P11 : DermAi Pro
| Students | Nicanor Uhart
Faisal Almasri |
| School | HDSB - Charles R. Beaudoin Public School - Burlington |
| Level | Junior 7/8 - Grade 8 |
| Group | Group 9 - Health Sciences II |
| Abstract | For our project, we developed an artificial intelligence tool designed to help detect skin cancer early and accurately. Using a deep learning framework, we trained a ResNet50 neural network on the HAM10000 dataset to classify seven distinct types of skin lesions. We wanted to move beyond just raw accuracy numbers, so we built a fully functional clinical dashboard. This interface features an explainability map that highlights the specific areas of the lesion the AI is analyzing, allowing medical professionals to actually verify the results. By addressing dataset imbalances and focusing on user experience, our goal was to create a practical and trustworthy tool that bridges the gap between computer science and real world dermatology. |