2 minute read

Dr Sally Cockburn

Next Article
Family dining

Family dining

Dr Sa y Cockburn GIVES HER UNIQUE VIEW ON ALL MATTERS MEDICAL

ARTIFICIAL INTELLIGENCE SET TO IMPROVE BREAST CANCER RISK IN YOUNG WOMEN Over the years I have had many patients frustrated after being told a mammogram is of little use to them as they have dense breasts. Some misinterpreted this as age discrimination but it’s not.

The reason is that dense breasts can make an accurate interpretation of breast cancer almost impossible. Radiologists have told me it’s like trying to locate something in a snowstorm.

It’s hard to reassure someone when you just don’t have information. For women at risk like this we rely on regular physical examination and ultrasound, and in some cases MRI.

But women obviously want to know their risk and living with an unknown, especially when you have close relatives with breast cancer, is agonising.

Well, there is something new on the horizon: A team of international researchers, led by Australia’s own Professor John Hopper from The University of Melbourne in collaboration with BreastScreen Victoria and Cancer Council Victoria, has developed some very smart software that is metaphorically like using fog lamps rather than high beams in a snowstorm.

Dense breast tissue is not necessarily a bad thing in itself, but we sometimes just can’t tell what’s going on with current mammograms. Dense breasts are chockers with glandular tissue that produces milk, which wanes later in life largely replaced by fat (let’s celebrate that there is something positive about mature breasts where a cancer may be easier to spot).

This is why the mammogram program starts at 50 — because it’s a very useful screening test in less dense breasts. However, we clinicians have always been uneasy simply reassuring women with dense breasts where glandular tissue shows up as white clouds on mammogram, when really we just don’t know because we can’t see whether something is behind the cloud. We know although rare, breast cancer happens in young women and we want to pick it and know the warning signs.

Researchers have long been looking for other ways

AD

to interpret breast cancer risk in women with dense breasts on mammograms. It was recognised that some women with dense breasts and more of these bright white patches on mammograms seemed to go on to develop cancer. Then again, lots of women didn’t, so the trick would be to develop a way to strip away the layers of white noise on the mammogram to see what’s underneath. This has only recently been made possible through digital subtraction technology and mammograms have only relatively recently been performed using digital technology. This technology was not possible on the old hard-copy photo-fi lm system.

The team has been working on semi-automated computer programming that measures brightness on mammograms. Using artifi cial intelligence, the computer peels back layers on the mammogram to reveal unseen texture in these cloudy mammograms that might well predict breast cancer. These measures are called cirrocumulus and cirrus (yes, as in clouds!).

Our current mammogram screening programs around the country are still vitally important and work well for most. This is an additional, tailored form of screening, personalised for a group of women who otherwise may not know their risk. It isn’t needed for those in whom mammograms already give enough information.

This is still in research phase so isn’t available just yet — but watch this space!

FOR MORE ON WOMEN’S HEALTH VISIT HOUSEOFWELLNESS.COM.AU

This article is from: