-
-
landing page
Inspiration
The inspiration came from the communication gap between patients and healthcare providers, especially during cancer diagnosis and treatment. The goal was to create a platform that not only aids medical detection but also ensures patients stay informed and connected with their doctors.
What it does
The app uses deep learning to detect lung and colon cancer from histopathology images and provides an intuitive interface for doctors to share diagnostic insights with patients. It bridges medical communication, offering real-time updates, progress tracking, and simplified result interpretation.
How we built it
The system was developed using Convolutional Neural Networks (CNNs) — including DenseNet121, InceptionV3, and MobileNetV3Large — trained on the LC25000 dataset. The web-based interface was built with HTML, CSS, and JavaScript, while Flask served as the backend API to handle image classification and communication flow.
Challenges we ran into
Integrating accurate cancer detection models with an easy-to-use communication system was complex. Ensuring real-time updates, maintaining data privacy, and optimizing model performance for web deployment also posed technical challenges.
Accomplishments that we're proud of
Successfully developing a unified platform that combines accurate histopathological image analysis with effective medical communication. The system’s interface promotes transparency between doctors and patients, improving trust and treatment understanding.
What we learned
We learned how powerful AI can be in bridging gaps beyond detection — enabling smoother, more empathetic communication in healthcare. We also gained experience in model optimization, web deployment, and user experience design.
What's next for Medical Communication App
Future plans include integrating electronic medical records, multilingual support, and mobile compatibility. There’s also a vision to expand detection to other cancer types and incorporate AI-driven patient assistance for early diagnosis and monitoring.
Log in or sign up for Devpost to join the conversation.