Enhancing Alzheimer's Disease Diagnosis: A Hybrid Model of Transfer Learning and SVM for Improved Classification Accuracy

Alzheimer's disease, marked by a debilitating decline in cognitive function due to neuronal loss, remains without a definitive cure. We have developed an innovative automated system to analyze brain Magnetic Resonance Imaging (MRI) scans, recognizing the importance of early detection in mitigating the disease's impact and improving patient care. Our objective extends beyond the simple detection of dementia; we are committed to accurately classifying the disease into its various progressive stages. To this end, we have tailored the AlexNet convolutional network specifically for this purpose by applying transfer learning and configuring it to train on unsegmented MRI images.Our model underwent rigorous testing on the widely used KAGGLE dataset, where it achieved an impressive 95.32% accuracy rate in multi-class classification tasks. This high level of accuracy is pivotal, especially considering the nuanced stages of Alzheimer's disease. The model's effectiveness is also highlighted by its performance across multiple diagnostic categories, achieving a specificity of 96.23%, a precision of 87.42%, a recall of 95.32%, and an F1-score of 83.65%. These metrics are particularly significant as they demonstrate the model's ability to not only identify the presence of Alzheimer's but also to differentiate between its stages with high reliability. The ability of our model to deliver such precise diagnostic outputs holds promise for significantly enhancing the strategic planning of treatment regimens and potentially paving the way for therapeutic advancements. By providing a reliable method for early and accurate stage classification, our approach could facilitate more targeted interventions, ultimately contributing to better patient outcomes and a deeper understanding of Alzheimer's progression. This innovative diagnostic tool could thus play a crucial role in the ongoing battle against Alzheimer's disease, providing clinicians and researchers with a powerful ally in their efforts to combat this challenging condition.