Breast cancer is the most common cancer diagnosed in women worldwide, with an estimated 1.38 million women across the world diagnosed in 2008. It the most common cause of death from cancer in women worldwide, and incidence is generally high in the developed countries, including the UK, which is ranked at 11th out of 184 countries worldwide.
It has been a long journey for researchers and doctors to deliver better tests that can catch breast cancer in its infancy, enabling women to be diagnosed at an earlier stage, which improves their chances of survival.
It is in this space that technology has played a big part, scientist Mark Kane Goldstein told Computing at Capgemini's North America media event in New York earlier this week.
Goldstein is chairman of MammaCare, a US-based firm with a mission to provide breast examination proficiency for every clinician who examines women, and which has the support of the US National Cancer Institute and the US National Science Foundation (NSF).
To enable this to happen, MammaCare has been working steadily, and in 10 to 15-year periods big developments have occurred with the technology, which have enabled MammaCare to improve its testing methods.
The organisation believes that X-rays and other scanning technology do not always catch cancers at an early stage, particularly as many women don't get screened, and that the human finger holds the sensors necessary for nurses to find the cancer at an early stage.
Planting the seeds for a thorough screening
Goldstein explained that first, through research studies and labs, the company found that it could produce a model that could accurately reflect the tactile properties of the human breast and of breast cancers.
"What we then built was a series of sensor systems, so that the model breast could talk back and say what it was feeling – the problem was that, until recently, technologically, they were unstable," he said.
This was then developed over a decade, and thereafter a second model was built, but it too was unstable, partly because the embedded firmware couldn't handle real-time data that was coming from palpitations.
"The idea at the time was to record every palpitation because the software knows where the lesion is, but the clinician doesn't," he said.
Finally, a small company of experts from the Massachusetts Institute of Technology (MIT) built a more advanced sensor that was more stable.
"They built us an array, the design of which was consistent with the size of three fingers. What became more important was the ability of the sensor to detect pressure levels within the range of humans' ability to exert pressure with as many steps in the pressure detection as the human finger can transmit," Goldstein explained.
"By that I mean we had to build a digital mirror of the pressure core muscles that had to tell the computer the same thing that the finger was telling the brain – that took another ten years," he added.
Once the final sensor was ready, the software that was necessary to decode the data input evolved with it.
Initial tests showed that students, who could not find lumps, went online and worked on the machine to work through the required skills, and could then find lumps on the same women. These were on women who had benign lesions that felt the same as cancerous ones.
Thereafter MammaCare could upload the data of every lump found and every lump missed, monitor the variability and sensitivity of the hand and ensure that all breast tissues are covered and that there was an ability to reject a false positive.