Scientists from the united states are using artificial intelligence to predict if the breast lesions identified from a biopsy become cancerous.
The machine learning system has been tested in 335 high-risk lesions, and correctly diagnosed 97% of the malignancy.
It reduced the number of unnecessary surgeries on more than 30%, the scientists said.
A breast cancer specialist said the research was “useful”.
The machine learning system was trained on the information about this type of injury, the system searches for patterns between a range of points of data, such as demographics, family history, biopsies and pathology reports.
“Because the diagnostic tools are so inaccurate, it is understandable that the trend of physicians across the screen for breast cancer,” said Regina Barzilay, MIT Delta Electronics Professor of Electrical Engineering and computer Science, and a breast cancer survivor herself.
“When there is a lot of uncertainty in the data, machine learning is exactly the tool we need to improve detection and avoid over-treatment.”First study
In the united states alone, 40,000 women die of breast cancer each year, but when you discover the cancer early enough it can often be cured.
Mammograms play a crucial role in the detection of this type of cancer, but also throw false positives, such as the lesions that appear suspicious.
Once operated, many of these injuries turn out to be benign.
“To our knowledge, this is the first study to apply machine learning to the task of distinguishing the high-risk lesions that require surgery from those that do not,” said Constance Lehman, professor of the Harvard medical School and chief of the Breast Imaging Division in the MGH Department of Radiology.
“We believe that this could help women make more informed decisions about your treatment, and that could give us more specific approaches to health care in general.”
Debashis Ghosh, a consultant breast surgeon based at the Royal Free hospital in London, said that the technology was good, but it may be of more use in the united states than in the uk.
“Here we have less than 5% of the patients who have these surgeries, while it is 30% in the united states.
“We try to make a definitive diagnosis before we operate, but this technology is certainly useful when there is a lack of experience.”
The research is being carried out by scientists from the Harvard Medical School, Massachusetts computer Science and Artificial Intelligence Laboratory, and Massachusetts General Hospital.