Wearable sensor-based healthcare systems have drawn a lot of attention from the scientific community and the industry during the past several years. As the healthcare cost is increasing and the population is aging, unobtrusive, low cost and accurate healthcare systems will potentially transform the future of healthcare. They are convenient to use, and can be utilized directly from home or community. These systems have various functionalities, such as bio-signal monitoring, signal analysis and biofeedback. They are consist of sub-systems including sensors, data aggregators, data processing units, connected by wired or wireless communication modules. The developing goal of the systems includes to increase the performance accuracy, to make the system smaller, to prolong the battery life and to lower the product cost. These systems are: (a) The Smart Glove with a 3D blooming flower for stroke rehabilitation. (b) The Smart Glove with a virtual piano for stroke rehabilitation. (c) The Smart Socks for activity analysis. (d) The Smart Shoes for rehabilitation of Cerebral Palsy. (e) The Smart Shoes for the treatment of Anorexia Nervosa. The major accomplishments of this thesis include the following four aspects: (1) The original idea of an intelligent system with an interactive virtual flower or a virtual piano for stroke rehabilitation. (2) The Smart Socks are proposed and designed for stroke rehabilitation. Comparing to traditional rehabilitation methods, it is convenient to use and very light weight. (3) The system designed for Cerebral Palsy with the Smart Shoe is the first intelligent wearable system designed specially for patient with this disease. The data analysis algorithms fit the specific feature of the patients with Cerebral Palsy. (4) We designed the Tweety System for the treatment of Anorexia Nervosa. It is a system includes a smart phone for usage in the hand the patient and the signal can be sent to a remote hospital for analysis in real time. The system can monitor fidgeting movement of the patient and can send biofeedback in realtime. In conclusion, this thesis provides different design methods and results for rehabilitation sensor systems. Further it propose how we prepare to further extend these systems.