Abstract
Background
The sit-to-stand (STS) task is a fundamental functional movement requiring substantial lower-limb joint torque and coordinated neuromuscular control. The knee joint plays a dominant role in generating extensor torque during the body elevation phase. However, existing exoskeleton assistive strategies often lack alignment with the temporal characteristics of joint mechanical demand, limiting their effectiveness.
Methods
Thirteen healthy male participants performed STS tasks under both non-assisted and exoskeleton-assisted conditions. A pressure-based motion intention recognition method was developed using anterior thigh pressure signals to detect movement onset. Three assistive torque levels (3.0 Nm, 4.5 Nm, and 6.0 Nm) were applied to the knee joint. Kinematic and kinetic data were collected using a motion capture system and force plates, and knee joint moment, power, and mechanical work were calculated to evaluate biomechanical changes.
Results
Recognition-performance analysis showed that the pressure-threshold algorithm identified the target pre-extension trigger in all six validation recordings, corresponding to a positive-event detection rate of 100% (6/6). The detected trigger occurred at 0.49 ± 0.05 of the pressure-rising phase, and the onset of assistive torque occurred within 0.072 s of knee-extension onset, advancing torque output by approximately 0.228 s compared with a kinematics-only trigger. Because the validation trials included intentional STS attempts only, specificity and overall classification accuracy against non-intention events were not estimated in the present study. Knee joint moment exhibited a clear phase-dependent pattern, characterized by a rapid increase following seat-off and a peak during the body elevation phase. Compared with the control condition, exoskeleton assistance significantly reduced mean and peak knee joint moment under the 4.5 Nm and 6.0 Nm conditions (p
Conclusions
In healthy young male participants under controlled laboratory conditions, knee joint mechanical demand during STS was strongly phase-dependent, and assistance aligned with this temporal structure reduced knee joint loading. The pressure-based intention recognition method provided a practical approach for improving system responsiveness during this standardized task. However, these findings should be interpreted as preliminary biomechanical evidence obtained from healthy participants, rather than direct evidence of clinical efficacy. Further studies involving older adults, patients with functional impairments, and real-world testing conditions are required before clinical or rehabilitation applications can be inferred.