Introduction: Rest tremor is a common symptom of neurological disorders such as Parkinson's disease (PD) that implies complications in swallowing function, such as oropharyngeal dysphagia. Objective: To explore the temporal and spectral characteristics of laryngeal one quantitatively analyze tremors in structures involved in swallowing observable in Fiberoptic Endoscopic Evaluation of Swallowing (FEES), using artificial intelligence Methods: A descriptive retrospective cross-sectional study was conducted, utilizing videos from a database of FEES at a Dysphagia Outpatient Clinic. Neural network training was performed using the DeepLabCut software with a pre-trained ResNet-50 neural network. Movement data from arytenoid points were analyzed in the time and frequency domains to identify tremor patterns associated with Parkinson's Disease (PD). Results: Individuals with PD associated tremors exhibited significantly higher signal crossings (16.42 vs. 13.00 on the X-axis, 16.62 vs. 13.00 on the Y-axis), amplitude peaks (270.42 vs. 32.48 on the X-axis, 69.06 vs. 9.75 on the Y-axis), and power spectral density (PSD) values (9.75 Hz vs. 2.92 Hz on the X-axis, 4.17 Hz vs. 0.41 Hz on the Y-axis) compared to those without tremor. Conclusion: The findings suggest that AI can effectively identify arytenoid laryngeal tremors associated with PD, offering valuable insights for personalized therapeutic planning in dysphagia management. Future research should aim to expand the dataset to enhance the generalizability of AI-based tremor detection methods.