Asian Journal of Research and Reports in Neurology

Huntington’s disease is a progressive neurodegenerative disorder characterised by motor, cognitive, psychiatric, and functional deterioration. Conventional assessment using genetic information, clinical scales, neuroimaging, and molecular biomarkers provides important information but may not adequately capture subtle longitudinal changes or interindividual variability in disease progression. This review examines the potential role of artificial intelligence in integrating multimodal biomarkers for monitoring and forecasting disease progression. Relevant data sources include CAG repeat length, clinical and functional assessments, neuroimaging, neurofilament light chain and other molecular biomarkers, together with digital measures derived from speech, smartphones, wearable sensors, video, and computerised cognitive assessments. Machine-learning and deep-learning approaches may support the integration of these heterogeneous data for disease classification, longitudinal monitoring, identification of progression patterns, and estimation of motor, cognitive, and functional decline. Digital biomarkers are particularly relevant because they permit repeated assessment in real-world settings and may complement periodic clinic-based examinations. However, routine clinical translation remains limited by small and heterogeneous datasets, incomplete longitudinal follow-up, differences in devices and data-collection protocols, overfitting, limited external validation, model interpretability, privacy concerns, and regulatory requirements. Artificial intelligence should therefore function as decision support rather than replace clinical assessment. Larger standardised longitudinal studies, independent validation, transparent models, and appropriate data governance are required before multimodal artificial-intelligence approaches can be reliably integrated into Huntington’s disease care and clinical research.