Intelligent Computational Models for Alzheimer's Disease Classification

Medical image is a type of digital image processing that has widened its

boundaries over the decades. It is a primary source for the visual representation

of a human system's internal organs for clinical interpretation and intervention.

The growth and developments in this field have triggered the development of

novel algorithms for data/ image processing.

The medical imaging modalities are medical imaging and processing,

Structural Magnetic Resonance Imaging (sMRI), Functional Magnetic

Resonance Imaging (fMRI), Digital Mammogram, Ultra Sound (US), Positron

Emission Tomography (PET), Electrocardiogram (ECG), and Computer

Tomography (CT) that have promoted research in many allied fields including

Geographical Information Systems (GIS), Forensic Science, Business Data

Analytics and Astronomy.

Digital Image is mathematically represented as a twodimensional

function, f(x, y) where x and y are the spatial coordinates, and the

amplitude f at any given pair of coordinates (x, y) is called intensity. Digital

Image Processing (DIP) is the subject area that deals with processing/

manipulating a digital image by a digital computer to obtain useful and necessary

information from it. The different elements of a DIP system include image

acquisition, image storage, and image processing.

Medical image processing is an offshoot of DIP, which involves using

technology to process the human body's medical images. Medical imaging is also

referred to as diagnostic imaging. Medical imaging aims to visualize the human

body's inner parts, which is required to identify internal abnormalities such as

broken bones, tumors, and leaking blood vessels. In medical imaging,

segmentation aims to study anatomical structure, demarcate and distinguish a

region and help in therapeutic planning and treatment.

Recent medical image diagnosis encompasses the advancement of the

latest technology for acquiring the medical images by different modalities used

for diagnosis, treatment, and research. Medical image processing covers a broad

array of operations categorized as low-level, mid-level, and high-level image

processing based on computation complexity. Medical image segmentation

contributes to analyzing human anatomy and diagnosing the disorders, using

automated or semi-automated computational systems. The foremost objective of

medical image segmentation is to extract the regions to study the intrinsic

anatomical structure and estimate the severity of abnormalities that help in

deciding the means and methods of treatment. These aspects motivate

researchers to develop competent computing techniques for pre-processing,

segmentation, and analysis. These techniques' novelty is commonly ascertained

by three metrics, namely, precision, recall, and F1-Score.