A team from the University of Calgary, in Canada, has managed to create a new software solution that can detect breast cancer at very early stage, even before the development of a tumor. The solution was created by engineers from the university's Schulich School of Engineering, along with radiologist Dr. Leo Desautels.
The team reviewed 106 mammograms from women who later on developed breast cancer. However, these women received a clean bill of health when they underwent initial screening. The new software solution was used to check those mammograms once again. The researchers managed to determine suspicious lesions that did not appear when the mammograms were reviewed in the first time.
On average, these lesions were detected 15 months before tumors or other cancer signs were identified clinically. Raj Rangayyan, the lead author in the research, commented "There is cancer there, but no tumor yet" adding "There is no mass or lump. . . . What we are identifying is what we call 'architectural distortions.' "
Dr. Rangayyan said that in normal breast, the tissues, blood vessels, and lymph ducts are all converging towards the nipple. However, when cancer is developing changes start to take place. For example, some breast tissues get pushed and pulled in other directions, forming what is called "architectural distortions" that appear on mammograms.
The problem with distortions is that they cannot be detected easily even by an experienced radiologist using the naked eye. Moreover, these distortions can be serious as a previous study indicated that 12-45% of misdiagnosed breast cancers are due to inability to identify architectural distortions.
Dr. Steven Narod, a professor at the University of Toronto's Women's College Research Institute and a Tier One Canada Research Chair in Breast Cancer, said that the new solution is offering a new option against breast cancer.
He explained "We can switch to another screening method like an MRI, and the problem there is they're very expensive, or we can improve the quality of the mammogram itself through more advance screening techniques," He added "Or we can get computers to read the mammograms better than the human eye . . . So I think what they're doing here is very interesting and exciting."