AI Model Suggests Nearly 40% of Breast Cancer Patients Could Avoid Axillary Surgery

A recent study conducted at Lund University in Sweden has demonstrated the potential of an AI-powered system to transform breast cancer treatment. The innovative model analyzes routine mammograms—images already collected during diagnosis—and predicts which patients may safely forgo axillary surgery, a procedure used to examine lymph node spread. Traditionally, this surgery involves removing lymph nodes from the armpit to determine if cancer has metastasized, which can lead to side effects such as pain, swelling, numbness, and fluid buildup.

The AI model has been trained using mammograms from 1,265 women diagnosed with early-stage breast cancer between 2009 and 2017. Its sophisticated analysis considers the entire mammogram, not solely the tumor area, to assess the risk of metastasis with high accuracy. The study's findings suggest that approximately 41.7% of cases could potentially avoid invasive lymph node procedures if this model was implemented during diagnosis.

This technological advancement supports more personalized cancer care. Rather than routinely performing sentinel lymph node biopsies on all patients, clinicians could use the AI's risk assessment to determine whether surgery is necessary. For patients with a low predicted risk, surgery could be skipped after a thorough discussion, reducing physical and emotional burden while maintaining effective treatment.

The research leverages existing mammography data, which is already part of standard diagnostic workflows. The AI analyzes various image features—such as edges, textures, and shapes—and integrates patient data like age and tumor type to improve its predictions. This approach underscores the move towards less invasive, more individualized cancer management strategies.

While the results are promising, further validation with independent patient data across different populations is needed. Lund University researchers are actively seeking other datasets worldwide to confirm the AI's effectiveness universally.

Looking ahead, the integration of such AI models into routine mammography could revolutionize breast cancer staging, allowing for early and accurate risk assessments. This could facilitate tailored treatment plans that minimize unnecessary surgeries, ultimately improving quality of life for many patients. Moreover, ongoing studies aim to expand the model's capabilities to also predict prognosis based on mammogram patterns.

This research highlights a significant step forward in leveraging artificial intelligence to enhance precision medicine and reduce invasive procedures in breast cancer care.