Kaiser score: breast MRI classification flowchart

The breast MRI classification flowchart (or Tree algorithm) is an evidence-based clinical decision rule to distinguish benign from malignant lesions in breast MRI. It incorporates 5 diagnostic criteria that are mainly consistent with BI-RADS though assigning diagnostic weights.

Diagnostic criteria:

• Root sign: A suspicious protrusion from the lesion surface that has concave lateral aspects. This includes classical spiculations, even subtle ones as well as dendritic branching patterns. NOT consistent with a root sign is a tongue-like aspect with rather convex lateral aspects such as can be seen in fibroadenoma and fibroadenomatoid hyperplasia
• Breast MRI enhancement curves: Simplified as Washout (type III, suspicious), Plateau (type II, somewhat suspicous) and Persistent (type 1, rather benign).
• Margins: General margins, distinguishes between circumscribed (benign) and non-circumscribed (suspicious). Note that most non-mass enhancements will be non-circumscribed by definition.
• Internal enhancement: Dichotomizes suspicious (Rim-like/centripetal and heterogeneous) and typical benign (homogeneous and centrifugal/central enhancement) patterns.
• EdemaOedema: defined as increased water content as visualized by T2w/STIR sequences around the lesion. This finding is highly suspicious and typically associated with lymhovascular invasion and lymph node metastases.

In case of an enhancing lesion, the Reader follows the flowchart in three steps, leading to a diagnostic score associated with an increasing risk of malignancy, ranging from 1 (minimal) to 11 (maximal). Diagnostic decisions and subsequent clinical recommendations should always consider the clinical situation and other imaging findings. As a rule of thumb, scores above 4 should be biopsied.