This archive report was first published on 2 January 2020.
Early detection of breast cancer is crucial for effective treatment and improved patient outcomes. In Britain, women over 50 are advised to undergo regular mammograms, which are then reviewed by two independent experts.
However, interpreting mammograms can be prone to errors, resulting in false positives or false negatives. A study published in Nature found that a small percentage of mammograms return incorrect results, with 5.7% of cases in the US and 1.2% in Britain being misdiagnosed, and 9.4% of cases in the US and 2.7% in Britain being missed.
Researchers at Google Health have developed an artificial intelligence model that can detect breast cancer with similar accuracy to expert radiographers. The model was trained on thousands of breast scans from Britain and the US, which had already been reviewed by doctors in real-life settings.
The AI model showed a significant reduction in incorrect diagnoses, with 5.7% of cases in the US and 1.2% in Britain being misdiagnosed, and 9.4% of cases in the US and 2.7% in Britain being missed. The technology has the potential to act as a 'second opinion' for cancer diagnoses, supporting and enabling expert radiographers to provide the best possible outcomes for patients.
According to Dominic King, UK lead at Google Health, "The earlier you identify a breast cancer, the better it is for the patient. We think about this technology in a way that supports and enables an expert, or a patient ultimately, to get the best outcome from whatever diagnostics they've had."
The study also found that using AI to verify the first human expert reviewer's diagnosis could save up to 88% of the workload for the second clinician, freeing up time for more critical tasks.
King noted, "Find me a country where you can find a nurse or doctor that isn't busy. There's the opportunity for this technology to support the existing excellent service of the (human) reviewers."
