Google Health embarked on the mission to help people live their best and healthiest life. For this purpose, Google Health released its findings recently to improve the process of Breast Cancer screening with the help of Artificial Intelligence (AI).
Early diagnosis and spotting of breast cancer remain a challenge even after so many innovations. Digital mammography is used to perform tests on the patient; however, reading the reports is a task, which can lead to both true and false results.
As a result of false predictions, the detection process can delay the treatment that can lead to unwanted results. Also, this increases the burden on the radiologists as well as patients.
The results of Google Health’s finding were published in Nature (a Science Journal) that suggested the use of Artificial Intelligence for the prediction, diagnosis, and detection of breast cancer.
AI helped in identifying breast cancer through de-identified screening mammograms with high accuracy. Also, there were very few negatives and false negatives that beats today’s human experts in their job.
AI assist doctors to identify metastatic breast cancer from lymph node specimens through its deep learning algorithms.
Google Health has also collaborated with DeepMind, Royal Surrey County Hospital, Cancer Research UK Imperial Centre, and Northwestern University to come to a conclusion if AI can actually be a great technological development in the medical science for breast cancer screening.
From the experiments, the results showed a 5.7% reduction in false positives in the US and 1.2% reduction in the UK. Also, it showed a 2.7% and 9.4% reduction in false negatives, which is a great improvement.
- Related: Epileptic Seizure Prediction becomes much easier with the new Artificial Intelligence technology
One thing to note here is that the AI system did not have any patient histories or their previous mammograms, which doctors usually have. The model was only trained by de-identified mammograms of 15,000 women in the US and 76,000 women in the UK.
The study with six radiologists showed how AI system to beat human experts with it are under the receiver operating characteristic curve (AUC-ROC) better than that for radiologists.
More research is going on the technology and let’s see how machine learning tools can benefit experts in their medical field.
Photo: Lester Lefkowitz via Getty Images
Read next: Cell phone injuries on the rise – affecting youngsters significantly