试析多模态数据融合在针灸调节糖尿病周围神经病变神经传导速度中的应用前景OA

Acupuncture has shown notable therapeutic benefits in diabetic peripheral neuropathy(DPN),in which improvement of nerve function is a major outcome of interest.Nerve conduction velocity is widely regarded as the gold standard for quantitative assessment of peripheral nerve conduction and is extensively used for diagnosing neuropathy and evaluating treatment response.However,when applied to acupuncture efficacy evaluation,nerve conduction velocity alone is often insufficient to comprehensively capture multilayered effects including symptom relief,nerve repair,and metabolic remodeling,resulting in a mismatch between clinical needs and evaluation capability.Multimodal data fusion,by integrating heterogeneous data sources such as electrophysiology,clinical phenotypes,imaging,and metabolic data,provides a methodological basis for building cross-dimensional association models.Accordingly,this paper systematically reviews clinical and experimental studies on acupuncture intervention for diabetic peripheral neuropathy and its regulation of nerve conduction velocity.From a hierarchical perspective of multimodal fusion,we discuss feasible analytical pathways for incorporating nerve conduction velocity together with metabolic profiles,neural structural information,and symptom scales.We further examine the potential utility of multimodal data in diabetic peripheral neuropathy subtype stratification and in optimizing individualized acupuncture protocols.This work aims to provide a reference for efficacy evaluation and protocol optimization of acupuncture treatment for DPN.