Train the Artificial Brain Diagnosis of Brain Tumors using Neural Network
Roma Vivek Pradhan
Friendswood, TX
Grade: 12
The early and accurate diagnosis of brain tumors is extremely critical to the patients' survival. About 13,000 people die from brain tumors every year in the US alone, with it being the second leading cause of death from cancer in children and young adults. In my project, I trained a neural network, an "artificial brain", to recognize features in brain tissue biopsy images such that the diagnostic result is an accurate and efficient delineation of malignant brain tumor from the benign one. The accuracy of my tool is over five times better than the manual image-based diagnosis conducted by neuropathologists.