CLOUD-BASED CNN FOR CYSTIC FIBROSIS DIAGNOSIS AND PREDICTION USING MEDICAL IMAGING

Cystic Fibrosis (CF) functions as a major genetic disease because it effectively harms both respiratory functionsand gastrointestinal tract functionality. The correct and immediate diagnosis plays an essential role in obtainingeffective patient treatment and proper management. The paper constructively demonstrates how integratingConvolutional Neural Networks with cloud computing framework enables efficient CF diagnostic operationsthrough image-based detection and severity evaluation. Expanded accuracy levels in classification are obtainedwhen the proposed approach extracts hierarchical deep features from CF-related image data sets using pre-trainedCNN layers. The proposed CNN model achieves superior results than the CART+PLS-SEM model throughcomparative evaluation because it reaches 99.35% accuracy with 95% precision and 96% recall that produces a98% F1-score. The research uses Adaptive Regularized Optimization techniques for both enhanced modelperformance along with cost-efficient operation and improved reliability dynamics. The study demonstrates CNNbased deep learning models perform better than other detection models for CF diagnosis which can operate inhealthcare environments both at the clinic and in cloud-based healthcare systems.