Researchers at the University of California Southwestern Medical Center have developed a new artificial intelligence (AI) model that analyzes the spatial arrangement of cells in tissue samples to predict cancer patient survival. results published In the journal Nature.
Scientists explained that the spatial organization of cells in tissues is like a complex puzzle in which all the pieces are carefully placed together to form a complete structure (a tissue or organ). To make a diagnosis, tissue samples are usually taken from patients and placed on special slides to be interpreted by specialists. However, this process is time consuming. Additionally, the human eye may miss some details that are important in assessing patient survival. New artificial intelligence algorithms trained to analyze images of tissue sample slides mimic the work of doctors.
The system detects the cells in the images and determines their positions. The algorithms then separate the cells by type and examine their morphology (structure, size).
Researchers have successfully used this tool to predict outcomes of various types of cancer. They used the technology to distinguish two subtypes of lung cancer (adenocarcinoma or squamous cell carcinoma). In another study, they used new algorithms to predict the likelihood of potentially malignant oral lesions (pre-cancerous lesions) turning into cancer. In a third experiment, the model sought to determine which lung cancer patients were most likely to respond favorably to a class of drugs called epidermal growth factor receptor inhibitors.
In each of the three trials, the new technology significantly outperformed traditional methods in predicting disease outcomes. Scientists hope that in the future, algorithms can be used to optimize the choice of treatment for cancer patients and identify preventive measures for people predisposed to this disease.
Biologists before there is Evidence for one of the theories of aging.