An Intelligent Healthcare System for Automated Alzheimer's Disease Prediction and Personalized Care

Alzheimer’s disease has posed the greatest threat among all the different types of neurodegenerative problems as it has assaulted humankind at quick pace than the others. Its manual revelation has become clinically insignificant because of the in expertise, high rate of false positives (FP) and false-negatives (FN). To reduce the false positive/ false negative rate, this paper frames a quick, affordable, and objective judgement of AD with a novel data mining method coalescing a global Maximum Relevance and Minimum Redundancy (MRMR) based ?lter heuristic with a globally optimised wrapper heuristic GANNIGMA with the intention of minimalising the consequence of an imbalanced healthcare dataset. The optimal feature subset yielding the best performance are utilised for model training of Decision Tree, k-NN, and SVM algorithms. The trial results on benchmark ADNI dataset using the proposed model displayed the Decision Tree attains TP rate of 0.778, and AUC of 0.798, k-NN acquires 0.764 TP Rate and 0.784 AUC, and SVM attains 0.997 TP Rate, and 0.996 as AUC. The results are far healthier than the separate results of these algorithms attained on the same dataset with fewer optimal feature subsets.