Breast cancer patterns found by AI could help treatment
Researchers at the University of Southampton have used artificial intelligence to identify previously unseen patterns inside breast cancers. Working with the city’s hospital, the team analysed more than 330,000 cells from 127 patients, with the aim of understanding features of tumours that may affect how the disease behaves and how it responds to treatment.
The study focused on centrosomes, which help cells divide properly and maintain their structure. Abnormalities in centrosomes have long been recognised as a hallmark of cancer, but the researchers said they have been difficult to study in patient tissue. Using an AI platform called CenSegNet, the team examined tumour samples at single-cell resolution across whole tumours, allowing them to see patterns that had previously been hard to detect.
One of the main findings was that two different kinds of centrosome changes had previously been treated as if they were the same. In fact, the study found that some cancer cells developed too many centrosomes, while others had centrosomes that were unusually large. The researchers said these changes could appear in different parts of the same tumour and may play different roles in cancer development.
The team also reported that tumours with higher numbers of enlarged centrosomes were more likely to show features associated with more aggressive disease. These included higher tumour grade, cancer that had spread to nearby lymph nodes and certain genetic changes. The study further found that patients with fewer enlarged centrosomes in the centre of their tumours tended to have better overall survival.
The researchers said the approach could help doctors better understand which cancers are more likely to grow, spread or resist treatment. It may also support the development of more personalised treatments by identifying tumours with specific weaknesses that could be targeted with new drugs. For now, though, the technology is not ready for routine use in hospitals. The team said it has also shown promise on tissue samples from other parts of the body, including the kidney, colon and appendix, and that CenSegNet is being made available free of charge as open-source software.
In short, the study suggests that AI can reveal detailed patterns inside breast cancer tissue that may help identify higher-risk patients and point toward more targeted treatment strategies.
Sources:
Naj Modak