Objective Assessment of Facial Expressions in Parkinson's Disease During Deep Brain Stimulation ON and OFF States: A...

Parkinson's disease (PD) affects both motor and non-motor functions, including facial expressivity. While subthalamic nucleus deep brain stimulation (STN-DBS) is effective for motor symptoms, its impact on facial expression remains unclear. This study aimed to objectively assess the impact of DBS on facial movement timing and amplitude using objective facial landmark-based motion analysis.

We analyzed 10 PD patients with STN-DBS (≥ 3 years) and 10 age- and sex-matched controls. Standardized video recordings were obtained using a fixed camera setup. Facial movements were analyzed using objective facial landmark-based motion analysis implemented via a pre-trained facial landmark detection framework (Mediapipe), extracting 468 facial landmarks across approximately 1800 frames per recording. Key facial parameters, including smile completion time and region-specific ranges of motion (ROM), were quantified.

DBS significantly increased oral commissure and frontal ROM (p < 0.05) while reducing smile completion time (p = 0.003). However, PD patients' smiles remained slower than those of controls (0.64 ± 0.21 s vs. 0.49 ± 0.09 s; p = 0.019), indicating incomplete normalization of facial expressivity. Medial eyebrow and palpebral movements showed no significant differences (p > 0.05).

STN-DBS enhances the amplitude of some facial movements but does not fully restore temporal dynamics of facial expression. Persistent delays in smile execution suggest that aspects of facial motor control may remain impaired despite effective stimulation. Objective facial landmark-based motion analysis provides a sensitive tool for detecting subtle alterations in facial motor dynamics and may support future studies investigating complementary strategies to optimize motor expressivity in PD.