Tuesday, October 22, 2013
Teen Develops Computer Algorithm to Diagnose Leukemia
Brittany Wenger isn't your average high-school senior: She taught the computer how to diagnose leukemia.
The 18-year-old student from Sarasota, Fla. built a custom, cloud-based "artificial neural network" to find patterns in genetic expression profiles to diagnose patients with an aggressive form of cancer called mixed-lineage leukemia (MLL). Simply put, this means Wenger taught the computer how to diagnose leukemia by creating a diagnostic tool for doctors to use.
Since artificial neural networks are programs that model the brain's neurons and their interconnections, Wenger told Mashable that they "can actually learn to detect things that transcend human knowledge."
Mixed-lineage leukemia generally has poor prognosis, and the five-year survival rate is only 40%. Since Wenger said "different types of cancer have different molecular fingerprints," she discovered four particular gene expressions in the body that can be targeted to create MLL-specific drugs. Not only did she create a powerful diagnostic tool for this cancer, but her findings might also help develop new treatments.
Wenger, who is graduating soon from The Out-of-Door Academy in Sarasota, previouslyused artificial-intelligence technology to diagnose breast cancer. With a non-invasive procedure, her technology was able to help determine whether a breast mass was malignant or benign. Wenger's new findings with leukemia prove that her Cloud4Cancer service (check it out, here) can be altered to improve diagnostics for multiple cancer classifications.
Images courtesy of Intel
Post courtesy of Mashable