Artificial intelligence researcher, engineer, physician and professor of pediatrics, pathology and laboratory medicine with the Institute for Computational Health Sciences at UC San Francisco, Dexter Hadley is applying big data analysis to solve a critical challenge: accurately screening and detecting breast cancer.
He and his colleagues at Hadley Labs are developing precision medicine techniques for doctors and their patients to help unravel the mechanisms of the disease in order to reduce morbidity and mortality.
Their network is designed to train AI algorithms to detect breast cancer and classify tumors far more accurately than doctors, by using GPU-accelerated deep learning that will be able to scan and analyze millions of images crowdsourced through their project initiative, breastwecan.org.
Hadley’s novel approach to improving detection is to obtain personal medical data from 5 million women by using Bitcoin’s blockchain technology. By crowdsourcing clinical imaging, he and his team hope to catch breast cancer before it progresses, and greatly improve diagnosis and management for women who will undergo treatment.
By using blockchain technology, the network will allow participants to store their data securely. Participants will also have full control over their data, granting permission to share, restrict or revoke access at any time.
While 1 in 8 or over 250,000 women are diagnosed with breast cancer annually, another five million will receive phone calls from their doctors that something looks amiss in their initial mammograms. Incorrect screenings will mean that these women will have to submit to additional testing. What ensues is stress, lab work, biopsies, more waiting and more stress. Another 1 in 4 breast cancers will go undetected by the mammograms, despite nearly 40 million mammograms being performed annually in the US.
Women are encouraged to visit the online portal breastwecan.org where they can sign up to participate in the research and accelerate efforts to train the neural network.
Don't Miss a Beat – Subscribe to get email alerts delivered directly to your inbox