Southampton researchers' CenSegNet distinguishes 2 centrosome abnormalities in breast cancer
Researchers at Southampton have built an AI tool called CenSegNet that's changing how we look at breast cancer.
By scanning tissue samples from 127 patients (over 330,000 cells), the AI found two distinct types of cell abnormalities (enlarged or multiple centrosomes) that were previously lumped together.
These patterns may play different roles in how cancers develop.
Enlarged centrosomes associated with poorer survival
The study showed that tumors with more enlarged centrosomes tend to be more aggressive and likely to spread. Patients with fewer enlarged centrosomes in the center of their tumors had better overall survival.
Dr. Salah Elias, who led the research, said CenSegNet's detailed analysis could help doctors spot high-risk patients and personalize treatments.
While the tool isn't ready for hospitals yet, it can be used on tissue samples from other parts of the body, including the kidney, colon, and appendix.