feat: add modular tumor size and severity analysis(#20)

[Akash1510]

Description

This PR enhances the brain tumor detection pipeline by extending the system beyond basic detection and segmentation to provide detailed tumor analysis and diagnostic insights.

Previously, the model returned only:

• Tumor detection status (Detected / Not Detected)
• Segmentation mask output

This update introduces advanced analytical outputs to improve diagnostic interpretation, research usability, and future extensibility.

---

Features Implemented

1. Tumor Size Analysis

Added quantitative tumor measurements including:

• Tumor area percentage
• Tumor pixel count
• Relative image occupancy

This enables more precise estimation of tumor size within the scanned image.

---

2. Severity Classification

Implemented automatic severity categorization based on tumor occupancy:

• Small → < 2%
• Medium → 2–8%
• Large → 8–15%
• Critical → > 15%

This provides quick clinical-level severity interpretation.

---

3. Multi-Region Detection

Added connected component analysis to detect:

• Number of disconnected tumor regions
• Largest tumor component
• Presence of multiple tumor regions

This improves structural understanding of tumor distribution.

---

4. Shape & Spread Analysis

Implemented morphological analysis for:

• Shape complexity scoring
• Spread pattern classification
• Structural region analysis

This helps characterize tumor growth behavior and spatial spread.

---

5. Future Comparison Support

Prepared backend support for longitudinal scan comparison by adding:

• Comparison-ready analytical output
• Foundation for future tumor progression tracking

---

New Modular Analysis Structure

analysis/
├── tumor_metrics.py
├── severity_analysis.py
├── region_analysis.py
├── shape_analysis.py
└── tumor_analysis_pipeline.py

This modular design improves maintainability and reduces merge conflicts.

---

API Response Enhancements

The segmentation response now includes:

• Tumor metrics
• Severity classification
• Region analysis
• Shape complexity score
• Spread pattern
• Comparison readiness metadata

---

Benefits

This enhancement provides:

• Richer diagnostic insights
• Improved visualization metadata
• Cleaner modular backend architecture
• Better scalability for future features
• Stronger research and clinical analysis support

---

Testing

Verified through:

• Local execution testing
• Flask API validation
• Prediction pipeline testing
• Segmentation response verification
• Analytical output consistency checks

Closes #20