AI-driven risk prediction and categorization in cystic fibrosis leveraging AttentiveLSTM and Fox Wolf Optimizer.

Cystic fibrosis (CF), a genetic disorder stemming from CFTR gene mutations, requires accurate risk prediction to improve management. Modulator therapies have advanced treatment but remain limited, as they don't cover all gene variants and face accessibility issues. To address these challenges, a novel Cystic Fibrosis Risk Prediction Framework (CGRPF) is proposed. CGRPF utilizes mean imputation for missing data, the Fox Wolf Optimizer (FWO) for effective feature selection, and an AttentiveLSTM to capture temporal patterns in time-series data, aiding chronic disease prediction. Fully connected layers and a softmax layer enhance model performance and ensure calibrated classification into high, medium, and low-risk categories. Tested on the CFTR-2 dataset, CGRPF achieved strong performance metrics - 97 % accuracy, 91 % precision, 97 % recall, 93 % F1-score, outperforming state-of-the-art models in CF risk prediction.