认知
仪表板
决策支持系统
机器学习
计算机科学
人工智能
临床决策支持系统
过程(计算)
认知模型
疾病
心理学
医学
临床决策
二元分类
深度学习
临床实习
预测建模
认知障碍
医疗保健
物理医学与康复
电子健康档案
心理干预
回归分析
认知评估系统
认知心理学
作者
Gurusigaamani Ayyanar Muthulingam,Poornima Nagaraj,R. Rajesh Kanna,Sakthi Sanjay S,Vijayakumar M,Kesani Rohith
出处
期刊:
[Institute of Electrical and Electronics Engineers]
日期:2025-11-19
卷期号:: 1283-1288
标识
DOI:10.1109/icoiics67115.2025.11390403
摘要
AD) Alzheimer Disease is a serious concern on development since it is a progression of (MCI) Mild Cognitive Impairment that is important to diagnose and intervene in early stages. This article describes a deep learning system using transformers to use longitudinal cognitive scores to predict the evolution of AD. The model gives two-output predictions; a binary classification to convert MCI to AD and regression to predict time-to-conversion. Based on ADNI data, the model is used to capture temporal dependencies of cognitive decline. Transformer model has been chosen due to its theoretical advantages over recurrent models (LSTM/GRU) in the process of modeling long-range dependencies. The model constructs an interactive clinical dashboard that provides time predictions and real-time actionable risk scores to clinicians. Having an AUC of 0.94 and 91.7% sensitivity, the model can contribute to personal care planning and decision-making and offer a promising approach to precision medicine in AD care. The paper contributes to AI in healthcare (SDG-3), providing an evidence-based instrument in the early detection of AD that enables clinical decision-making.
科研通智能强力驱动
Strongly Powered by AbleSci AI