Article Preview
Top1. Introduction
Advancement in biological science with modern technologies has given us large volumes of clinical and biological data, including analytical diagnosis images of diseases. These data help scientists understand human health, diseases conditions, and required drug design for specific disease. Regardless of the ongoing advancement in biomedical research, a more significant part remains unused, whereas, for example, the brain image data can immensely help to identify Alzheimer's patients' stages of health condition. Alzheimer disease (AD) is one of the neurological disorders such as Parkinson's disease (PD), Prion disease, Motor neuron diseases (MND), Huntington's disease (HD), Amyotrophic lateral sclerosis (ALS) ( Eric E Hall, 2014; Sarraf & Tofighi, 2016). These major neurological disorders have mutual cellular and molecular mechanisms resulting in structure plus protein aggregation or formation occuring over the disease's progression ( Nussbaum & Ellis, 2003; Ross & Poirier, 2004) . It is a progressive disease, and over time it worsens the working function of the brain. The main risk factors of AD are family history, age, and genetics; if anyone has Alzheimer's in the family history like immediate family members, there is a high probability of getting AD. Some genes are involved in AD-like apolipoprotein E (APOE), and other forms of APOE will appear on late-onset AD. Again, amyloid precursor protein (APP), Presenilin 1 (PSEN1), Presenilin 2 (PSEN2) will appear in early-onset Alzheimer's disease (Giri et al., 2016).
In the late AD stage, the surface layer covers the cerebrum, where it perishes and decreases the mind's biggest aspect. This harm to the cortex plays destruction with the mind's ordinary capacity to prepare, review, and concentrate. Alzheimer's sickness likewise influences the hippocampus, which assumes a significant part in memory. The malady makes the hippocampus wilt. This damages the mind's capacity to make new recollections (Farooq et al., 2017). Postmortem biopsy is one of the determination processes of AD depends on histological information. Also, Mini-Mental State Examination (MMSE) tests, Mild Cognitive Impairment (MCI), which cannot help that much to identify actions for a temporary period between normal aging and AD through a cognitive test, because the patients do not have extreme memory issues (Dubois et al., 2007; Leandrou et al., 2018). Clinical tests like Frontotemporal lobar degeneration (FTLD) are used to separate different subtypes of dementia, which is also not certain due to the right imaging biomarkers (Muñoz-Ruiz et al., 2012; Moller et al., 2016 ). Brain MRI Imaging plays a significant role in exploring the AD affect people using the brain's image. The entorhinal cortex and the hippocampus are the two most basic regions of interest (ROIs) utilized in both in vivo and postmortem examinations on pathophysiology in AD. Moreover, it is not sufficient to locate the misfolded protein or tissues in the brain with a visual subjective evaluation of magnetic resonance imaging (MRI), and quantitative measures are basic for the evaluation of the disease.
Over the past decade, Artificial Intelligence (AI) techniques have become a promising field in biomedical sciences to medical imaging. Research scientists are working to take care of complex clinical issues like neurodegenerative diseases. Machine learning (ML) and deep learning (DL) techniques are so powerful in feature extraction, prediction, and classification. The DL helps computational models made of various layers to extract the information in the most efficient way. The DL techniques like Convolutional Neural Network (NN) have improved the way of researching computer vision areas. It utilizes the backpropagation calculation to show how the weights are processed in each layer from the past layer (Lecun et al., 2015).