Sailfish Optimization based Enhanced Dementia Detection Using Faster R CNN Architecture

Abstract :- Due to its degenerative nature, which makes early diagnosis challenging, dementia especially Alzheimer’s has become one of the major global health concerns. Finding the true causes of dementia is essential to implementing the right kind of treatment. In order to automatically extract features that contribute to some signal of brain shrinkage and other pertinent dementia biomarkers for more accurate and early diagnosis, the suggested model makes use of deep learning techniques in a CNN. The model’s performance is improved by using publically accessible datasets with sophisticated preprocessing and data augmentation approaches, such as the Australian Imaging, Biomarkers & Lifestyle dataset and the Alzheimer’s Disease Neuroimaging Initiative. CNN’s promise in early dementia diagnosis is demonstrated by our tests, which show a significant improvement in detection accuracy when compared to superior findings that can be obtained using well-known classical methods like SVM. The CNN model’s capacity to generalize over a wide range of patients and imaging situations is enhanced by our method, which not only increases detection accuracy but also tackles the problem of sparse datasets through data augmentation and transfer learning. Our studies’ results show that, in comparison to more conventional machine learning techniques like SVM and Random Forests, we achieve superior accuracy, sensitivity, and specificity. The model’s capacity to distinguish between Alzheimer’s disease, MCI, and normal cognitive states makes it a potentially reliable tool for identifying dementia in its early stages and tracking its progression. The results validate CNN’s efficacy in medical imaging and its ability to assist clinicians in diagnosing dementia.