Utilizing Sequential Information of General Lab-test Results and Diagnoses History for Differential Diagnosis of Dementia

2025-02-21

AI Summary

Early diagnosis of Alzheimer’s disease (AD) is hindered by high inter-individual variability in biomarkers, limited access to specialized neuroimaging or cerebrospinal fluid (CSF) assays, and overreliance on single diagnostic indicators. To address these challenges, we propose a non-invasive, low-cost differential diagnostic framework leveraging routinely collected longitudinal laboratory test data. We innovatively model sequential lab test items as “sentences” and integrate word embeddings (Word2Vec/GloVe) with a hybrid multi-scale temporal architecture combining LSTM and Transformer modules. Key technical advances include temporal feature alignment and multi-task joint training to enhance robustness and generalizability. Crucially, our method eliminates dependence on neuroimaging or CSF analysis and demonstrates strong cross-institutional generalization. Evaluated on multicenter clinical cohorts, it achieves an AUC of 0.89 for early AD identification—significantly outperforming conventional approaches—and exhibits strong potential for scalable, interpretable, and low-barrier population-level screening.

Technology Category

• Machine Learning: Deep Learning Algorithms
• Humans and AI: Brain-Sensing and Analysis
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning

Application Category

• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

Early diagnosis of Alzheimer's Disease (AD) faces multiple data-related challenges, including high variability in patient data, limited access to specialized diagnostic tests, and overreliance on single-type indicators. These challenges are exacerbated by the progressive nature of AD, where subtle pathophysiological changes often precede clinical symptoms by decades. To address these limitations, this study proposes a novel approach that takes advantage of routinely collected general laboratory test histories for the early detection and differential diagnosis of AD. By modeling lab test sequences as"sentences", we apply word embedding techniques to capture latent relationships between tests and employ deep time series models, including long-short-term memory (LSTM) and Transformer networks, to model temporal patterns in patient records. Experimental results demonstrate that our approach improves diagnostic accuracy and enables scalable and costeffective AD screening in diverse clinical settings.

Problem

• Early diagnosis of Alzheimer's Disease using routine lab tests.
• Addressing data variability and limited specialized diagnostic access.
• Improving diagnostic accuracy with deep time series models.

Innovation

• Sequential lab-test data modeling
• Word embedding for test relationships
• LSTM and Transformer temporal modeling