Can deep learning detect breast cancer as well as a radiologist?

Deep learning (DL) is a type of artificial intelligence that can analyze medical images. Several studies have compared DL algorithms to radiologists for detecting breast cancer on mammography and tomosynthesis. Current evidence shows that DL performs at a level comparable to radiologists, especially for certain tasks like classifying lesions. However, DL is not yet ready to replace human readers, and more research is needed before it can be widely used in clinical practice.

What the research says

A 2025 systematic review and meta-analysis of 13 studies with over 38,000 patients found that stand-alone DL algorithms for digital breast tomosynthesis (DBT) achieved a pooled sensitivity of 0.88 and specificity of 0.74, with an area under the curve (AUC) of 0.89 ². This performance was not statistically different from that of all radiologists (AUC 0.89 vs 0.88) or senior radiologists (AUC 0.89 vs 0.90) ². Another review from 2022 noted that studies show similar and even better performances of DL algorithms compared to radiologists, but emphasized that large trials are still needed, especially for ultrasound and MRI ¹⁰. Additionally, machine learning techniques are being applied to improve HER2 status assessment in breast cancer, which is important for treatment decisions ⁹. While these results are promising, DL is currently best viewed as a tool to assist radiologists rather than replace them.

What to ask your doctor

• Is deep learning used in the imaging center where I get my mammograms?
• How does the radiologist incorporate AI or deep learning results into my breast cancer screening?
• What are the limitations of deep learning for detecting breast cancer that I should be aware of?
• Are there any ongoing studies or trials at this hospital testing AI for breast imaging?
• If my mammogram is read by AI, will a radiologist still review the images?