Rash Decisions

• 

  detector page 2

• 

  detector page

• 

  welcome page

Inspiration

Rash Decisions was born out of the need to bridge the gap in accessible healthcare. We wanted to empower individuals with the ability to take charge of their skin health, especially in situations where quick access to dermatologists is challenging or costly. By leveraging AI and the Teachable Machine API, we aimed to make reliable skin rash identification available to anyone with a webcam, promoting early intervention and informed decision-making.

What it does

Rash Decisions leverages AI to identify skin rashes in real-time using a webcam. Users simply point their camera at the affected area, and our system provides a rapid diagnosis, offering potential insights into rash type and treatments. Through Rash Decisions, we aim to...

• Reduce Rash Misidentification: By offering accurate AI-assisted diagnoses, we aim to lower the rate of misidentifying skin conditions, ensuring individuals receive appropriate guidance and care.
• Increase Healthcare Accessibility: Rash Decisions addresses the accessibility gap in healthcare by enabling people to assess their skin health without the need for immediate medical appointments. This is especially valuable in situations with limited access to healthcare services.
• Prevent Self-Misdiagnosis: It minimizes self-misdiagnosis by offering users a reliable resource for understanding their skin conditions and making informed decisions.
• Enable Early Detection: It allows for the early identification of skin rashes, enabling users to address potential skin conditions before it becomes severe.
• Increase Health Awareness: It encourages individuals to take a proactive approach to their skin health.

How we built it

We used Google's Teachable Machine API, incorporating machine learning models to teach our system to recognize and classify various skin rashes. Our locally hosted website offers an intuitive user interface and seamless integration with the webcam, ensuring a user-friendly experience.

Challenges we ran into

Balancing data reliability and quantity for training our model presented a significant challenge. We needed a substantial dataset with accurately labeled images of different rashes. To address this, we began by sourcing images from trusted medical sources for a solid foundation. Then, we expanded our dataset with images from diverse internet sources and datasets, maintaining a strong emphasis on data quality and precision throughout the collection process.

Accomplishments that we're proud of

• Successfully integrating Google's Teachable Machine API into our project.
• Displayed an integrated webcam to allow users to engage with our AI detection software in real time.
• Developing an intuitive user interface for a seamless user experience.
• Achieving high accuracy in rash classification and diagnosis.

What we learned

During the development of Rash Decisions, we gained invaluable experience in machine learning, AI model training, and web development. We also deepened our understanding of dermatological conditions and their visual characteristics.

What's next for Rash Decisions

Our vision for Rash Decisions goes beyond the hackathon. We plan to refine and expand our model's capabilities, incorporating a wider range of skin conditions and enhancing its accuracy. In the future, we aim to provide users with personalized treatment plans based on severity of the skin condition and potentially partner with healthcare providers to offer telemedicine consultations for more serious cases. Rash Decisions is poised to revolutionize the way people approach skin health, making it more accessible and convenient for everyone.

Built With

• css
• html
• javascript
• teachable-machine-api