Secure and Transparent Lung and Colon Cancer Classification Using Blockchain and Microsoft Azure

Area of Science:

• Oncology
• Computer Science
• Health Informatics

Background:

• Global healthcare systems face diagnostic and management challenges for lung and colon cancers.
• Traditional diagnostic methods are inefficient and raise data privacy concerns.

Purpose of the Study:

• To develop a secure and transparent framework for remote cancer diagnosis using blockchain and Microsoft Azure.
• To leverage advanced machine learning models for accurate classification of lung and colon cancers.

Main Methods:

• Integration of Microsoft Azure cloud services with a permissioned blockchain network.
• Secure data handling including anonymization, encryption, and controlled access via smart contracts.
• Utilizing Azure Machine Learning for training and deploying models on the LC25000 histopathological image dataset.

Main Results:

• Achieved 100% accuracy in lung and colon cancer classification using DenseNet, ResNet50, and MobileNet models.
• Demonstrated exceptional performance with F1-scores and cross-validation accuracies exceeding 99.9%.
• Enhanced diagnostic process efficiency and transparency through real-time notifications and secure remote consultations.

Conclusions:

• The proposed framework effectively enhances diagnostic accuracy and data security for lung and colon cancer.
• Blockchain and Azure integration offers a robust solution for secure remote consultations and cancer care management.
• The study highlights the potential of advanced technologies in improving patient outcomes and streamlining healthcare processes.