Advancements in High-Performance Computing in Early Diagnosis of Alzheimer's Disease: A Systematic Review

Abstract

Background: The key histopathological features of Alzheimer's disease (AD) include extensive brain shrinkage and amyloid plaques made up of amyloid beta (Aβ) deposition, as well as neurofibrillary tangles (NFTs) made up of hyperphosphorylated microtubule-associated Tau protein. The existence of AD is primarily a complicated biological and neurodegenerative illness, and various computer modeling techniques to overcome the complexity of the disease are needed. Moreover, the study of brain disorders, large-scale brain models, or connectomes, depends heavily on high-performance computing (HPC). Objectives: This systematic review intends to answer five key questions about the function of the HPC in the early diagnosis of the AD, as the best contender for dealing with the high-dimensional problems. This review also intends to review the previous articles on the HPC for the early diagnosis of the AD and the computational performance solutions in the literature, which could be used for increasing the diagnostic speed and the research output. Methods: The previous articles on the topic are found using the following search codes: ALL= ("parallel processing"AND "Alzheimer's disease"OR "high performance computing"AND "Alzheimer's disease"OR "parallel computing"AND "Alzheimer's disease"OR "parallel processing"AND "Alzheimer"OR "high performance computing"AND "Alzheimer"OR "parallel computing"AND "Alzheimer") between the years 2014 and 2023. The scientific literature analysis based on the visualizations is conducted using the VOSviewer program. Using both the inclusion and exclusion criteria, 298 results, which include 196 open access papers, 71 papers with enriched cited references and 12 review articles, are analyzed. Results: According to the findings of the classifications, AD is the most frequent cause of dementia globally, and there is currently no specific therapy that slows its course. However, the HPC could provide some rapid computations of various modeling methodologies, allowing each aspect of the brain to be seen. Some institutes are attempting to simulate prospective remedies for the disease. Conclusion: Various approaches are integrated to be utilized in the literature with the help of the HPC in this review, and to enlighten readers about the current status of the HPC technologies for the early detection and treatment of the AD across the world. The high-performance algorithms and approaches for the AD are major scientific challenges that may encounter several issues as well. © 2013 IEEE.