New Study Shows Artificial Intelligence Can Improve Breast Cancer Detection
A recent trial conducted in Sweden has shown that artificial intelligence (AI) can accurately detect 20 percent more breast cancers from mammograms compared to traditional screening by radiologists. The study, which is the first randomized controlled trial of its kind, published its interim results in the Lancet Oncology on Tuesday.
The trial involved over 80,000 women, with half of the participants having their mammograms analyzed by AI and a radiographer, while the other half had their mammograms examined by two radiologists without AI. The results indicated that using AI alongside radiologists was as effective as using two radiologists without AI, but led to a 20 percent increase in cancer detection. Additionally, the workload for radiologists was reduced by 44 percent, as they spent less time reading mammograms.
Lead author Kristina Lång from Lund University in Sweden noted that although the interim safety results were positive, further results from the trial are needed to determine whether the use of AI reduces the number of cancers detected between screenings. Lång also mentioned that the greatest potential of AI is to alleviate the burden on radiologists, allowing them to assist more patients.
The use of AI in healthcare settings, including medical imaging, has generated both excitement and concerns. While AI has proven beneficial in improving efficiency and accuracy, there are concerns surrounding algorithm training, validation, bias, and over-diagnosis. The European Union is currently planning stringent regulations for the use of AI, and the European Medicines Agency is assessing the risks and benefits of AI in drug development.
Stephen Duffy, professor of cancer screening at Queen Mary University of London, emphasized the importance of reducing the burden on radiologists’ time, but cautioned that over-detection of harmless lesions could be a potential concern when using AI.
This trial not only sheds light on the potential of AI in breast cancer screening but also highlights the need for randomized controlled trials to monitor the safety and effectiveness of AI systems in medical settings. AI has the ability to revolutionize healthcare, but rigorous testing and validation are necessary to ensure its reliability and accuracy.