To evaluate the effectiveness of smartphone video technology in assessing gait impairments in individuals with Parkinson's Disease (PD), highlighting the importance of accurate gait assessment for tailored treatment.
Key Findings:
The deep learning model closely tracks expert consensus on gait impairment, suggesting high reliability.
The model provides interpretable outputs that highlight movement features driving predictions, which could enhance clinician trust.
Smartphone-based assessments improve access to care and can be integrated into telehealth, potentially transforming patient management.
Interpretation:
The use of smartphone video technology for gait assessment in PD offers a promising, objective, and accessible method that may enhance patient care and treatment personalization.
Limitations:
Video analysis is episodic and may not capture at-home function, which is critical for comprehensive assessment.
Exclusion of patients with mobility devices limits generalizability, necessitating further research in diverse populations.
All videos were obtained by the research team, which may affect feasibility in routine practice and raise questions about real-world applicability.
Conclusion:
Smartphone video technology represents a significant advancement in gait assessment for PD, potentially improving treatment outcomes and access to care, while paving the way for future research into its broader applications.
