关键词: 脊柱侧弯, 筛查设备, 间接飞行时间深度成像, 动态中线, 躯干旋转角, 足底压力, 表面肌电图, 多模态融合, 风险分级

Abstract: Objective To develop a multimodal, radiation-free screening device combining back depth imaging, plantar pressure monitoring, and back surface electromyography, to address current issues in adolescent idiopathic scoliosis （AIS） screening, including high X-ray irradiation, strong subjectivity of handheld devices, and lack of functional assessment. Methods An indirect time of flight （iToF） depth camera for 3D point cloud of the back, a pressure-sensing balance platform, and a high-sampling surface electromyography module were integrated into a hardware architecture with unified triggering and timestamps for multimodal synchronous acquisition at millisecond level. The software pipeline proposed a dynamic midline tracking and angle of trunk rotation （ATR） estimation algorithm based on minimum energy method, and constructed an adaptive risk grading model by integrating centre of pressure （COP） stability and the symmetry index （SI） of electromyography. Results The system synchronously acquired three-modal data at 30 fps, and the error between ATR and standard instrument measurements was less than 2°. The fusion algorithm effectively reduced false positive rates in the case of posture disturbance with COP shift > 10 mm, and a complete screening workflow was achieved within 90 s. Conclusion The proposed quantitative and visualized multimodal scoliosis primary screening system could elevate screening efficiency and provide an objective functional basis for early risk warning and rehabilitation follow-up.

Key words: Scoliosis, Screening device, Indirect time of flight depth imaging, Dynamic midline, Angle of trunk rotation, Plantar pressure, Surface electromyography, Multimodal fusion, Risk grading