🩺 DermaVision AI — AI-Powered Skin Disease Analysis & Clinical Decision Support System

DermaVision AI (PRCP-1027) is an enterprise-grade artificial intelligence healthcare web application designed to classify skin disorders into 6 distinct clinical categories with 98.65% accuracy. Developed by Sushameendra H, the application features a modern glassmorphic interface, interactive analytics hub, AI lesion scanner, dynamic PDF report generator, and developer portfolio.

---

👨‍💻 Developer Profile

• Developer: Sushameendra H
• Education: B.E. Computer Science & Engineering | Jain College of Engineering and Technology (2022–2026)
• GitHub: https://github.com/Sushameendra07
• Email: sushameendrah@gmail.com

---

🌟 Key Application Features

1. 🏠 Executive Dashboard:

   • Modern glassmorphism UI layout with dark/light background styling.
   • Real-time animated KPI cards displaying Accuracy (98.65%), Weighted Precision, Recall, and F1-Score.
   • Interactive Machine Learning workflow breakdown and disease category index.

2. 📊 Interactive Analytics Hub:

   • Reuses notebook Exploratory Data Analysis (EDA) visualizations powered by Plotly.
   • Clinical feature correlation matrix heatmap.
   • Interactive horizontal and grouped bar charts comparing 6 ML models:
     • Logistic Regression (Best Model - 98.65%)
     • Random Forest (97.30%)
     • Support Vector Machine (95.95%)
     • Decision Tree (94.59%)
     • K-Nearest Neighbors (94.59%)
     • Gaussian Naive Bayes (87.84%)
   • Confusion matrix heatmap and classification metrics breakdown.

3. 🔬 AI Lesion Scanner & Image Analysis:

   • Skin lesion image upload drop zone with animated CSS scan beam effect.
   • Interactive clinical and histopathological symptom parameter controls.
   • Real-time AI prediction returning disease category, confidence score, ICD-10 code, symptoms, and care guidelines.
   • Modular architecture ready for seamless integration with CNN deep learning image models.

4. 📋 PDF Clinical Report Generator:

   • Dynamic patient report compiler using ReportLab.
   • Exports printable PDF reports complete with patient info, prediction findings, confidence metrics, and medical disclaimer.

5. 👨‍💻 Developer Portfolio:

   • Premium developer showcase featuring technical skills, project highlights, GitHub, LinkedIn, and contact links.

---

📂 Project Structure

DermaVision_AI/
│
├── app.py                      # Main Streamlit application shell & sidebar navigation
├── requirements.txt            # Project Python dependencies
├── README.md                   # Complete application documentation
├── model.pkl                   # Trained Logistic Regression model
├── scaler.pkl                  # Trained StandardScaler preprocessor
│
├── assets/                     # Application visual graphics & media
│   ├── logo.png                # DermaVision AI brand logo
│   ├── hero.png                # Dashboard hero banner
│   └── profile.jpg             # Developer photo profile
│
├── pages/                      # Multi-page Streamlit application views
│   ├── 1_Dashboard.py          # Executive Dashboard & Metrics
│   ├── 2_Analytics.py          # Interactive EDA & Model Comparison
│   ├── 3_Image_Analysis.py     # AI Lesion Scanner & Diagnostic Assistant
│   ├── 4_Report.py             # Patient Report Generator & PDF Export
│   └── 5_About_Developer.py    # Developer Portfolio Page
│
├── utils/                      # Core backend utility modules
│   ├── load_model.py           # Cached model loader & auto-training script
│   ├── preprocessing.py        # Missing value treatment & input dataframe formatting
│   ├── prediction.py           # Model inference engine & confidence calculator
│   ├── visualizations.py       # Reusable Plotly visualizers
│   └── report_generator.py     # ReportLab PDF compiler
│
└── data/                       # Dataset & Metadata
    ├── dermatology.csv         # Dermatology clinical dataset (366 records, 35 features)
    └── disease_info.json       # Clinical descriptions & medical guidelines for 6 classes

---

📊 Model Performance Benchmarks

Machine Learning Model	Accuracy (%)	Precision (%)	Recall (%)	F1-Score (%)	Status
Logistic Regression	98.65%	98.78%	98.65%	98.66%	🏆 Selected Best Model
Random Forest Classifier	97.30%	97.45%	97.30%	97.32%	Evaluated
Support Vector Machine (SVM)	95.95%	96.12%	95.95%	95.98%	Evaluated
Decision Tree Classifier	94.59%	94.80%	94.59%	94.60%	Evaluated
K-Nearest Neighbors (KNN)	94.59%	94.75%	94.59%	94.60%	Evaluated
Gaussian Naive Bayes	87.84%	89.20%	87.84%	88.05%	Evaluated

---

⚡ Installation & Execution Guide

Prerequisites

• Python 3.10 or higher
• pip package manager

1. Clone or Navigate to Project Directory

cd DermaVision_AI

2. Install Required Dependencies

pip install -r requirements.txt

3. Launch the Application

streamlit run app.py

The application will open automatically in your browser at http://localhost:8501.

---

📄 License & Disclaimer

This application is developed strictly for educational, portfolio demonstration, and preliminary clinical decision support purposes. It does not replace professional dermatological diagnosis.