AI model makes it easier to remove a cancerous tumor
Researchers have taught an artificial intelligence model to predict whether surgeons removed all cancerous tissue during surgery to remove a breast tumor.Researchers at the University of North Carolina (UNC) School of Medicine have developed an artificial intelligence model that can predict in real time whether a surgeon removed all cancerous tissue during surgery to remove a malignant breast tumor. The model performed as well or better than doctors.
To "teach" the artificial intelligence model what the "negative" and "positive" borders of pathologic tissue looked like, the researchers used 821 mammography images taken immediately after resection. They then compared the data with specimen reports from pathologists. Just over half (53%) of the images had positive margins. They also provided the model with patient demographics such as age, race, tumor type and size.
It turned out that the AI model showed a sensitivity of 85%. This is a measure of how well the model can detect positive cases.
The researchers say that compared to the accuracy of human interpretation, the AI model performed just as well, if not better, than humans.
The study is published in the journal Annals of Surgical Oncology.