• Prachi Nag
• Kavita Rawat

The successful development of amyloid-based biomarkers and diagnostic tests for Alzheimer's disease (AD) marks a significant achievement in advancing AD diagnosis. However, two key challenges persist. First, amyloid-based biomarkers provide limited insight into the overall disease process, and second, they are ineffective in identifying individuals with AD before significant amyloid-beta accumulation in the brain occurs. This study aims to address these limitations by developing a method for identifying potential blood-based non-amyloid biomarkers for the early detection of AD. Blood is an attractive medium due to its accessibility and relatively low cost. Our approach utilizes machine learning (ML) techniques, specifically support vector machines, which are adept at building multi-variable models by identifying patterns within complex datasets. Through novel feature selection and evaluation methods, we identified five novel panels of non-amyloid proteins that show promise as early AD biomarkers. Notably, the combination of A2M, ApoE, BNP, Eot3, RAGE, and SGOT emerged as a potential key biomarker profile for early-stage AD. Detection models built using these panels achieved a sensitivity (SN) of over 99.27%, specificity (SP) exceeding 99.24%, and an area under the receiver operating curve (AUC) of at least 98.87% during the prodromal stage of the disease, with even higher performance at later stages. Comparatively, existing ML models demonstrated poor performance at this early stage, suggesting that the previously used protein panels may not be suitable for detecting AD in its early phase. Our findings underscore the feasibility of early Alzheimer's detection using non-amyloid-based biomarkers, offering a promising alternative to traditional amyloid-based diagnostics.

Alzheimer’s Disease, Blood Plasma Biomarkers, Bi-Directional Attention Mechanism, Deep Learning

"Early Detection of Alzheimer's Disease Using Blood Plasma Biomarkers with Deep Learning Techniques", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.11, Issue 10, page no.e543-e553, October-2024, Available :http://www.jetir.org/papers/JETIR2410459.pdf

"Early Detection of Alzheimer's Disease Using Blood Plasma Biomarkers with Deep Learning Techniques", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.11, Issue 10, page no. ppe543-e553, October-2024, Available at : http://www.jetir.org/papers/JETIR2410459.pdf