Contribution of AI in diagnosing and treating cancer

Cancer of unknown primary sites often causes pleural and peritoneal effusions, and cytological analysis by pathologists using traditional methods can be inaccurate.  Researchers in China used advanced artificial intelligence (AI) methods to study 57,000 cases of fluid buildup in the chest and abdomen, where the origin of the cancer was initially unknown but later identified.

They tested the AI system on nearly 30,000 additional samples to see if it could accurately distinguish between benign (non-cancerous) and malignant (cancerous) cells and determine where the cancer started. They found the AI to be highly accurate, matching the skill level of experienced pathologists and outperforming less experienced ones. When junior pathologists reviewed their diagnoses after seeing the AI's results, their accuracy improved significantly.

Using AI predictions in treatments also showed promising results. Patients whose treatments matched the AI's tumour origin predictions lived longer (27 months on average) compared to those whose treatments did not align with the AI's suggestions (17 months).

In summary, AI technology not only helps pathologists make more accurate diagnoses but also improves patient outcomes by guiding more effective treatments.