Jin Peng, Yu Liu, Xiaohui Wang
Parkinson's disease (PD) involves not only dopaminergic degeneration but also pathological changes in cortico-basal ganglia-thalamocortical circuits and broader disease-relevant biological processes. Deep-brain neuromodulation has emerged as an important therapeutic strategy for motor dysfunction. Among the available approaches, deep brain stimulation (DBS) is the most established modality, whereas low-intensity focused ultrasound (LIFUS), a form of transcranial ultrasound stimulation, represents a promising but earlier-stage, non-invasive platform. This review discusses DBS and LIFUS from a shared mechanistic and translational perspective. Current evidence suggests that the two modalities may engage partially overlapping mechanistic domains associated with motor deficit improvement, including modulation of abnormal network activity, promotion of synaptic and axonal remodeling, attenuation of neuroinflammation and cellular stress, and possible interaction with α-synuclein-related pathology. At the same time, they differ substantially in evidentiary depth, clinical maturity and translational readiness. DBS remains the clinical benchmark, with durable motor benefits and an expanding mechanistic framework that now extends beyond circuit correction to neurotrophic, proteinopathic, neuroimmune and adaptive biomarker-guided mechanisms. By contrast, LIFUS offers non-invasive access to deep brain targets and shows encouraging pre-clinical effects on inflammation, apoptosis, synaptic integrity and neurovascular function, but its clinical evidence remains limited. Overall, deep-brain neuromodulation in PD should be viewed as a multi-dimensional therapeutic framework rather than a group of isolated technologies. Future progress will depend on tighter integration of circuit physiology, pathology-relevant biomarkers, model selection and standardized translational endpoints.