EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

Breast MRI update for the radiologist: emerging sequences and artificial intelligence

Michael Haydon, Jared Lim, Harnoor Singh

Journal of Gender-Specific Imaging · 2026-11 · 1 · 100004

DOI: 10.1016/j.gesim.2026.100004 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Gender-Specific Imaging, 2026-11. Verfasst von Michael Haydon und 2 weiteren Autor:innen. Die Arbeit nennt 49 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Purpose

Breast magnetic resonance imaging (MRI) is the most sensitive modality for detecting breast cancer and continues to evolve as new techniques enter clinical use. It is central to high-risk screening, problem-solving of equivocal cases, local staging, and treatment-response assessment. This review offers a radiologist-oriented update on breast MRI, focusing on newer sequences, multiparametric imaging, and the incorporation of artificial intelligence (AI), and, in contrast to existing general overviews, appraises each technique by its current level of clinical readiness and its principal barrier to routine implementation.

Methods

We conducted a structured narrative review of English-language, peer-reviewed articles indexed in PubMed/MEDLINE, Scopus, and Web of Science (January 2010 to May 2026), supplemented by hand-searching reference lists of key reviews and society guidelines. Search terms combined breast MRI with dynamic contrast-enhanced MRI (DCE-MRI), diffusion-weighted imaging (DWI), abbreviated and ultrafast MRI, non-contrast and deep-learning reconstruction, synthetic contrast, and radiomics/AI. Prospective studies, multicenter trials, meta-analyses, and consensus statements were prioritized. A supplementary targeted search for meta-analyses, prospective validation studies, consensus recommendations, and validated diagnostic decision rules was performed during revision.

Results

Structured interpretation of DCE-MRI with a validated decision rule (the Kaiser score) and an apparent diffusion coefficient (ADC) rule-out threshold of approximately 1.5 × 10 −3 mm 2 /s are supported by meta-analytic and multicenter evidence for reducing unnecessary biopsies; the incremental value of DWI over structured DCE-MRI interpretation, however, appears modest, and advanced diffusion techniques have not been shown to outperform standard DWI. Abbreviated protocols have shown accuracy comparable to full-protocol MRI in screening cohorts and higher invasive-cancer detection than tomosynthesis in a prospective multicenter study of women with dense breasts; evidence in diagnostic, staging, and treatment-response settings is limited. Ultrafast DCE-MRI, non-contrast protocols, and AI-generated synthetic contrast remain investigational, and deep-learning reconstruction is supported mainly by single-center image-quality studies. AI applications for detection, segmentation, and classification show early promise for workflow and consistency, yet most evidence derives from retrospective, single-institution datasets and requires external and prospective validation before routine use.

Conclusion

Breast MRI is becoming more quantitative, multiparametric, and efficient. DCE-MRI remains the cornerstone, and structured interpretation with validated decision rules is the best-supported means of improving its specificity; diffusion, non-contrast, and AI approaches have supporting evidence of variable maturity. We conclude with an explicit “ready now / approaching translation / requires prospective validation” synthesis so that readers can gauge which developments are suitable for practice today.

Zitieren

APA

Haydon, M., Lim, J., & Singh, H. (2026). Breast MRI update for the radiologist: emerging sequences and artificial intelligence. Journal of Gender-Specific Imaging, 1, 100004. https://doi.org/10.1016/j.gesim.2026.100004

Vancouver

Haydon M, Lim J, Singh H. Breast MRI update for the radiologist: emerging sequences and artificial intelligence. Journal of Gender-Specific Imaging. 2026;1:100004. doi:10.1016/j.gesim.2026.100004

BibTeX
@article{Haydon2026,
  author = {Haydon, Michael and Lim, Jared and Singh, Harnoor},
  title = {Breast MRI update for the radiologist: emerging sequences and artificial intelligence},
  journal = {Journal of Gender-Specific Imaging},
  year = {2026},
  volume = {1},
  pages = {100004},
  doi = {10.1016/j.gesim.2026.100004}
}
Kostenlos starten →