This study will integrate wireless wearable sensors, smartphone imaging, and multimodal artificial intelligence (AI) to address the rehabilitation needs of patients with lumbar degeneration. Patients will undergo comprehensive functional assessments, and individualized exercise instruction with real-time feedback will be provided through a smartphone application. The goals of this research are to: (1) develop a multimodal AI-based digital health system combining IMU sensors and smartphone cameras for real-time assessment and interactive rehabilitation training, (2) construct biomechanics- and gait-analysis models to support personalized rehabilitation for patients with lumbar degeneration, and (3) investigate the mechanisms and clinical efficacy of pelvic control exercise training combined with real-time smartphone feedback in improving function and quality of life for aging patients.