The research team used pathology images to train algorithms to detect the localization of breast cancer that had spread to lymph nodes adjacent to the breast. In its current form, the localization score for the deep learning algorithm reached 89 percent — higher than the 73 percent score pathologists tend to see, even without time constraints.
Since pathologists are often under pressure to review and diagnose patient tissue samples with limited time, the researchers hope that this algorithm can relieve some of this work. The researchers also noted that pathologists often disagree on cancer diagnosis, with agreement for some forms of breast cancer being as low as 48 percent.
“These algorithms perform well for the tasks for which they are trained, but lack the breadth of knowledge and experience of human pathologists,” according to the blog post. ” To ensure the best clinical outcome for patients, these algorithms need to be incorporated in a way that complements the pathologist’s workflow.”
More articles on health IT:
Epic unveils App Orchard for third-party developers
Office of Management and Budget: 4 assessments on HHS’ cybersecurity performance in 2016
Former Mayo Clinic IT executive named North Dakota CIO
At the Becker's 11th Annual IT + Revenue Cycle Conference: The Future of AI & Digital Health, taking place September 14–17 in Chicago, healthcare executives and digital leaders from across the country will come together to explore how AI, interoperability, cybersecurity, and revenue cycle innovation are transforming care delivery, strengthening financial performance, and driving the next era of digital health. Apply for complimentary registration now.