Related Experiment Video
Updated: Jun 15, 2025
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
Advancing the early detection of canine cognitive dysfunction syndrome with machine learning-enhanced blood-based
Chae Young Kim1, Jinhye Kim2, Sunmi Yoon2
1BK21 Four program, Department of Medical Sciences, Soonchunhyang University, Asan, Republic of Korea.
Abstract:
Up to half of the senior dogs suffer from canine cognitive dysfunction syndrome (CCDS), the diagnosis method relies on subjective questionnaires such as canine cognitive dysfunction rating (CCDR) scores. Therefore, the necessity of objective diagnosis is emerging. Here, we developed blood-based biomarkers for CCDS early detection. Blood samples from dogs with CCDR scores above 25 were analyzed, and the biomarkers retinol-binding protein 4 (RBP4), C-X-C-motif chemokine ligand 10 (CXCL10), and NADPH oxidase 4 (NOX4) were validated against neurodegenerative models. Lower biomarker levels were correlated with higher CCDR scores, indicating cognitive decline. Machine-learning analysis revealed the highest predictive accuracy when analyzing the combination of RBP4 and NOX4 using the support vector machine algorithm and confirmed potential diagnostic biomarkers. These results suggest that blood-based biomarkers can notably improve CCDS early detection and treatment, with implications for neurodegenerative disease management in both animals and humans.
