Introduction
MRI is an established technique for diagnosing breast cancer; its diagnostic performance depends on image quality, limited by misregistration artefacts due to patient movement. We propose a standardized image quality score (BreastMRI-QUAL).

Methods

Two independent readers with 3 years of experience assigned a 4-level score to each sequence of 50 consecutive 1.5-T examinations at a tertiary cancer centre, as follows: 0 = not diagnostic for any cause; 1 = relevant artefacts/malpositioning with conserved diagnostic value for the specific case; 2 = slight artefacts/malpositioning with conserved diagnostic value; 3 = excellent image quality with full diagnostic value. The score was: (T2-weighted*1) + (DWI-b=0*0.5) + (ADCmaps*0.5) + (T1-weighted-precontrast*1) + (T1-weighted-postcontrast*2) + (T1-weighted-subtracted*3). To obtain a global score (GS) from 0 to 10, the sum was divided by 2.4, with scores below 6 considered as insufficient.

Results

The average GS between the two readers was 8.0 ± 1.0 (mean ± standard deviation), with 42 cases (84%%) receiving a score ≥ 7. Only 1 case (2%) was scored <6 by both readers. The difference between the GSs assigned by the readers was ≤ 1 for 32 cases (64%), >1 but ≤ 2 for 17 cases (34%), and > 2 for 1 case (2%). Bland-Altman analysis showed a mean difference (bias) of 0.28, with the limits of agreement ranging from -1.82 to 2.38, indicating the level of agreement between readers.

Conclusion

BreastMRI-QUAL is a reproducible quality score system. Breast MRI image quality at a tertiary cancer center was good-to-excellent in over 80% of cases.