see most/all pathology.
i. Sometimes, the dynamic images can more convincingly demonstrate lesions that enhance just before the background parenchyma.
d. Describe enhancing findings as masses or non-mass enhancement (NME).
i. Masses:
Shape: Oval, round, lobulated (returned to the lexicon), or irregular.
Margin: Circumscribed or non-circumscribed (specifically indistinct or spiculated; "irregular" margin was retired to avoid duplication with irregular shape and harmonize across modalities).
Internal enhancement pattern: Homogeneous, heterogeneous, thick rim enhancement, or dark internal septations.
T2 signal intensity: hyperintense vs. not hyperintense.
ii. Non-mass enhancement (NME):
Distribution: Focal, linear, segmental, regional, or diffuse (note: "multiple regions" was retired/consolidated).
Internal enhancement pattern: Homogeneous, heterogeneous, clumped, or clustered ring.
iii. For all enhancing findings, evaluate enhancement kinetics:
Initial phase (first 1–2 minutes): Slow, medium, or rapid.
Delayed phase: Persistent (Type I), plateau (Type II), or washout (Type III).
iv. Non-enhancing findings
High T1 signal lesions (e.g., proteinaceous cysts, hematomas)
Signal voids (clips/calcifications)
Non-enhancing masses.
v. Associated / Secondary features:
Involvement of the skin, nipple, pectoralis muscle, chest wall.
Evaluate for peritumoral edema (new dedicated descriptor).
Architectural distortion (T1 non-fat-sat images)
e. Correlate all findings with T1 (non-fat-sat) and T2/STIR characteristics.
g. Compare to priors as necessary.
Assess any implants (if the patient happens to have them on a study done for other reasons).
a. Use implant specific sequences for silicone (water suppressed STIR). Otherwise rely on T2.
b. Assess type/position: saline vs silicone, single vs double lumen, retropectoral vs prepectoral.
c. Look for signs of intra-capsular rupture, extra-capsular rupture.
d. Assess any fluid/collection around implants.
e. Refer to the section on breast implant evaluation for further details.
Assign a BI-RADS score to each breast (unless without contrast/only for implant rupture).
Last checks and proofread.