According to the results of research carried out by scientists at the University of California at San Francisco (UCSF) and the Mayo Clinic, software aimed at evaluating breast density could predict the risk of breast cancer just as well as conventional mammograms.
A study published on 1 May in the 'Annals of Internal Medicine' has shown that BI-RADS (Breast Imaging Reporting and Data System) software designed by New Zealand company Volpara aimed at measuring the density of breast tissue can provide results as good as those obtained by radiologists when it comes to evaluating the risks of breast cancer, says Health Data Management.
Potential solution to lack of examination standards
At the moment, US women can have regular breast checks in 30 states, but there is an issue with reliability linked to the lack of standardisation of the examinations. The results can therefore vary according to the radiologist and the type of mammograms, explains a concerned Karla Kerlikowske, professor of medicine at UCSF.
"You have these compulsory examinations, but there is no standard for interpreting them," says Professor Kerlikowske, a researcher who codirected the study with Céline Vachon, professor of epidemiology at the Mayo Clinic. "The advantage of an automated measure is that it remains constant from one medical establishment to another and is easier to reproduce."
Extremely accurate results based on density of breast tissue
According to the study, involving 1,609 women suffering from breast cancer, Volpara’s software is reported to have demonstrated that patients with heavy-density breast tissue were 5.65 times more likely to develop an interval cancer (one which appears between two examinations) than those with a normal breast tissue density.
"Extremely high predictability rates," is how a delighted Professor Kerlikowske terms them. "One of the interesting aspects of our study is that [the measuring of] breast density gives a better predictability of interval cancers or those which have gone undetected than you would get from a conventional mammogram."
According to the professor, using his new fully automated method of examination would also enable radiologists to save precious time – "an hour a day" – if they do not have to interpret the results of mammograms themselves, according to a radiologist who tested the software.
Contact Allianz Partners
Mar 15, 2018
DeepMind – the Google subsidiary dedicated to artificial intelligence –and the Moorfields Eye Hospital in London have developed a form of IA which can detect anomalies on retinal sc [...]
Nov 28, 2017
A big step towards the personalisation of drug treatments for cancer has recently been made by US researchers, who have designed an algorithm equipped with artificial intelligence aimed [...]