Automatic Detection and Classification of Respiratory Diseases Using Deep Learning

User Interface

Welcome Screen

The AuscultaTech Welcome Screen is designed to set the tone for an innovative and user-friendly experience. With a striking visual design and intuitive layout, the screen offers:

Navigation Options: Easy access to essential features like "Sign Up," "Log In," and "Check Your Health." Inspirational Tagline: "Transforming Respiratory Health with AI," emphasizing the platform's mission to revolutionize diagnostics. Call-to-Action Button: A prominently displayed "Check Your Health" button encourages users to start their diagnostic journey.

This welcoming interface reflects the platform's commitment to simplifying respiratory health diagnostics through AI-driven technology.

Feature Highlights

Here are some of the advanced features that make AUSCULTATECH a comprehensive Health platform:

Developer Team

Meet Our Developers

Our dedicated team is responsible for building and enhancing the VYC Verification Platform.

1. Manav Rai – Frontend Web Developer

   • Responsible for designing and developing the user interface for the web application.

2. Vishwas Kisaniya – Model Integration

   • Integrated the deep learning model into the web application, ensuring smooth communication between the backend and frontend.

3. Jashanpreet Singh Salh – Model Development

   • Developed the deep learning models to classify respiratory diseases based on lung auscultation sounds.

4. Pranshu Chauhan – Model Development

   • Worked on the deep learning models, focusing on data preprocessing, model training, and optimization.

---

Login and Signup Pages

Signup Page

New users can create an account by providing basic information such as their full name, email address, password, and uploading a profile photo. The signup page includes real-time validation to ensure the correctness of data.

• Signup Features:
  • Input fields for full name, email, password, and confirmation.
  • Real-time validation to prevent errors.
  • Secure password storage with encryption.

Login Page

The login page provides a simple and secure interface for users to access their accounts. Users must input their registered email and password to log in to the platform.

• Login Features:
  • Simple, user-friendly design.
  • Error handling for incorrect logins.

Dashboard Overview

The AuscultaTech Dashboard provides a seamless and intuitive experience for users. Upon logging in, users are presented with the option to upload their lung auscultation audio files for a quick health check. Powered by advanced deep learning models, the platform analyzes the uploaded audio and provides a detailed report, including:

Predicted Disease Class: Identifies the specific respiratory disease (e.g., COPD, pneumonia, or bronchiectasis). Prediction Confidence: Displays the accuracy level of the prediction. Waveform Graph: Offers a visual representation of the lung sound's waveform, enhancing understanding of the audio data. The dashboard ensures that healthcare professionals and individuals can receive rapid and accurate insights into respiratory health.

Project Overview

Respiratory diseases such as COPD, pneumonia, and bronchiectasis are leading causes of death worldwide. This project aims to improve the diagnostic process for respiratory diseases by utilizing deep learning techniques to automatically detect and classify lung auscultation sounds. By leveraging Convolutional Neural Networks (CNNs), the system is designed to aid healthcare professionals in making faster, more accurate diagnoses, reducing the potential for misdiagnoses and delays in treatment.

Our system processes lung sound data to identify key patterns and provides fast, accurate classification, helping to reduce misdiagnoses and delays in patient care.

Mission

• Improve Diagnostic Accuracy: Leveraging AI to classify respiratory diseases accurately from lung sounds.
• Faster Intervention: Helping healthcare professionals detect respiratory conditions early.
• Reduce Human Error: Automating the diagnostic process to ensure consistent results and faster decisions.

---

Key Features

• Automatic Classification: Detect and classify respiratory diseases like COPD, pneumonia, and bronchiectasis.
• AI-Powered: Utilizes Convolutional Neural Networks (CNNs) for high-accuracy classification.
• Real-Time Processing: Optimized for real-time lung sound analysis, making it suitable for clinical environments.
• Data Preprocessing: Includes bandpass filtering, resampling, and feature extraction for better model performance.

---

Technologies Used

• Deep Learning: Convolutional Neural Networks (CNNs)
• Python Libraries:
  • TensorFlow and Keras (for model building and training)
  • Librosa (for audio processing)
• Data Processing: Mel-Frequency Cepstral Coefficients (MFCC)
• Audio Format: WAV files for lung sound recordings

---

Contributing

We welcome contributions from the community! If you'd like to contribute, please follow these guidelines:

Steps to Contribute:

1. Fork this repository to your GitHub account.
2. Clone the forked repository:

   git clone https://github.com/coderiders22/Minor-Project-PEC-Chd.git
   git checkout -b feature-name
   git add .
   git commit -m "Description of your changes"
   git push origin feature-name