Welcome to ChestVision AI, the cutting-edge web application designed to empower healthcare professionals with AI-driven chest X-ray analysis. Our platform uses advanced deep learning to provide rapid, accurate, and reliable detection of a variety of chest conditions, significantly improving diagnostic accuracy and speed.
For more information, explore our official website:
chestvision.us
At ChestVision AI, we are dedicated to pushing the boundaries of medical technology, combining artificial intelligence with medical imaging to provide real-time insights directly on your browser. Whether you are a healthcare professional or a researcher, ChestVision AI offers an indispensable tool for analyzing chest X-ray images.
Our platform’s key benefits include:
- 🌍 Accessible Anywhere: Accessible via web browsers, providing a responsive design for desktops, tablets, and mobile devices.
- ⚡ Real-Time Results: AI-powered analysis provides instant results with confidence scores, enhancing clinical workflow.
- 🔒 Secure & Private: All image analysis happens client-side, ensuring that your data remains private and secure.
- 🧠 AI Specialization: Our models specialize in detecting a wide range of chest-related conditions, including pneumonia, atelectasis, cardiomegaly, and more.
By utilizing state-of-the-art machine learning technologies like TensorFlow.js, our platform offers powerful diagnostics directly in your browser without the need for external servers, ensuring both speed and data privacy.
Discover the future of medical imaging with ChestVision AI.
Visit our website at chestvision.us to learn more and get started today!
- 🔍 Multi-Disease Detection: Capable of identifying 14 different chest conditions
- 🏥 Pneumonia Specialization: Dedicated model for pneumonia detection
- 📊 Real-time Analysis: Instant results with confidence scores
- 📱 Responsive Design: Seamless experience across all devices
- 🔒 Secure Processing: Client-side analysis ensuring data privacy
- Training: 78,468 images from ChestX-ray14 Dataset
- Validation: 11,218 images from ChestX-ray14 Dataset
- Testing: 22,432 images from ChestX-ray14 Dataset
Js
```const detectedConditions = [
"Atelectasis", "Cardiomegaly", "Effusion",
"Infiltration", "Mass", "Nodule",
"Pneumonia", "Pneumothorax", "Consolidation",
"Edema", "Emphysema", "Fibrosis",
"Pleural Thickening", "Hernia"
]
'''
- Clone the repository:
git clone https://github.com/2006coder/ChestVision_AI_HackPrinceton2024.git
- Open index.html in your browser or set up a local server:
python -m http.server 8000
Then, visit http://localhost:8000 in your browser.
Follow these simple steps to get started with ChestVision AI:
-
Select your preferred model
Choose between Multi-Disease Detection or Pneumonia Detection to tailor the analysis for your needs. -
Upload a chest X-ray image
Supported formats: DICOM, JPG, or PNG. Simply drag and drop the image into the uploader. -
Click "Analyze Image"
Let the AI work its magic and process the image. -
View the detailed analysis results
Get instant results with confidence scores for each detected condition.
This AI-assisted diagnostic tool is designed to supplement, not replace, professional medical judgment.
Always consult with qualified healthcare professionals for medical decisions.
- TensorFlow.js powers the machine learning models deployed directly in the browser, enabling real-time AI inference. This means your data never leaves your device, ensuring privacy and speed.
- The Multi-Disease Detection Model was trained by Bach using a complex residual network that predicts 14 different chest-related conditions from X-ray images. Here's a breakdown:
- Training Time: ⏱️ 3.68 hours
- Network Architecture: 🧬 Residual Network for efficient feature extraction.
- Dataset: 🖼️ ChestX-ray14 Dataset
- Training Set: 78,468 images
- Validation Set: 11,218 images
- Testing Set: 22,432 images
- This model is able to detect conditions like Atelectasis, Cardiomegaly, Effusion, and more!
- Particles.js creates stunning and interactive particle effects in the background. These beautiful animations enhance the overall user experience by responding to scroll and mouse movements.
- With Particles.js, the background remains lively and dynamic, engaging users without distracting from the core content.
- AOS adds smooth, engaging animations that trigger when the user scrolls through the page. This helps keep the user interface dynamic, making the app feel more fluid and interactive.
- Whether it’s fade-ins, slide-ins, or zoom effects, AOS ensures a captivating user experience that highlights important content as users scroll.
- Flexbox and CSS Grid were employed to create a flexible and responsive layout, ensuring the app looks great on all screen sizes, from mobile to desktop.
- Glassmorphism adds a frosted-glass effect, making the design feel modern and sophisticated with translucent backgrounds and soft blur effects.
- Together, these technologies provide a clean, minimalistic, and intuitive design that’s easy to navigate.
- Training Time: ⏱️ 3.68 hours
- Network Architecture: 🧬 Residual Network - A complex deep learning architecture optimized for high accuracy.
- Dataset: 🖼️ ChestX-ray14 Dataset
- Training Set: 78,468 images
- Validation Set: 11,218 images
- Testing Set: 22,432 images
- Supported Conditions:
This model can detect a range of chest-related diseases such as:- Atelectasis, Cardiomegaly, Effusion, Mass, Nodule, and more!
- Training Time: ⏱️ 22 minutes
- Network Architecture: DenseNet-121 💥 - A highly efficient and advanced deep learning network optimized for medical imaging.
- Dataset: 🖼️ Kaggle Chest X-ray Dataset
- Training Set: 5,863 images of chest X-rays labeled as Pneumonia or Non-Pneumonia.
- Output: This model provides a Pneumonia vs. Non-Pneumonia classification for X-ray images, offering a quick diagnosis for suspected cases.
| Member | Role | GitHub |
|---|---|---|
| Bach | ML Engineer | @2006coder |
| Samadhi | Web Developer | @samadhichandrasena |
| Sora | ML Engineer | @sowada23 |
Copyright © 2024 ChestVision AI. All rights reserved.