Home / Science / AI Spots Hidden Breast Cancer Clues
AI Spots Hidden Breast Cancer Clues
18 Aug
Summary
- AI platform CenSegNet analyzes microscopic patterns in breast cancer.
- Two distinct abnormalities in centrosomes were discovered by the AI.
- Enlarged centrosomes linked to aggressive tumors and lower survival.
A newly developed artificial intelligence platform, CenSegNet, is uncovering microscopic patterns within breast cancer tumors that were previously unseen. This innovation, stemming from research at the University of Southampton, could lead to more tailored therapies for patients by providing a deeper understanding of the disease's progression.
Centrosomes, critical for cell division, have long been recognized as a hallmark of cancer. However, their small size and fluid nature made them challenging to study in patient samples. CenSegNet analyzes hundreds of thousands of cells in tumor samples, revealing two distinct centrosome abnormalities: cells acquiring too many centrosomes and abnormally enlarged centrosomes.
Researchers discovered these defects can behave independently and are spatially distributed within tumors. Tumors with higher levels of enlarged centrosomes were found to be more aggressive, correlating with lower patient survival chances. This discovery opens avenues for developing new biomarkers and personalized treatment strategies.
