Dementia Diagnosis through Artificial Intelligence in Medical Images

Dementia, a progressive cognitive impairment affecting memory and cognitive functions, poses significant challenges for timely diagnosis and intervention. This work presents a comprehensive methodology for predicting dementia by harnessing the power of image processing and machine learning techniques. The preprocessing stage employs Gaussian filtering to enhance image quality by reducing noise and preserving important structural information. This step prepares the input images for subsequent analyses with improved clarity and noise resilience. Subsequently, Otsu's segmentation technique is applied to the preprocessed images. Otsu's method optimally segments images into distinct regions based on variations in pixel intensity, aiding in the extraction of relevant regions of interest that contribute to dementia prediction. To capture intricate textural details indicative of dementia-related changes, gray-level co-occurrence matrix (GLCM) based feature extraction is employed. GLCM quantifies the spatial relationships between pixel values, providing valuable insights into image texture. This feature extraction process yields a high-dimensional feature space that encapsulates complex texture patterns within the segmented regions. The core of the proposed methodology lies in the Probabilistic Neural Network (PNN) based classification. PNNs are well-suited for pattern recognition tasks due to their ability to model complex relationships in data and provide probabilistic classification which achieves accuracy of 97.82%. By synergizing these techniques, the proposed methodology showcases promising results in accurately identifying dementia improving the quality of life for individuals affected by dementia.