Rush student develops a way of speeding up the detection of breast cancer using Artificial Intelligence
Stephen Flynn from St Joseph’s Secondary School, Rush with their BT Young Scientist project, Sláinte: The use of artificial intelligence to detect Breast Cancer.
A student from St Joseph’s Secondary School in Rush believes he had found a way to more quickly identify breast cancer that could be vital in clearing the backlog in cancer detections that has built up during the pandemic.
The project is entitled “Sláinte - A.I. Breast Cancer Detction” and was taken on by Rush student, Stephen Flynn.
Stephen embarked on a project using image classification and a self sufficient data engine to detect breast cancer in x-rays and mammograms for faster breast cancer screening and detection.
Stephen first identified the problem he was trying to address and said: “At this very moment, there is a massive cancer detection backlog that is estimated to take 10 years to clear.”
He began to think of ways the process of cancer detection might be speeded up and thought Artificial Intelligence or AI might provide the answer.
Beginning with a database of x-rays and mammograms, he set to work training an AI programme to detect the more common forms of cancer, particularly breast cancer.
He found that due to the quality of these black and white images, the accuracy of the AI in identifying cancers was lower than he expected.
He then switched his attention to a much larger database of slides and images taken after biopsy and began training the AI on these much shaper images.
The results were staggering, delivering a detection accuracy of between 88 per cent 94 per cent.
Stephen believes the programme he has developed can significantly speed up cancer detection rates but not satisfied with that, he has left the programme as an open source project, and invited other developers to see if they can improve it.
The hugely impressive project has obvious real world application and could help clear the backlog of cancer detections that has built up during the pandemic, in a much shorter time-frame.
