顏面神經麻痺的病患，透過臉部復健運動可以促進臉部血液循環，刺激麻痺肌肉，從而改善病患的症狀，甚至達到完全痊癒的效果。除了固定到診所在復健師協助下進行復健動作外，患者在家自我復健也相當重要。因此，如何運用先進資訊科技的輔助，讓顏面神經麻痺患者能夠在家中隨時進行自我復健療程，以提升復健的方便性和正確性是一個值得探討的議題。 本研究以AI臉部辨識技術和深度學習演算法為主軸，結合機器學習、TensorFlow、Dlib、MediaPipe、OpenCV、Google文字轉語音和Pillow影像處理套件等技術，開發了一套輔助顏面神經麻痺患者在家自主復健和復健師追蹤治療的互動式復健系統。該系統通過AI臉部辨識，與預先建立的機器學習復健動作訓練模型進行比對，並即時以語音和文字的方式提供患者改善復健動作的指導，幫助患者及時調整以達到最佳的復健效果。同時，系統將復健者自我復建過程中的重要資訊記錄在雲端資料庫中，協助復健師定期追蹤和觀察患者的自我復健過程，並適時調整患者的復健計劃。 本研究完成的系統可以協助顏面神經麻痺患者隨時隨地進行自我復健運動，透過系統的語音提示，正確地完成復健師指示的復健療程，實現復健的智慧化。同時，系統提供給復健師的後端功能，提高在家自我復健的成效和完整性，讓復健師可以遠端監控顏面神經麻痺患者在家自我復健的療程。

The peripheral facial palsy patients can improve their symptoms and even achieve complete recovery by promoting facial blood circulation and stimulating paralyzed muscles through facial rehabilitation exercises. In addition to receiving rehabilitation exercises with the assistance of a rehabilitation therapist at a clinic, self-rehabilitation at home is also crucial for patients. Exploring how advanced information technology can be utilized to provide facial nerve palsy patients with the ability to perform self-rehabilitation at home, thereby enhancing the convenience and accuracy of their rehabilitation process, is an important topic worth investigating. This study focuses on AI facial recognition technology and deep learning algorithms, combined with machine learning, TensorFlow, Dlib, MediaPipe, OpenCV, Google Text-to-Speech, and Pillow image processing libraries, to develop an interactive rehabilitation system that assists peripheral facial palsy patients in self-rehabilitation at home while enabling remote monitoring by rehabilitation therapists. The study involves AI facial recognition for peripheral facial palsy patients, comparing their facial movements with a pre-established training model of rehabilitative exercises using machine learning. Real-time feedback is provided to the patients through voice and text, helping them make immediate adjustments for optimal rehabilitation outcomes. Additionally, important rehabilitation information during the self-rehabilitation process is recorded in a cloud-based database, facilitating regular monitoring and adjustment of the patient's rehabilitation plan by the rehabilitation therapist. The developed system allows peripheral facial palsy patients to engage in self-rehabilitation exercises anytime and anywhere, with voice prompts guiding them to accurately perform the prescribed exercises. The system's backend functionality enables rehabilitation therapists to monitor and observe the patient's self- rehabilitation process remotely, thereby improving the effectiveness and comprehensiveness of home-based self-rehabilitation.