• Prachi Nag
• Kavita Rawat

The development of amyloid-based biomarkers and diagnostic tests for Schizophrenia (SD) marks a significant advancement in diagnostic tools. However, two key challenges remain: amyloid-based biomarkers offer only a limited view of the disease's progression, and they fail to detect SD before substantial amyloid-beta build-up in the brain. This study seeks to overcome these challenges by establishing blood-based, non-amyloid biomarkers for early SD detection. Blood is an appealing medium due to its accessibility and cost-effectiveness. Leveraging machine learning (ML), particularly transformers, we can develop multi-variable models that discern complex patterns within large datasets. Through innovative feature selection and assessment, we identified five promising non-amyloid protein panels for early SD detection. Notably, a biomarker profile composed of A2M, ApoE, BNP, Eot3, RAGE, and SGOT showed strong potential for early-stage SD detection. Models built with these panels achieved a sensitivity of over 99.28%, specificity above 99.73%, and an area under the receiver operating characteristic curve (AUC) of at least 98.87% in the prodromal stage, with even higher accuracy in later stages. By contrast, existing ML models have shown limited success in early SD detection, indicating that previous protein panels may not be optimal for early-phase diagnosis. These findings highlight the potential of non-amyloid biomarkers for early SD detection, offering a viable alternative to traditional amyloid-based diagnostics.

Schizophrenia Disease, Blood Plasma Biomarkers, Bi-Directional Attention Mechanism, Deep Learning

"Early Detection of Schizophrenia Disease Using Blood Plasma Biomarkers with Deep Learning", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.11, Issue 11, page no.b253-b263, November-2024, Available :http://www.jetir.org/papers/JETIR2411129.pdf

"Early Detection of Schizophrenia Disease Using Blood Plasma Biomarkers with Deep Learning", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.11, Issue 11, page no. ppb253-b263, November-2024, Available at : http://www.jetir.org/papers/JETIR2411129.pdf