Researchers from Washington State University have developed an artificial intelligence system that can recognize signs of disease in human tissues. It has been recorded that this system detects even pathologies missed by doctors.
Researchers from Washington State University have developed an artificial intelligence system that can recognize disease symptoms from images of human tissues.
This system uses scalable deep learning to identify infections and surpasses human capabilities in terms of both time and accuracy.
A new study published in Scientific Reports has revealed a new deep learning model that efficiently identifies errors.
This new approach could speed up disease-related research, which requires hours for pathologists to identify samples.
“COULD REVOLUTIONIZE MEDICINE”
Scientist Michael Skinner said, “This AI-based deep learning program was very, very accurate when looking at these tissues. This AI for both animals and humans could revolutionize medicine, essentially making these types of analyses easier.”
DETECTED MISSED PATHOLOGIES
To add extra input to the AI, the researchers fed it more gigapixel images, including breast cancer and lymph node metastases.
During testing, the researchers found the new AI model to be more efficient and accurate than previous systems.
Interestingly, it even detected pathologies that doctors had missed.
“I think we now have a faster and more accurate way to identify diseases and tissues than humans,” said Holder, one of the study’s authors.
