You often hear about (somewhat exaggerated) articles about how bots and especially AI will be replacing human workers in the not so distant future. That is a possibility but one that is still far from becoming reality, especially when life-critical decisions have to be made. Still, it’s quite impressive how far AI has come, especially when it manages to outdo professionals in accuracy, like when detecting breast cancer.
The threat of cancer looms over the entire population of the world. While doctors race for a more permanent cure, the best chances women have against breast cancer is early detection. That is often performed through X-ray imaging or digital mammography but the actual process of deciphering scans can be painstaking and error-prone.
That’s where Google’s latest AI research comes in. It collaborated with its DeepMind sibling company, Cancer Research UK Imperial Centre, Northwestern University, and Royal Surrey County Hospital to train neural networks and AI models to detect signs of breast cancer from de-identified mammograms. Suffice it to say, the results were quite impressive.
Trained on data from US and UK patients, the model produced 5.7% (US) and 1.2% (UK) reduction in false positives as well as a 9.4% (US) and 2.7% (UK) reduction in false negatives. These numbers are in relation to the success rates of radiologists trained specifically in this line of cancer detection. And this is from AI that will not tire.
Of course, that’s not to say doctors are being put on notice. Such an AI is still at a very early research stage and is hardly certified to be used in real-world situations. Such systems will most likely be used to augment rather than displace the work of healthcare professionals, acting more as assistants rather than inhuman replacements.