Back to Research

PhD Thesis Summary

Advances in Deep Learning for Medical Image Segmentation and Classification

Overview

My doctoral research addressed critical challenges in medical image analysis through the development of innovative deep learning architectures and methodologies. The work focused on improving accuracy, computational efficiency, and generalization capabilities of AI models for medical diagnostics, with particular emphasis on breast ultrasound imaging and tumor detection.

Key Research Contributions

EfficientU-Net: Parameter-Optimized Medical Image Segmentation

Developed EfficientU-Net, a novel deep learning model integrating EfficientNet-B7 and atrous convolution into the U-Net architecture, optimizing breast tumor segmentation and classification in ultrasound images with reduced computational complexity and enhanced accuracy.

  • 13x reduction in parameters (1.31M vs. 17.27M in U-Net)
  • Superior segmentation of malignant tumors with irregular shapes
  • Enhanced boundary localization with adaptive receptive fields
  • 97.905% accuracy in tumor classification (benign, malignant, normal)
  • Validated on two public datasets using 5-fold cross-validation

2. UMA-Net with Adaptive Loss Functions

Developed UMA-Net, an advanced U-Net variant integrating residual connections, attention mechanisms, and atrous convolutions, enhanced by a dynamic ensemble loss function to optimize medical image segmentation across diverse datasets.

  • Residual connections and attention blocks enhance feature integration and focus on critical regions
  • Atrous convolutions enable multi-scale feature capture without compromising resolution
  • Dynamic ensemble loss (BCE, Dice, Hausdorff, Tversky) adapts weights for balanced optimization
  • Achieves superior generalization across five breast ultrasound datasets (BUET, BUSI, Mendeley, OMI, UDIAT)

3. Fuzzy Rough Set Loss for Boundary Precision

Introduced a novel loss function based on fuzzy rough set theory to handle boundary uncertainties in medical images.

Innovations:

  • Enhanced sensitivity to uncertain predictions and ambiguous lesion boundaries
  • Reduced computational complexity while improving segmentation accuracy
  • Effective handling of irregular shapes and overlapping edges in medical imaging
  • Novel similarity functions for better uncertainty handling in boundary regions

4. Deep Learning and Genetic Algorithm-based Ensemble Model for Feature Selection and Classification

Developed a unified approach integrating MobileNet for feature extraction, Genetic Algorithms (GA) for feature selection, and an ensemble model for classification, optimizing medical image analysis with enhanced accuracy and efficiency.

  • MobileNet as an optimal feature extractor for medical images with minimal parameters
  • GA-based feature selection to navigate complex feature spaces effectively
  • Novel ensemble model with soft voting for robust classification decisions
  • Addresses overfitting in limited data scenarios, enhancing model generalizability

5. Saliency-Guided AttentionNet (SGAN)

Designed a dual-branch architecture leveraging Grad-CAM saliency maps for breast ultrasound classification.

Performance Metrics:

  • 90.51% accuracy on multi-center validation
  • 87.95% F1-score and 94.08% AUC across five datasets
  • Explicit foreground-background decomposition for lesion and peritumoral analysis
  • Adaptive attention fusion maintaining transfer learning benefits with minimal parameter increase

Impact and Validation

The research was comprehensively validated across multiple medical imaging modalities including ultrasound, MRI, CT scans, and various anatomical regions. Key achievements include:

  • Cross-dataset validation demonstrating robust generalization (78.46% accuracy across held-out datasets)
  • Computational efficiency suitable for clinical deployment and real-time processing
  • State-of-the-art performance with significant improvements over baseline methods
  • Clinical applicability addressing real-world challenges in resource-limited healthcare settings

Broader Significance

This work bridges the gap between theoretical deep learning advances and practical medical applications, providing computationally efficient solutions that maintain high accuracy while being deployable in clinical environments. The research contributes to the democratization of advanced medical AI, particularly benefiting healthcare systems with limited access to expert radiological interpretation.

The developed methodologies have been published in top-tier conferences and journals, with open-source implementations available to facilitate further research and clinical adoption.

Back to Research