Table 1 Summary of related works on medical image analysis and disease detection techniques.

EfficientNet, Deep Learning²⁵

COVID-19 Detection from X-ray Images

Achieved high accuracy through transfer learning to address data scarcity during the pandemic.

Integrating feature extraction and selection techniques, exploring MRI and CT scan data.

Machine Learning (SVM, DT, KNN, ANN), Data Preprocessing²⁶

Diabetes Prediction

Optimized data preprocessing to tackle data imbalance, achieved 95.5% accuracy.

Investigating feature selection, ensemble methods, and deep learning techniques.

Blockchain Technology²⁷

Data Privacy and Secure Data Transfer

Enhanced public access to medical records with privacy architecture to address security concerns.

Improve privacy mechanisms and integrate with advanced AI models.

Deep CNN, Blockchain³¹

Pediatric Radiology

Developed open-source software for pediatric chest X-ray classification, addressing privacy and security through blockchain.

Enhance communication efficiency and address free-riding and model poisoning attacks.

Blockchain with Combined Learning³⁵

Federated Learning

Reduced single-point failures by using decentralized blockchain for model verification.

Implement personalized models for different healthcare settings.

Combined NAS (FedNAS), Neural Architecture Search⁴⁰

Federated Learning Automation

Improved model integration from local ML models, highlighting the challenge of non-IID data.

Improve compatibility and optimization for non-IID client data.

Capsule Network (CapsNet)³⁸

COVID-19 Detection from X-ray Images

Outperformed CNN-based models on small datasets, offering enhanced robustness and performance.

Fine-tune model hyperparameters and explore multi-modal data fusion.

Automated Combined Learning, FedNAS⁴¹

Federated Learning

Demonstrated the limitations of default local ML model parameters in combined learning environments.

Develop adaptive parameter optimization strategies.

Blockchain and AI Integration¹⁴

Vaccine Distribution and Monitoring

Enhanced traceability and monitoring of vaccine distribution during the pandemic.

Integrate predictive analytics for proactive decision-making.

Machine Learning (SVM, DT, KNN, ANN)²⁰

Parkinson’s Disease Recognition

Compared various ML models for accurate disease recognition, achieving high accuracy with specific algorithms.

Incorporate deep learning models for feature extraction and classification.

Privacy-Preserving Architecture, Blockchain³⁰

Privacy and Data Security

Developed a robust architecture to protect data privacy while enabling secure communication.

Integrate lightweight cryptographic methods for faster data processing.

Blockchain with Federated Learning³³

Secure Medical Imaging Transmission

Improved data transmission security and resilience against model poisoning and free-riding attacks.

Optimize communication cost and latency during federated learning.