Researchers at the University of Regina have created a computer system which could help improve the accuracy of breast cancer diagnoses by providing a digital “second opinion”.
Radiologists use mammogram readings to help diagnose breast cancer, but misdiagnoses have happened, leading to lawsuits from patients and delayed treatment of tumours.
- Quebec mammogram review finds 109 errors
- Quebec to review 15,000 mammograms
Esam Hussein is part of a U of R research team which has developed a system aimed at preventing human error when radiologists are reading mammograms.
The technology trains a computer to read mammograms using a database of correctly diagnosed readings, providing radiologists with a digital ‘second opinion’.
Mammogram mistakes2:36
“If the computer diagnosis agrees with the radiologist, the chance that the diagnosis is correct becomes higher,” said Hussein.
“If there is a disagreement, the radiologist can take another look.”
Hussein told CBC Radio’s Saskatoon Morning he hoped it would reduce the chance of error, reassure patients and give radiologists peace of mind when making a diagnosis.
It is possible for breast cancer to be misdiagnosed due to human error.
The researchers are now looking for an opportunity to put the new computer system to work.
They say it has a 92 per cent accuracy rate.
“We look forward to hearing from any radiologists in town and the province or anywhere really, who wants to work with us and try to bring this technology into the field and make it useful.”