Abstract

Background:
Gait and postural impairments in patients with Parkinson’s disease (PD) severely affect daily activities and are important targets for treatment.

Aim:
This study aimed to analyze gait and posture in PD-model rats during the beam test using DeepLabCut (DLC), a deep neural network for motion tracking.

Methods:
Ten rats with unilateral 6-hydroxydopamine (6-OHDA) lesions and five sham-operated rats were tested. The beam test was used to evaluate foot slips, execution time, head and trunk inclination, and stride length using images captured from frontal, lateral, and top views.

Results:
The 6-OHDA group exhibited significantly longer task times. Frontal-view images revealed a greater head tilt in the PD group than in the SHAM group. Lateral-view images showed a shorter stride length, and top-view images indicated larger mean and maximum trunk inclination angles in the PD group than in the SHAM group.

Conclusion:
DLC successfully detected these gait and posture abnormalities, indicating that it is an effective method for evaluating motor function in PD-model rats. This approach may serve as a reliable tool for assessing the therapeutic effects of PD.

Key words: Beam test; DeepLabCut; Gait analysis; Image analysis; Parkinson's disease.