Abstract
Automated diagnosis can help detect, treat, and prevent threatening diseases. This project utilizes a convolutional neural network (CNN) built using the FastAI Library, OpenCV Library, and Google Colaboratory that classify Melanoma images as cancerous vs. noncancerous. Melanoma can become deadly in as little as 6 weeks. Thus, it is important for a solution to be developed that aids in the early-stage detection of cancer. The app allows the user to upload a picture. Once they do that, the image is fed to the Flask server and the CNN classifier returns a result (positive or negative). The app also features a comprehensive platform so users can access resources such as MayoClinic, cancer support centers, and other information. Due to its malignant nature and tendency to affect relatively young individuals, the timely diagnosis of Melanoma is very important.
About
Sneha Iyer is a Freshman at the University of Virginia. She is very passionate about computer science and planning a career in the research and application of software to solve real-world problems. Some areas she is interested in include CS Research, Artificial Intelligence/Machine Learning, Data Science, Mobile/Web app development, and Robotics.
​
She is very passionate about applying technology to solve problems and make an impact on the community and even on a global scale. Throughout high school and college, she has been involved in research and entrepreneurship competitions that center on designing and presenting creative innovations. Sneha is a Siemens Foundation semifinalist (Computer science). Another research program she has been involved with is the Aspiring Scientists Summer Internship Program at GMU where she researched Programming Strategies under Dr. LaToza.
​
Through this project, she has learned that there are many opportunities to develop deep-learning algorithms for a variety of real-life applications from automation to assistive and autonomous technologies. She hopes to continue her research and use computer science to solve modern-day issues and help people.
Slideshow
Technical Presentation Video
